From 1ddf5e091932e6cc1c419f3745985cc847fa5840 Mon Sep 17 00:00:00 2001
From: danielinteractive <danielinteractive@users.noreply.github.com>
Date: Fri, 11 Oct 2024 14:16:02 +0000
Subject: [PATCH] deploy: 9de9a7e0bc49e6129201f38360c830ecbf07db82

---
 main/coverage-report/index.html | 25163 +++++++++++++++---------------
 1 file changed, 12641 insertions(+), 12522 deletions(-)

diff --git a/main/coverage-report/index.html b/main/coverage-report/index.html
index 591be7781..8294ff1a7 100644
--- a/main/coverage-report/index.html
+++ b/main/coverage-report/index.html
@@ -95,7 +95,7 @@
   font-size: 11px;
 }</style>
   <div class="col-md-8 col-md-offset-2">
-    <h2>mmrm coverage - 97.05%</h2>
+    <h2>mmrm coverage - 96.91%</h2>
     <div class="tabbable">
       <ul class="nav nav-tabs" data-tabsetid="covr">
         <li class="active">
@@ -107,8 +107,8 @@ <h2>mmrm coverage - 97.05%</h2>
       </ul>
       <div class="tab-content" data-tabsetid="covr">
         <div class="tab-pane active" title="Files" data-value="Files" id="tab-covr-1">
-          <div class="datatables html-widget html-fill-item" id="htmlwidget-fcc1214b9a26b602b500" style="width:100%;height:500px;"></div>
-          <script type="application/json" data-for="htmlwidget-fcc1214b9a26b602b500">{"x":{"filter":"none","vertical":false,"fillContainer":false,"data":[["<a href=\"#\">src/testthat-helpers.h<\/a>","<a href=\"#\">R/zzz.R<\/a>","<a href=\"#\">src/test-chol_cache.cpp<\/a>","<a href=\"#\">R/satterthwaite.R<\/a>","<a href=\"#\">src/test-derivatives.cpp<\/a>","<a href=\"#\">R/testing.R<\/a>","<a href=\"#\">R/utils.R<\/a>","<a href=\"#\">src/test-covariance.cpp<\/a>","<a href=\"#\">src/test-utils.cpp<\/a>","<a href=\"#\">R/tidiers.R<\/a>","<a href=\"#\">R/kenwardroger.R<\/a>","<a href=\"#\">src/jacobian.cpp<\/a>","<a href=\"#\">R/interop-parsnip.R<\/a>","<a href=\"#\">R/fit.R<\/a>","<a href=\"#\">R/cov_struct.R<\/a>","<a href=\"#\">R/tmb-methods.R<\/a>","<a href=\"#\">src/covariance.h<\/a>","<a href=\"#\">R/interop-car.R<\/a>","<a href=\"#\">R/tmb.R<\/a>","<a href=\"#\">R/mmrm-methods.R<\/a>","<a href=\"#\">src/derivatives.h<\/a>","<a href=\"#\">src/predict.cpp<\/a>","<a href=\"#\">src/utils.h<\/a>","<a href=\"#\">src/mmrm.cpp<\/a>","<a href=\"#\">src/empirical.cpp<\/a>","<a href=\"#\">R/component.R<\/a>","<a href=\"#\">src/chol_cache.h<\/a>","<a href=\"#\">R/between-within.R<\/a>","<a href=\"#\">src/kr_comp.cpp<\/a>","<a href=\"#\">src/exports.cpp<\/a>","<a href=\"#\">R/interop-emmeans.R<\/a>","<a href=\"#\">R/utils-formula.R<\/a>","<a href=\"#\">R/utils-nse.R<\/a>","<a href=\"#\">R/residual.R<\/a>","<a href=\"#\">R/skipping.R<\/a>","<a href=\"#\">R/empirical.R<\/a>"],[57,144,67,259,130,158,550,151,232,199,191,62,83,517,461,700,182,210,575,263,224,123,184,155,91,150,138,131,71,84,96,94,132,32,31,27],[13,61,54,103,101,59,195,103,167,65,89,45,49,199,85,267,91,117,270,127,115,90,75,75,69,58,55,53,53,46,40,26,15,8,8,7],[8,39,49,94,94,55,184,98,160,63,87,44,48,196,84,264,90,116,269,127,115,90,75,75,69,58,55,53,53,46,40,26,15,8,8,7],[5,22,5,9,7,4,11,5,7,2,2,1,1,3,1,3,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],["75","1","6","601","5","168","297","7","2","8","334","135982","1","105","245","2196","15615","23","210","14","7784","547178","1763504","3763123","39381","4077","167201","19","274891","4860","13","299","368","1","11","34"],["61.54%","63.93%","90.74%","91.26%","93.07%","93.22%","94.36%","95.15%","95.81%","96.92%","97.75%","97.78%","97.96%","98.49%","98.82%","98.88%","98.90%","99.15%","99.63%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%"]],"container":"<table class=\"row-border\">\n  <thead>\n    <tr>\n      <th>File<\/th>\n      <th>Lines<\/th>\n      <th>Relevant<\/th>\n      <th>Covered<\/th>\n      <th>Missed<\/th>\n      <th>Hits / Line<\/th>\n      <th>Coverage<\/th>\n    <\/tr>\n  <\/thead>\n<\/table>","options":{"searching":false,"dom":"t","paging":false,"columnDefs":[{"targets":6,"createdCell":"function(td, cellData, rowData, row, col) {\n  var percent = cellData.replace(\"%\", \"\");\n  if (percent > 90) {\n    var grad = \"linear-gradient(90deg, #edfde7 \" + cellData + \", white \" + cellData + \")\";\n  } else if (percent > 75) {\n    var grad = \"linear-gradient(90deg, #f9ffe5 \" + cellData + \", white \" + cellData + \")\";\n  } else {\n    var grad = \"linear-gradient(90deg, #fcece9 \" + cellData + \", white \" + cellData + \")\";\n  }\n  $(td).css(\"background\", grad);\n}\n"},{"className":"dt-right","targets":[1,2,3,4]},{"name":"File","targets":0},{"name":"Lines","targets":1},{"name":"Relevant","targets":2},{"name":"Covered","targets":3},{"name":"Missed","targets":4},{"name":"Hits / Line","targets":5},{"name":"Coverage","targets":6}],"order":[],"autoWidth":false,"orderClasses":false},"callback":"function(table) {\ntable.on('click.dt', 'a', function() {\n  files = $('div#files div');\n  files.not('div.hidden').addClass('hidden');\n  id = $(this).text();\n  files.filter('div[id=\\'' + id + '\\']').removeClass('hidden');\n  $('ul.nav a[data-value=Source]').text(id).tab('show');\n});\n}"},"evals":["options.columnDefs.0.createdCell","callback"],"jsHooks":[]}</script>
+          <div class="datatables html-widget html-fill-item" id="htmlwidget-5cb8e6107df003a33a13" style="width:100%;height:500px;"></div>
+          <script type="application/json" data-for="htmlwidget-5cb8e6107df003a33a13">{"x":{"filter":"none","vertical":false,"fillContainer":false,"data":[["<a href=\"#\">R/zzz.R<\/a>","<a href=\"#\">src/testthat-helpers.h<\/a>","<a href=\"#\">src/test-chol_cache.cpp<\/a>","<a href=\"#\">R/satterthwaite.R<\/a>","<a href=\"#\">src/test-derivatives.cpp<\/a>","<a href=\"#\">R/testing.R<\/a>","<a href=\"#\">R/utils.R<\/a>","<a href=\"#\">src/test-covariance.cpp<\/a>","<a href=\"#\">src/test-utils.cpp<\/a>","<a href=\"#\">R/tidiers.R<\/a>","<a href=\"#\">R/kenwardroger.R<\/a>","<a href=\"#\">src/jacobian.cpp<\/a>","<a href=\"#\">R/interop-parsnip.R<\/a>","<a href=\"#\">R/fit.R<\/a>","<a href=\"#\">R/cov_struct.R<\/a>","<a href=\"#\">R/tmb-methods.R<\/a>","<a href=\"#\">src/covariance.h<\/a>","<a href=\"#\">R/interop-car.R<\/a>","<a href=\"#\">R/tmb.R<\/a>","<a href=\"#\">R/mmrm-methods.R<\/a>","<a href=\"#\">src/derivatives.h<\/a>","<a href=\"#\">src/predict.cpp<\/a>","<a href=\"#\">src/utils.h<\/a>","<a href=\"#\">src/mmrm.cpp<\/a>","<a href=\"#\">src/empirical.cpp<\/a>","<a href=\"#\">R/component.R<\/a>","<a href=\"#\">src/chol_cache.h<\/a>","<a href=\"#\">R/between-within.R<\/a>","<a href=\"#\">src/kr_comp.cpp<\/a>","<a href=\"#\">src/exports.cpp<\/a>","<a href=\"#\">R/interop-emmeans.R<\/a>","<a href=\"#\">R/utils-formula.R<\/a>","<a href=\"#\">R/utils-nse.R<\/a>","<a href=\"#\">R/skipping.R<\/a>","<a href=\"#\">R/residual.R<\/a>","<a href=\"#\">R/empirical.R<\/a>"],[149,57,67,259,130,158,562,151,232,199,191,62,83,517,461,700,182,210,575,263,224,123,184,155,91,150,138,131,71,84,96,94,132,31,32,27],[65,13,54,103,101,59,198,103,167,65,89,45,49,199,85,267,91,117,270,127,115,90,75,75,69,58,55,53,53,46,40,26,15,8,8,7],[39,8,49,94,94,55,187,98,160,63,87,44,48,196,84,264,90,116,269,127,115,90,75,75,69,58,55,53,53,46,40,26,15,8,8,7],[26,5,5,9,7,4,11,5,7,2,2,1,1,3,1,3,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],["1","75","6","601","5","168","295","7","2","8","334","135982","1","105","245","2196","15615","23","210","14","7784","547178","1763504","3763123","39381","4077","167201","19","274891","4860","13","299","368","11","1","34"],["60.00%","61.54%","90.74%","91.26%","93.07%","93.22%","94.44%","95.15%","95.81%","96.92%","97.75%","97.78%","97.96%","98.49%","98.82%","98.88%","98.90%","99.15%","99.63%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%","100.00%"]],"container":"<table class=\"row-border\">\n  <thead>\n    <tr>\n      <th>File<\/th>\n      <th>Lines<\/th>\n      <th>Relevant<\/th>\n      <th>Covered<\/th>\n      <th>Missed<\/th>\n      <th>Hits / Line<\/th>\n      <th>Coverage<\/th>\n    <\/tr>\n  <\/thead>\n<\/table>","options":{"searching":false,"dom":"t","paging":false,"columnDefs":[{"targets":6,"createdCell":"function(td, cellData, rowData, row, col) {\n  var percent = cellData.replace(\"%\", \"\");\n  if (percent > 90) {\n    var grad = \"linear-gradient(90deg, #edfde7 \" + cellData + \", white \" + cellData + \")\";\n  } else if (percent > 75) {\n    var grad = \"linear-gradient(90deg, #f9ffe5 \" + cellData + \", white \" + cellData + \")\";\n  } else {\n    var grad = \"linear-gradient(90deg, #fcece9 \" + cellData + \", white \" + cellData + \")\";\n  }\n  $(td).css(\"background\", grad);\n}\n"},{"className":"dt-right","targets":[1,2,3,4]},{"name":"File","targets":0},{"name":"Lines","targets":1},{"name":"Relevant","targets":2},{"name":"Covered","targets":3},{"name":"Missed","targets":4},{"name":"Hits / Line","targets":5},{"name":"Coverage","targets":6}],"order":[],"autoWidth":false,"orderClasses":false},"callback":"function(table) {\ntable.on('click.dt', 'a', function() {\n  files = $('div#files div');\n  files.not('div.hidden').addClass('hidden');\n  id = $(this).text();\n  files.filter('div[id=\\'' + id + '\\']').removeClass('hidden');\n  $('ul.nav a[data-value=Source]').text(id).tab('show');\n});\n}"},"evals":["options.columnDefs.0.createdCell","callback"],"jsHooks":[]}</script>
         </div>
         <div class="tab-pane" title="Source" data-value="Source" id="tab-covr-2">
           <div id="files">
@@ -3737,14 +3737,14 @@ <h2>mmrm coverage - 97.05%</h2>
                 </tbody>
               </table>
             </div>
-            <div id="R/utils.R" class="hidden">
+            <div id="R/testing.R" class="hidden">
               <table class="table-condensed">
                 <tbody>
                   <tr class="never">
                     <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Capture all Output</pre>
+                      <pre class="language-r">#' Calculation of Degrees of Freedom for One-Dimensional Contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -3758,21 +3758,21 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This function silences all warnings, errors &amp; messages and instead returns a list</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' containing the results (if it didn't error), as well as the warnings, errors</pre>
+                      <pre class="language-r">#' Calculates the estimate, adjusted standard error, degrees of freedom,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' and messages and divergence signals as character vectors.</pre>
+                      <pre class="language-r">#' t statistic and p-value for one-dimensional contrast.</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -3786,483 +3786,483 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param expr (`expression`)\cr to be executed.</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param remove (`list`)\cr optional list with elements `warnings`, `errors`,</pre>
+                      <pre class="language-r">#' @param contrast (`numeric`)\cr contrast vector. Note that this should not include</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `messages` which can be character vectors, which will be removed from the</pre>
+                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   results if specified.</pre>
+                      <pre class="language-r">#'   actually estimated coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param divergence (`list`)\cr optional list similar as `remove`, but these</pre>
+                      <pre class="language-r">#' @return List with `est`, `se`, `df`, `t_stat` and `p_val`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">12</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   character vectors will be moved to the `divergence` result and signal</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   that the fit did not converge.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return</pre>
+                      <pre class="language-r">#' object &lt;- mmrm(</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' A list containing</pre>
+                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">17</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   data = fev_data</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">18</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `result`: The object returned by `expr` or `list()` if an error was thrown.</pre>
+                      <pre class="language-r">#' )</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `warnings`: `NULL` or a character vector if warnings were thrown.</pre>
+                      <pre class="language-r">#' contrast &lt;- numeric(length(object$beta_est))</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `errors`: `NULL` or a string if an error was thrown.</pre>
+                      <pre class="language-r">#' contrast[3] &lt;- 1</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `messages`: `NULL` or a character vector if messages were produced.</pre>
+                      <pre class="language-r">#' df_1d(object, contrast)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">22</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `divergence`: `NULL` or a character vector if divergence messages were caught.</pre>
+                      <pre class="language-r">df_1d &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">23</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">338<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">24</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">338<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")), any.missing = FALSE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">25</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">338<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_record_all_output &lt;- function(expr,</pre>
+                      <pre class="language-r">  contrast &lt;- as.vector(contrast)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">26</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">338<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                remove = list(),</pre>
+                      <pre class="language-r">  switch(object$method,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">27</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">318<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                divergence = list()) {</pre>
+                      <pre class="language-r">    "Satterthwaite" = h_df_1d_sat(object, contrast),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">28</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Note: We don't need to and cannot assert `expr` here.</pre>
+                      <pre class="language-r">    "Kenward-Roger" = h_df_1d_kr(object, contrast),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="missed">
                     <td class="num">29</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(remove, types = "character")</pre>
+                      <pre class="language-r">    "Residual" = h_df_1d_res(object, contrast),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">30</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(divergence, types = "character")</pre>
+                      <pre class="language-r">    "Between-Within" = h_df_1d_bw(object, contrast),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="missed">
                     <td class="num">31</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  env &lt;- new.env()</pre>
+                      <pre class="language-r">    stop("Unrecognized degrees of freedom method: ", object$method)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">32</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  result &lt;- withCallingHandlers(</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">33</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    withRestarts(</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">34</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      expr,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">35</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      muffleStop = function(e) structure(e$message, class = "try-error")</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ),</pre>
+                      <pre class="language-r">#' Calculation of Degrees of Freedom for Multi-Dimensional Contrast</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">37</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    message = function(m) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">38</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      msg_without_newline &lt;- gsub(m$message, pattern = "\n$", replacement = "")</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">39</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      env$message &lt;- c(env$message, msg_without_newline)</pre>
+                      <pre class="language-r">#' Calculates the estimate, standard error, degrees of freedom,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">40</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      invokeRestart("muffleMessage")</pre>
+                      <pre class="language-r">#' t statistic and p-value for one-dimensional contrast, depending on the method</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">41</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    },</pre>
+                      <pre class="language-r">#' used in [mmrm()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">42</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning = function(w) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">43</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      env$warning &lt;- c(env$warning, w$message)</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">44</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      invokeRestart("muffleWarning")</pre>
+                      <pre class="language-r">#' @param contrast (`matrix`)\cr numeric contrast matrix, if given a `numeric`</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">45</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    },</pre>
+                      <pre class="language-r">#'   then this is coerced to a row vector. Note that this should not include</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">46</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    error = function(e) {</pre>
+                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">47</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      env$error &lt;- c(env$error, e$message)</pre>
+                      <pre class="language-r">#'   actually estimated coefficients.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">48</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      invokeRestart("muffleStop", e)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#' @return List with `num_df`, `denom_df`, `f_stat` and `p_val` (2-sided p-value).</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">51</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">52</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    result = result,</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">53</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warnings = setdiff(env$warning, c(remove$warnings, divergence$warnings)),</pre>
+                      <pre class="language-r">#' object &lt;- mmrm(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">54</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    errors = setdiff(env$error, c(remove$errors, divergence$errors)),</pre>
+                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">55</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    messages = setdiff(env$message, c(remove$messages, divergence$messages)),</pre>
+                      <pre class="language-r">#'   data = fev_data</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">56</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    divergence = c(</pre>
+                      <pre class="language-r">#' )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">57</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      intersect(env$warning, divergence$warnings),</pre>
+                      <pre class="language-r">#' contrast &lt;- matrix(data = 0, nrow = 2, ncol = length(object$beta_est))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">58</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      intersect(env$error, divergence$errors),</pre>
+                      <pre class="language-r">#' contrast[1, 2] &lt;- contrast[2, 3] &lt;- 1</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">59</td>
-                    <td class="coverage">201<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      intersect(env$message, divergence$messages)</pre>
+                      <pre class="language-r">#' df_md(object, contrast)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">60</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">df_md &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">61</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">62</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  assert_numeric(contrast, any.missing = FALSE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">63</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  if (!is.matrix(contrast)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">64</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">113<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Trace of a Matrix</pre>
+                      <pre class="language-r">    contrast &lt;- matrix(contrast, ncol = length(contrast))</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">65</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">66</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtain the trace of a matrix if the matrix is diagonal, otherwise raise an error.</pre>
+                      <pre class="language-r">  assert_matrix(contrast, ncols = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">67</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  if (nrow(contrast) == 0) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">68</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`matrix`)\cr square matrix input.</pre>
+                      <pre class="language-r">    return(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">69</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">70</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The trace of the square matrix.</pre>
+                      <pre class="language-r">        num_df = 0,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">71</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        denom_df = NA_real_,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">72</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">        f_stat = NA_real_,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">73</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_tr &lt;- function(x) {</pre>
+                      <pre class="language-r">        p_val = NA_real_</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">74</td>
-                    <td class="coverage">1790<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (nrow(x) != ncol(x)) {</pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">75</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("x must be square matrix")</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -4274,72 +4274,72 @@ <h2>mmrm coverage - 97.05%</h2>
                   </tr>
                   <tr class="covered">
                     <td class="num">77</td>
-                    <td class="coverage">1789<em>x</em></td>
+                    <td class="coverage">149<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  sum(Matrix::diag(x))</pre>
+                      <pre class="language-r">  switch(object$method,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">78</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">145<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    "Satterthwaite" = h_df_md_sat(object, contrast),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">79</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    "Kenward-Roger" = h_df_md_kr(object, contrast),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="missed">
                     <td class="num">80</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Split Control List</pre>
+                      <pre class="language-r">    "Residual" = h_df_md_res(object, contrast),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">81</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    "Between-Within" = h_df_md_bw(object, contrast),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="missed">
                     <td class="num">82</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Split the [mmrm_control()] object according to its optimizers and use additional arguments</pre>
+                      <pre class="language-r">    stop("Unrecognized degrees of freedom method: ", object$method)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' to replace the elements in the original object.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">84</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param control (`mmrm_control`)\cr object.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">86</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... additional parameters to update the `control` object.</pre>
+                      <pre class="language-r">#' Creating T-Statistic Test Results For One-Dimensional Contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -4353,8808 +4353,8465 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A `list` of `mmrm_control` entries.</pre>
+                      <pre class="language-r">#' @description Creates a list of results for one-dimensional contrasts using</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' a t-test statistic and the given degrees of freedom.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">90</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_split_control &lt;- function(control, ...) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">91</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(control, "mmrm_control")</pre>
+                      <pre class="language-r">#' @inheritParams df_1d</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">92</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  l &lt;- length(control$optimizers)</pre>
+                      <pre class="language-r">#' @param df (`number`)\cr degrees of freedom for the one-dimensional contrast.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">93</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  lapply(seq_len(l), function(i) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">94</td>
-                    <td class="coverage">22<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ret &lt;- utils::modifyList(control, list(...))</pre>
+                      <pre class="language-r">#' @return List with `est`, `se`, `df`, `t_stat` and `p_val` (2-sided p-value).</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">95</td>
-                    <td class="coverage">22<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ret$optimizers &lt;- control$optimizers[i]</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">96</td>
-                    <td class="coverage">22<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ret</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">97</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
+                      <pre class="language-r">h_test_1d &lt;- function(object,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">98</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">                      contrast,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">99</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">                      df) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">100</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain Optimizer according to Optimizer String Value</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">101</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">102</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description This function creates optimizer functions with arguments.</pre>
+                      <pre class="language-r">  assert_number(df, lower = .Machine$double.xmin)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">103</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">104</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param optimizer (`character`)\cr names of built-in optimizers to try, subset</pre>
+                      <pre class="language-r">  est &lt;- sum(contrast * component(object, "beta_est"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">105</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   of "L-BFGS-B", "BFGS", "CG" and "nlminb".</pre>
+                      <pre class="language-r">  var &lt;- h_quad_form_vec(contrast, component(object, "beta_vcov"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">106</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param optimizer_fun (`function` or `list` of `function`)\cr alternatively to `optimizer`,</pre>
+                      <pre class="language-r">  se &lt;- sqrt(var)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">107</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   an optimizer function or a list of optimizer functions can be passed directly here.</pre>
+                      <pre class="language-r">  t_stat &lt;- est / se</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">108</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param optimizer_args (`list`)\cr additional arguments for `optimizer_fun`.</pre>
+                      <pre class="language-r">  p_val &lt;- 2 * stats::pt(q = abs(t_stat), df = df, lower.tail = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">109</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param optimizer_control (`list`)\cr passed to argument `control` in `optimizer_fun`.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">110</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">111</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">    est = est,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">112</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' If you want to use only the built-in optimizers:</pre>
+                      <pre class="language-r">    se = se,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">113</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `optimizer` is a shortcut to create a list of built-in optimizer functions</pre>
+                      <pre class="language-r">    df = df,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">114</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   passed to `optimizer_fun`.</pre>
+                      <pre class="language-r">    t_stat = t_stat,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">115</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">486<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - Allowed are "L-BFGS-B", "BFGS", "CG" (using [stats::optim()] with corresponding method)</pre>
+                      <pre class="language-r">    p_val = p_val</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">116</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   and "nlminb" (using [stats::nlminb()]).</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">117</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - Other arguments should go into `optimizer_args`.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">118</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">119</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' If you want to use your own optimizer function:</pre>
+                      <pre class="language-r">#' Creating F-Statistic Test Results For Multi-Dimensional Contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">120</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - Make sure that there are three arguments: parameter (start value), objective function</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">121</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   and gradient function are sequentially in the function arguments.</pre>
+                      <pre class="language-r">#' @description Creates a list of results for multi-dimensional contrasts using</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">122</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - If there are other named arguments in front of these, make sure they are correctly</pre>
+                      <pre class="language-r">#' an F-test statistic and the given degrees of freedom.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">123</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   specified through `optimizer_args`.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">124</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - If the hessian can be used, please make sure its argument name is `hessian` and</pre>
+                      <pre class="language-r">#' @inheritParams df_md</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">125</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   please add attribute `use_hessian = TRUE` to the function,</pre>
+                      <pre class="language-r">#' @param contrast (`matrix`)\cr numeric contrast matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">126</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   using `attr(fun, "use_hessian) &lt;- TRUE`.</pre>
+                      <pre class="language-r">#' @param df (`number`)\cr denominator degrees of freedom for the multi-dimensional contrast.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">127</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param f_stat_factor (`number`)\cr optional scaling factor on top of the standard F-statistic.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">128</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Named `list` of optimizers created by [h_partial_fun_args()].</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">129</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @return List with `num_df`, `denom_df`, `f_stat` and `p_val` (2-sided p-value).</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">130</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">131</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_optimizers &lt;- function(optimizer = c("L-BFGS-B", "BFGS", "CG", "nlminb"),</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">132</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             optimizer_fun = h_optimizer_fun(optimizer),</pre>
+                      <pre class="language-r">h_test_md &lt;- function(object,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">133</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             optimizer_args = list(),</pre>
+                      <pre class="language-r">                      contrast,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">134</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             optimizer_control = list()) {</pre>
+                      <pre class="language-r">                      df,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">135</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if ("automatic" %in% optimizer) {</pre>
+                      <pre class="language-r">                      f_stat_factor = 1) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">136</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    lifecycle::deprecate_warn(</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">137</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      when = "0.2.0",</pre>
+                      <pre class="language-r">  assert_matrix(contrast, ncols = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">138</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      what = I("\"automatic\" optimizer"),</pre>
+                      <pre class="language-r">  num_df &lt;- nrow(contrast)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">139</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      details = "please just omit optimizer argument"</pre>
+                      <pre class="language-r">  assert_number(df, lower = .Machine$double.xmin)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">140</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">  assert_number(f_stat_factor, lower = .Machine$double.xmin)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">141</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    optimizer_fun &lt;- h_optimizer_fun()</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">142</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  prec_contrast &lt;- solve(h_quad_form_mat(contrast, component(object, "beta_vcov")))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">143</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert(</pre>
+                      <pre class="language-r">  contrast_est &lt;- component(object, "beta_est") %*% t(contrast)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">144</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    test_function(optimizer_fun),</pre>
+                      <pre class="language-r">  f_statistic &lt;- as.numeric(f_stat_factor / num_df * h_quad_form_mat(contrast_est, prec_contrast))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">145</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    test_list(optimizer_fun, types = "function", names = "unique")</pre>
+                      <pre class="language-r">  p_val &lt;- stats::pf(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">146</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    q = f_statistic,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">147</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.function(optimizer_fun)) {</pre>
+                      <pre class="language-r">    df1 = num_df,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">148</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    optimizer_fun &lt;- list(custom_optimizer = optimizer_fun)</pre>
+                      <pre class="language-r">    df2 = df,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">149</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    lower.tail = FALSE</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">150</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  lapply(optimizer_fun, function(x) {</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">151</td>
-                    <td class="coverage">924<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    do.call(h_partial_fun_args, c(list(fun = x, control = optimizer_control), optimizer_args))</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">152</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
+                      <pre class="language-r">  list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">153</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    num_df = num_df,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">154</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    denom_df = df,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">155</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain Optimizer Function with Character</pre>
+                      <pre class="language-r">    f_stat = f_statistic,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">156</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtain the optimizer function through the character provided.</pre>
+                      <pre class="language-r">    p_val = p_val</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">157</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param optimizer (`character`)\cr vector of optimizers.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">158</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/utils.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">159</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A (`list`)\cr of optimizer functions generated from [h_partial_fun_args()].</pre>
+                      <pre class="language-r">#' Capture all Output</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">160</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">161</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_optimizer_fun &lt;- function(optimizer = c("L-BFGS-B", "BFGS", "CG", "nlminb")) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">162</td>
-                    <td class="coverage">240<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  optimizer &lt;- match.arg(optimizer, several.ok = TRUE)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">163</td>
-                    <td class="coverage">240<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  lapply(stats::setNames(optimizer, optimizer), function(x) {</pre>
+                      <pre class="language-r">#' This function silences all warnings, errors &amp; messages and instead returns a list</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">164</td>
-                    <td class="coverage">920<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">4</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    switch(x,</pre>
+                      <pre class="language-r">#' containing the results (if it didn't error), as well as the warnings, errors</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">165</td>
-                    <td class="coverage">229<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">5</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "L-BFGS-B" = h_partial_fun_args(fun = stats::optim, method = x),</pre>
+                      <pre class="language-r">#' and messages and divergence signals as character vectors.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">166</td>
-                    <td class="coverage">230<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">6</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "BFGS" = h_partial_fun_args(fun = stats::optim, method = x),</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">167</td>
-                    <td class="coverage">228<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">7</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "CG" = h_partial_fun_args(fun = stats::optim, method = x),</pre>
+                      <pre class="language-r">#' @param expr (`expression`)\cr to be executed.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">168</td>
-                    <td class="coverage">233<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">8</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "nlminb" = h_partial_fun_args(fun = stats::nlminb, additional_attr = list(use_hessian = TRUE))</pre>
+                      <pre class="language-r">#' @param remove (`list`)\cr optional list with elements `warnings`, `errors`,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">169</td>
+                    <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#'   `messages` which can be character vectors, which will be removed from the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">170</td>
+                    <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
+                      <pre class="language-r">#'   results if specified.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">171</td>
+                    <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @param divergence (`list`)\cr optional list similar as `remove`, but these</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">172</td>
+                    <td class="num">12</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   character vectors will be moved to the `divergence` result and signal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">173</td>
+                    <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Create Partial Functions</pre>
+                      <pre class="language-r">#'   that the fit did not converge.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">174</td>
+                    <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Creates partial functions with arguments.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">175</td>
+                    <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @return</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">176</td>
+                    <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param fun (`function`)\cr to be wrapped.</pre>
+                      <pre class="language-r">#' A list containing</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">177</td>
+                    <td class="num">17</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... Additional arguments for `fun`.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">178</td>
+                    <td class="num">18</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param additional_attr (`list`)\cr of additional attributes to apply to the result.</pre>
+                      <pre class="language-r">#' - `result`: The object returned by `expr` or `list()` if an error was thrown.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">179</td>
+                    <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `warnings`: `NULL` or a character vector if warnings were thrown.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">180</td>
+                    <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details This function add `args` attribute to the original function,</pre>
+                      <pre class="language-r">#' - `errors`: `NULL` or a string if an error was thrown.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">181</td>
+                    <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' and add an extra class `partial` to the function.</pre>
+                      <pre class="language-r">#' - `messages`: `NULL` or a character vector if messages were produced.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">182</td>
+                    <td class="num">22</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `args` is the argument for the function, and elements in `...` will override the existing</pre>
+                      <pre class="language-r">#' - `divergence`: `NULL` or a character vector if divergence messages were caught.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">183</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' arguments in attribute `args`. `additional_attr` will override the existing attributes.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">184</td>
+                    <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">185</td>
+                    <td class="num">25</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Object with S3 class `"partial"`, a `function` with `args` attribute (and possibly more</pre>
+                      <pre class="language-r">h_record_all_output &lt;- function(expr,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">186</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' attributes from `additional_attr`).</pre>
+                      <pre class="language-r">                                remove = list(),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">187</td>
+                    <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">                                divergence = list()) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">188</td>
+                    <td class="num">28</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_partial_fun_args &lt;- function(fun, ..., additional_attr = list()) {</pre>
+                      <pre class="language-r">  # Note: We don't need to and cannot assert `expr` here.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">189</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">29</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_function(fun)</pre>
+                      <pre class="language-r">  assert_list(remove, types = "character")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">190</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">30</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(additional_attr, names = "unique")</pre>
+                      <pre class="language-r">  assert_list(divergence, types = "character")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">191</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">31</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  a_args &lt;- list(...)</pre>
+                      <pre class="language-r">  env &lt;- new.env()</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">192</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">32</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(a_args, names = "unique")</pre>
+                      <pre class="language-r">  result &lt;- withCallingHandlers(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">193</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">33</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  args &lt;- attr(fun, "args")</pre>
+                      <pre class="language-r">    withRestarts(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">194</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">34</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.null(args)) {</pre>
+                      <pre class="language-r">      expr,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">195</td>
-                    <td class="coverage">932<em>x</em></td>
+                    <td class="num">35</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    args &lt;- list()</pre>
+                      <pre class="language-r">      muffleStop = function(e) structure(e$message, class = "try-error")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">196</td>
+                    <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    ),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">197</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">37</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  do.call(</pre>
+                      <pre class="language-r">    message = function(m) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">198</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">38</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    structure,</pre>
+                      <pre class="language-r">      msg_without_newline &lt;- gsub(m$message, pattern = "\n$", replacement = "")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">199</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">39</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    args = utils::modifyList(</pre>
+                      <pre class="language-r">      env$message &lt;- c(env$message, msg_without_newline)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">200</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">40</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      list(</pre>
+                      <pre class="language-r">      invokeRestart("muffleMessage")</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">41</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    },</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">201</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">42</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        .Data = fun,</pre>
+                      <pre class="language-r">    warning = function(w) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">202</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">43</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        args = utils::modifyList(args, a_args),</pre>
+                      <pre class="language-r">      env$warning &lt;- c(env$warning, w$message)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">203</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">44</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        class = c("partial", "function")</pre>
+                      <pre class="language-r">      invokeRestart("muffleWarning")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">204</td>
+                    <td class="num">45</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ),</pre>
+                      <pre class="language-r">    },</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">205</td>
-                    <td class="coverage">1848<em>x</em></td>
+                    <td class="num">46</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      additional_attr</pre>
+                      <pre class="language-r">    error = function(e) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">206</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">47</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">      env$error &lt;- c(env$error, e$message)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">207</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">48</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">      invokeRestart("muffleStop", e)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">208</td>
+                    <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">209</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">210</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">51</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain Default Covariance Method</pre>
+                      <pre class="language-r">  list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">211</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">52</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    result = result,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">212</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">53</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtain the default covariance method depending on</pre>
+                      <pre class="language-r">    warnings = setdiff(env$warning, c(remove$warnings, divergence$warnings)),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">213</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">54</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' the degrees of freedom method used.</pre>
+                      <pre class="language-r">    errors = setdiff(env$error, c(remove$errors, divergence$errors)),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">214</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">55</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    messages = setdiff(env$message, c(remove$messages, divergence$messages)),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">215</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">56</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param method (`string`)\cr degrees of freedom method.</pre>
+                      <pre class="language-r">    divergence = c(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">216</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">57</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      intersect(env$warning, divergence$warnings),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">217</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">58</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details The default covariance method is different for different degrees of freedom method.</pre>
+                      <pre class="language-r">      intersect(env$error, divergence$errors),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">218</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">59</td>
+                    <td class="coverage">201<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' For "Satterthwaite" or "Between-Within", "Asymptotic" is returned.</pre>
+                      <pre class="language-r">      intersect(env$message, divergence$messages)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">219</td>
+                    <td class="num">60</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' For "Kenward-Roger" only, "Kenward-Roger" is returned.</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">220</td>
+                    <td class="num">61</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' For "Residual" only, "Empirical" is returned.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">221</td>
+                    <td class="num">62</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">}</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">63</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">64</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' Trace of a Matrix</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">65</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">222</td>
+                    <td class="num">66</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return String of the default covariance method.</pre>
+                      <pre class="language-r">#' @description Obtain the trace of a matrix if the matrix is diagonal, otherwise raise an error.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">223</td>
+                    <td class="num">67</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">224</td>
+                    <td class="num">68</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_cov_default &lt;- function(method = c("Satterthwaite", "Kenward-Roger", "Residual", "Between-Within")) {</pre>
+                      <pre class="language-r">#' @param x (`matrix`)\cr square matrix input.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">225</td>
-                    <td class="coverage">197<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">69</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(method)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">226</td>
-                    <td class="coverage">197<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">70</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  method &lt;- match.arg(method)</pre>
+                      <pre class="language-r">#' @return The trace of the square matrix.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">227</td>
-                    <td class="coverage">196<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">71</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  switch(method,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">228</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">72</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Residual" = "Empirical",</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">229</td>
-                    <td class="coverage">158<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">73</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Satterthwaite" = "Asymptotic",</pre>
+                      <pre class="language-r">h_tr &lt;- function(x) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">230</td>
-                    <td class="coverage">35<em>x</em></td>
+                    <td class="num">74</td>
+                    <td class="coverage">1790<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Kenward-Roger" = "Kenward-Roger",</pre>
+                      <pre class="language-r">  if (nrow(x) != ncol(x)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">231</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">75</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Between-Within" = "Asymptotic"</pre>
+                      <pre class="language-r">    stop("x must be square matrix")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">232</td>
+                    <td class="num">76</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">77</td>
+                    <td class="coverage">1789<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  sum(Matrix::diag(x))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">233</td>
+                    <td class="num">78</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">234</td>
+                    <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">235</td>
+                    <td class="num">80</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Complete `character` Vector Names From Values</pre>
+                      <pre class="language-r">#' Split Control List</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">236</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">237</td>
+                    <td class="num">82</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`character` or `list`)\cr value whose names should be completed</pre>
+                      <pre class="language-r">#' @description Split the [mmrm_control()] object according to its optimizers and use additional arguments</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">238</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   from element values.</pre>
+                      <pre class="language-r">#' to replace the elements in the original object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">239</td>
+                    <td class="num">84</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">240</td>
+                    <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A named vector or list.</pre>
+                      <pre class="language-r">#' @param control (`mmrm_control`)\cr object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">241</td>
+                    <td class="num">86</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @param ... additional parameters to update the `control` object.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">87</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">242</td>
+                    <td class="num">88</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @return A `list` of `mmrm_control` entries.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">243</td>
+                    <td class="num">90</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">fill_names &lt;- function(x) {</pre>
+                      <pre class="language-r">h_split_control &lt;- function(control, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">244</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">91</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n &lt;- names(x)</pre>
+                      <pre class="language-r">  assert_class(control, "mmrm_control")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">245</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">92</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  is_unnamed &lt;- if (is.null(n)) rep_len(TRUE, length(x)) else n == ""</pre>
+                      <pre class="language-r">  l &lt;- length(control$optimizers)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">246</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">93</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  names(x)[is_unnamed] &lt;- x[is_unnamed]</pre>
+                      <pre class="language-r">  lapply(seq_len(l), function(i) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">247</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">94</td>
+                    <td class="coverage">22<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x</pre>
+                      <pre class="language-r">    ret &lt;- utils::modifyList(control, list(...))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">95</td>
+                    <td class="coverage">22<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    ret$optimizers &lt;- control$optimizers[i]</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">96</td>
+                    <td class="coverage">22<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    ret</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">248</td>
+                    <td class="num">97</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  })</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">98</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">249</td>
+                    <td class="num">99</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">250</td>
+                    <td class="num">100</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Drop Items from an Indexible</pre>
+                      <pre class="language-r">#' Obtain Optimizer according to Optimizer String Value</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">251</td>
+                    <td class="num">101</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">252</td>
+                    <td class="num">102</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Drop elements from an indexible object (`vector`, `list`, etc.).</pre>
+                      <pre class="language-r">#' @description This function creates optimizer functions with arguments.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">253</td>
+                    <td class="num">103</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">254</td>
+                    <td class="num">104</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x Any object that can be consumed by [seq_along()] and indexed by a</pre>
+                      <pre class="language-r">#' @param optimizer (`character`)\cr names of built-in optimizers to try, subset</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">255</td>
+                    <td class="num">105</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   logical vector of the same length.</pre>
+                      <pre class="language-r">#'   of "L-BFGS-B", "BFGS", "CG" and "nlminb".</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">256</td>
+                    <td class="num">106</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n (`integer`)\cr the number of terms to drop.</pre>
+                      <pre class="language-r">#' @param optimizer_fun (`function` or `list` of `function`)\cr alternatively to `optimizer`,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">257</td>
+                    <td class="num">107</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   an optimizer function or a list of optimizer functions can be passed directly here.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">258</td>
+                    <td class="num">108</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A subset of `x`.</pre>
+                      <pre class="language-r">#' @param optimizer_args (`list`)\cr additional arguments for `optimizer_fun`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">259</td>
+                    <td class="num">109</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @param optimizer_control (`list`)\cr passed to argument `control` in `optimizer_fun`.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">110</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">260</td>
+                    <td class="num">111</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">261</td>
+                    <td class="num">112</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">drop_elements &lt;- function(x, n) {</pre>
+                      <pre class="language-r">#' If you want to use only the built-in optimizers:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">262</td>
-                    <td class="coverage">819<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">113</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x[seq_along(x) &gt; n]</pre>
+                      <pre class="language-r">#' - `optimizer` is a shortcut to create a list of built-in optimizer functions</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">263</td>
+                    <td class="num">114</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'   passed to `optimizer_fun`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">264</td>
+                    <td class="num">115</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' - Allowed are "L-BFGS-B", "BFGS", "CG" (using [stats::optim()] with corresponding method)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">265</td>
+                    <td class="num">116</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Ask for Confirmation on Large Visit Levels</pre>
+                      <pre class="language-r">#'   and "nlminb" (using [stats::nlminb()]).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">266</td>
+                    <td class="num">117</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' - Other arguments should go into `optimizer_args`.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">118</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">267</td>
+                    <td class="num">119</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Ask the user for confirmation if there are too many visit levels</pre>
+                      <pre class="language-r">#' If you want to use your own optimizer function:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">268</td>
+                    <td class="num">120</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' for non-spatial covariance structure in interactive sessions.</pre>
+                      <pre class="language-r">#' - Make sure that there are three arguments: parameter (start value), objective function</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">269</td>
+                    <td class="num">121</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   and gradient function are sequentially in the function arguments.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">270</td>
+                    <td class="num">122</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`numeric`)\cr number of visit levels.</pre>
+                      <pre class="language-r">#' - If there are other named arguments in front of these, make sure they are correctly</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">271</td>
+                    <td class="num">123</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   specified through `optimizer_args`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">272</td>
+                    <td class="num">124</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Logical value `TRUE`.</pre>
+                      <pre class="language-r">#' - If the hessian can be used, please make sure its argument name is `hessian` and</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">273</td>
+                    <td class="num">125</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'   please add attribute `use_hessian = TRUE` to the function,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">274</td>
+                    <td class="num">126</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_confirm_large_levels &lt;- function(x) {</pre>
+                      <pre class="language-r">#'   using `attr(fun, "use_hessian) &lt;- TRUE`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">275</td>
-                    <td class="coverage">297<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">127</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(x)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">276</td>
-                    <td class="coverage">297<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">128</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  allowed_lvls &lt;- x &lt;= getOption("mmrm.max_visits", 100)</pre>
+                      <pre class="language-r">#' @return Named `list` of optimizers created by [h_partial_fun_args()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">277</td>
-                    <td class="coverage">297<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">129</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (allowed_lvls) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">278</td>
-                    <td class="coverage">295<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">130</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(TRUE)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">279</td>
+                    <td class="num">131</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">h_get_optimizers &lt;- function(optimizer = c("L-BFGS-B", "BFGS", "CG", "nlminb"),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">280</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">132</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!interactive()) {</pre>
+                      <pre class="language-r">                             optimizer_fun = h_optimizer_fun(optimizer),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">281</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">133</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("Visit levels too large!", call. = FALSE)</pre>
+                      <pre class="language-r">                             optimizer_args = list(),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">282</td>
+                    <td class="num">134</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">                             optimizer_control = list()) {</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">283</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">135</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  proceed &lt;- utils::askYesNo(</pre>
+                      <pre class="language-r">  if ("automatic" %in% optimizer) {</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">284</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">136</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    paste(</pre>
+                      <pre class="language-r">    lifecycle::deprecate_warn(</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">285</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">137</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Visit levels is possibly too large.",</pre>
+                      <pre class="language-r">      when = "0.2.0",</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">286</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">138</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "This requires large memory. Are you sure to continue?",</pre>
+                      <pre class="language-r">      what = I("\"automatic\" optimizer"),</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">287</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">139</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      collapse = " "</pre>
+                      <pre class="language-r">      details = "please just omit optimizer argument"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">288</td>
+                    <td class="num">140</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">    )</pre>
                     </td>
                   </tr>
+                  <tr class="covered">
+                    <td class="num">141</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    optimizer_fun &lt;- h_optimizer_fun()</pre>
+                    </td>
+                  </tr>
                   <tr class="never">
-                    <td class="num">289</td>
+                    <td class="num">142</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">290</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">143</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!identical(proceed, TRUE)) {</pre>
+                      <pre class="language-r">  assert(</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">291</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">144</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("Visit levels too large!", call. = FALSE)</pre>
+                      <pre class="language-r">    test_function(optimizer_fun),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">145</td>
+                    <td class="coverage">246<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    test_list(optimizer_fun, types = "function", names = "unique")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">292</td>
+                    <td class="num">146</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">293</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">147</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  return(TRUE)</pre>
+                      <pre class="language-r">  if (is.function(optimizer_fun)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">294</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">148</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    optimizer_fun &lt;- list(custom_optimizer = optimizer_fun)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">295</td>
+                    <td class="num">149</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">150</td>
+                    <td class="coverage">246<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  lapply(optimizer_fun, function(x) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">151</td>
+                    <td class="coverage">924<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    do.call(h_partial_fun_args, c(list(fun = x, control = optimizer_control), optimizer_args))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">296</td>
+                    <td class="num">152</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Default Value on NULL</pre>
+                      <pre class="language-r">  })</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">297</td>
+                    <td class="num">153</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Return default value when first argument is NULL.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">298</td>
+                    <td class="num">154</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">299</td>
+                    <td class="num">155</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x Object.</pre>
+                      <pre class="language-r">#' Obtain Optimizer Function with Character</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">300</td>
+                    <td class="num">156</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param y Object.</pre>
+                      <pre class="language-r">#' @description Obtain the optimizer function through the character provided.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">301</td>
+                    <td class="num">157</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param optimizer (`character`)\cr vector of optimizers.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">302</td>
+                    <td class="num">158</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details If `x` is NULL, returns `y`. Otherwise return `x`.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">303</td>
+                    <td class="num">159</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @return A (`list`)\cr of optimizer functions generated from [h_partial_fun_args()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">304</td>
+                    <td class="num">160</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">305</td>
+                    <td class="num">161</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_default_value &lt;- function(x, y) {</pre>
+                      <pre class="language-r">h_optimizer_fun &lt;- function(optimizer = c("L-BFGS-B", "BFGS", "CG", "nlminb")) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">306</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="num">162</td>
+                    <td class="coverage">240<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.null(x)) {</pre>
+                      <pre class="language-r">  optimizer &lt;- match.arg(optimizer, several.ok = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">307</td>
-                    <td class="coverage">277<em>x</em></td>
+                    <td class="num">163</td>
+                    <td class="coverage">240<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    y</pre>
+                      <pre class="language-r">  lapply(stats::setNames(optimizer, optimizer), function(x) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">308</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">164</td>
+                    <td class="coverage">920<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">    switch(x,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">309</td>
-                    <td class="coverage">35<em>x</em></td>
+                    <td class="num">165</td>
+                    <td class="coverage">229<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x</pre>
+                      <pre class="language-r">      "L-BFGS-B" = h_partial_fun_args(fun = stats::optim, method = x),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">310</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">166</td>
+                    <td class="coverage">230<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">      "BFGS" = h_partial_fun_args(fun = stats::optim, method = x),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">311</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">167</td>
+                    <td class="coverage">228<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      "CG" = h_partial_fun_args(fun = stats::optim, method = x),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">312</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">168</td>
+                    <td class="coverage">233<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      "nlminb" = h_partial_fun_args(fun = stats::nlminb, additional_attr = list(use_hessian = TRUE))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">313</td>
+                    <td class="num">169</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Warn on na.action</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">314</td>
+                    <td class="num">170</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  })</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">315</td>
+                    <td class="num">171</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_warn_na_action &lt;- function() {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">316</td>
-                    <td class="coverage">260<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  if (!identical(getOption("na.action"), "na.omit")) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">317</td>
-                    <td class="coverage">6<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    warning("na.action is always set to `na.omit` for `mmrm` fit!")</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">318</td>
+                    <td class="num">172</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">319</td>
+                    <td class="num">173</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' Create Partial Functions</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">320</td>
+                    <td class="num">174</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @description Creates partial functions with arguments.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">321</td>
+                    <td class="num">175</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain `na.action` as Function</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">322</td>
+                    <td class="num">176</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @param fun (`function`)\cr to be wrapped.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">323</td>
+                    <td class="num">177</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_na_action &lt;- function(na_action) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">324</td>
-                    <td class="coverage">56<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  if (is.function(na_action) &amp;&amp; identical(methods::formalArgs(na_action), c("object", "..."))) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">325</td>
-                    <td class="coverage">5<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    return(na_action)</pre>
+                      <pre class="language-r">#' @param ... Additional arguments for `fun`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">326</td>
+                    <td class="num">178</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @param additional_attr (`list`)\cr of additional attributes to apply to the result.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">327</td>
-                    <td class="coverage">51<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">179</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.character(na_action) &amp;&amp; length(na_action) == 1L) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">328</td>
-                    <td class="coverage">51<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">180</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_subset(na_action, c("na.omit", "na.exclude", "na.fail", "na.pass", "na.contiguous"))</pre>
+                      <pre class="language-r">#' @details This function add `args` attribute to the original function,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">329</td>
-                    <td class="coverage">51<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">181</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(get(na_action, mode = "function", pos = "package:stats"))</pre>
+                      <pre class="language-r">#' and add an extra class `partial` to the function.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">330</td>
+                    <td class="num">182</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' `args` is the argument for the function, and elements in `...` will override the existing</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">331</td>
+                    <td class="num">183</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' arguments in attribute `args`. `additional_attr` will override the existing attributes.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">332</td>
+                    <td class="num">184</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">333</td>
+                    <td class="num">185</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Validate mmrm Formula</pre>
+                      <pre class="language-r">#' @return Object with S3 class `"partial"`, a `function` with `args` attribute (and possibly more</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">334</td>
+                    <td class="num">186</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula (`formula`)\cr to check.</pre>
+                      <pre class="language-r">#' attributes from `additional_attr`).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">335</td>
+                    <td class="num">187</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">336</td>
+                    <td class="num">188</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details In mmrm models, `.` is not allowed as it introduces ambiguity of covariates</pre>
+                      <pre class="language-r">h_partial_fun_args &lt;- function(fun, ..., additional_attr = list()) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">337</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">189</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' to be used, so it is not allowed to be in formula.</pre>
+                      <pre class="language-r">  assert_function(fun)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">338</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">190</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_list(additional_attr, names = "unique")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">339</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">191</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  a_args &lt;- list(...)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">340</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">192</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_valid_formula &lt;- function(formula) {</pre>
+                      <pre class="language-r">  assert_list(a_args, names = "unique")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">341</td>
-                    <td class="coverage">183<em>x</em></td>
+                    <td class="num">193</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_formula(formula)</pre>
+                      <pre class="language-r">  args &lt;- attr(fun, "args")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">342</td>
-                    <td class="coverage">183<em>x</em></td>
+                    <td class="num">194</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if ("." %in% all.vars(formula)) {</pre>
+                      <pre class="language-r">  if (is.null(args)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">343</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">195</td>
+                    <td class="coverage">932<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("`.` is not allowed in mmrm models!")</pre>
+                      <pre class="language-r">    args &lt;- list()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">344</td>
+                    <td class="num">196</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">345</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">197</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  do.call(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">346</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">198</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    structure,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">347</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">199</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Standard Starting Value</pre>
+                      <pre class="language-r">    args = utils::modifyList(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">348</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">200</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">349</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">201</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtain standard start values.</pre>
+                      <pre class="language-r">        .Data = fun,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">350</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">202</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        args = utils::modifyList(args, a_args),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">351</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">203</td>
+                    <td class="coverage">1848<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param cov_type (`string`)\cr name of the covariance structure.</pre>
+                      <pre class="language-r">        class = c("partial", "function")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">352</td>
+                    <td class="num">204</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_visits (`int`)\cr number of visits.</pre>
+                      <pre class="language-r">      ),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">205</td>
+                    <td class="coverage">1848<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      additional_attr</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">353</td>
+                    <td class="num">206</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_groups (`int`)\cr number of groups.</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">354</td>
+                    <td class="num">207</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... not used.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">355</td>
+                    <td class="num">208</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">356</td>
+                    <td class="num">209</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">357</td>
+                    <td class="num">210</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `std_start` will try to provide variance parameter from identity matrix.</pre>
+                      <pre class="language-r">#' Obtain Default Covariance Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">358</td>
+                    <td class="num">211</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' However, for `ar1` and `ar1h` the corresponding values are not ideal because the</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">359</td>
+                    <td class="num">212</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \eqn{\rho} is usually a positive number thus using 0 as starting value can lead to</pre>
+                      <pre class="language-r">#' @description Obtain the default covariance method depending on</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">360</td>
+                    <td class="num">213</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' incorrect optimization result, and we use 0.5 as the initial value of \eqn{\rho}.</pre>
+                      <pre class="language-r">#' the degrees of freedom method used.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">361</td>
+                    <td class="num">214</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">362</td>
+                    <td class="num">215</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A numeric vector of starting values.</pre>
+                      <pre class="language-r">#' @param method (`string`)\cr degrees of freedom method.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">363</td>
+                    <td class="num">216</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">364</td>
+                    <td class="num">217</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#' @details The default covariance method is different for different degrees of freedom method.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">365</td>
+                    <td class="num">218</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">std_start &lt;- function(cov_type, n_visits, n_groups, ...) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">366</td>
-                    <td class="coverage">502<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(cov_type)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">367</td>
-                    <td class="coverage">502<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(cov_type, cov_types(c("abbr", "habbr")))</pre>
+                      <pre class="language-r">#' For "Satterthwaite" or "Between-Within", "Asymptotic" is returned.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">368</td>
-                    <td class="coverage">502<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">219</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_int(n_visits, lower = 1L)</pre>
+                      <pre class="language-r">#' For "Kenward-Roger" only, "Kenward-Roger" is returned.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">369</td>
-                    <td class="coverage">502<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">220</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_int(n_groups, lower = 1L)</pre>
+                      <pre class="language-r">#' For "Residual" only, "Empirical" is returned.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">370</td>
-                    <td class="coverage">502<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">221</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  start_value &lt;- switch(cov_type,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">371</td>
-                    <td class="coverage">502<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">222</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    us = rep(0, n_visits * (n_visits + 1) / 2),</pre>
+                      <pre class="language-r">#' @return String of the default covariance method.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">372</td>
-                    <td class="coverage">502<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">223</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    toep = rep(0, n_visits),</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">373</td>
-                    <td class="coverage">502<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">224</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    toeph = rep(0, 2 * n_visits - 1),</pre>
+                      <pre class="language-r">h_get_cov_default &lt;- function(method = c("Satterthwaite", "Kenward-Roger", "Residual", "Between-Within")) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">374</td>
-                    <td class="coverage">502<em>x</em></td>
+                    <td class="num">225</td>
+                    <td class="coverage">197<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ar1 = c(0, 0.5),</pre>
+                      <pre class="language-r">  assert_string(method)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">375</td>
-                    <td class="coverage">502<em>x</em></td>
+                    <td class="num">226</td>
+                    <td class="coverage">197<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ar1h = c(rep(0, n_visits), 0.5),</pre>
+                      <pre class="language-r">  method &lt;- match.arg(method)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">376</td>
-                    <td class="coverage">502<em>x</em></td>
+                    <td class="num">227</td>
+                    <td class="coverage">196<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ad = rep(0, n_visits),</pre>
+                      <pre class="language-r">  switch(method,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">377</td>
-                    <td class="coverage">502<em>x</em></td>
+                    <td class="num">228</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    adh = rep(0, 2 * n_visits - 1),</pre>
+                      <pre class="language-r">    "Residual" = "Empirical",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">378</td>
-                    <td class="coverage">502<em>x</em></td>
+                    <td class="num">229</td>
+                    <td class="coverage">158<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cs = rep(0, 2),</pre>
+                      <pre class="language-r">    "Satterthwaite" = "Asymptotic",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">379</td>
-                    <td class="coverage">502<em>x</em></td>
+                    <td class="num">230</td>
+                    <td class="coverage">35<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    csh = rep(0, n_visits + 1),</pre>
+                      <pre class="language-r">    "Kenward-Roger" = "Kenward-Roger",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">380</td>
-                    <td class="coverage">502<em>x</em></td>
+                    <td class="num">231</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    sp_exp = rep(0, 2)</pre>
+                      <pre class="language-r">    "Between-Within" = "Asymptotic"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">381</td>
+                    <td class="num">232</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">382</td>
-                    <td class="coverage">502<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  rep(start_value, n_groups)</pre>
-                    </td>
-                  </tr>
                   <tr class="never">
-                    <td class="num">383</td>
+                    <td class="num">233</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">384</td>
+                    <td class="num">234</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">385</td>
+                    <td class="num">235</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Empirical Starting Value</pre>
+                      <pre class="language-r">#' Complete `character` Vector Names From Values</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">386</td>
+                    <td class="num">236</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">387</td>
+                    <td class="num">237</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtain empirical start value for unstructured covariance</pre>
+                      <pre class="language-r">#' @param x (`character` or `list`)\cr value whose names should be completed</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">388</td>
+                    <td class="num">238</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   from element values.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">389</td>
+                    <td class="num">239</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param data (`data.frame`)\cr data used for model fitting.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">390</td>
+                    <td class="num">240</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param model_formula (`formula`)\cr the formula in mmrm model without covariance structure part.</pre>
+                      <pre class="language-r">#' @return A named vector or list.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">391</td>
+                    <td class="num">241</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param visit_var (`string`)\cr visit variable.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">392</td>
+                    <td class="num">242</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param subject_var (`string`)\cr subject id variable.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">393</td>
+                    <td class="num">243</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param subject_groups (`factor`)\cr subject group assignment.</pre>
+                      <pre class="language-r">fill_names &lt;- function(x) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">394</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">244</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... not used.</pre>
+                      <pre class="language-r">  n &lt;- names(x)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">395</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">245</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  is_unnamed &lt;- if (is.null(n)) rep_len(TRUE, length(x)) else n == ""</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">396</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">246</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">  names(x)[is_unnamed] &lt;- x[is_unnamed]</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">247</td>
+                    <td class="coverage">4<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  x</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">397</td>
+                    <td class="num">248</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This `emp_start` only works for unstructured covariance structure.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">398</td>
+                    <td class="num">249</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' It uses linear regression to first obtain the coefficients and use the residuals</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">399</td>
+                    <td class="num">250</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' to obtain the empirical variance-covariance, and it is then used to obtain the</pre>
+                      <pre class="language-r">#' Drop Items from an Indexible</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">400</td>
+                    <td class="num">251</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' starting values.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">401</td>
+                    <td class="num">252</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' Drop elements from an indexible object (`vector`, `list`, etc.).</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">253</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">402</td>
+                    <td class="num">254</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @note `data` is used instead of `full_frame` because `full_frame` is already</pre>
+                      <pre class="language-r">#' @param x Any object that can be consumed by [seq_along()] and indexed by a</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">403</td>
+                    <td class="num">255</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' transformed if model contains transformations, e.g. `log(FEV1) ~ exp(FEV1_BL)` will</pre>
+                      <pre class="language-r">#'   logical vector of the same length.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">404</td>
+                    <td class="num">256</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' drop `FEV1` and `FEV1_BL` but add `log(FEV1)` and `exp(FEV1_BL)` in `full_frame`.</pre>
+                      <pre class="language-r">#' @param n (`integer`)\cr the number of terms to drop.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">405</td>
+                    <td class="num">257</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">406</td>
+                    <td class="num">258</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A numeric vector of starting values.</pre>
+                      <pre class="language-r">#' @return A subset of `x`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">407</td>
+                    <td class="num">259</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">408</td>
+                    <td class="num">260</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">409</td>
+                    <td class="num">261</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">emp_start &lt;- function(data, model_formula, visit_var, subject_var, subject_groups, ...) {</pre>
+                      <pre class="language-r">drop_elements &lt;- function(x, n) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">410</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">262</td>
+                    <td class="coverage">819<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_formula(model_formula)</pre>
+                      <pre class="language-r">  x[seq_along(x) &gt; n]</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">411</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">263</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_data_frame(data)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">412</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">264</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(all.vars(model_formula), colnames(data))</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">413</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">265</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(visit_var)</pre>
+                      <pre class="language-r">#' Ask for Confirmation on Large Visit Levels</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">414</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">266</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(subject_var)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">415</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">267</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_factor(data[[visit_var]])</pre>
+                      <pre class="language-r">#' @description Ask the user for confirmation if there are too many visit levels</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">416</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">268</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_visits &lt;- length(levels(data[[visit_var]]))</pre>
+                      <pre class="language-r">#' for non-spatial covariance structure in interactive sessions.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">417</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">269</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_factor(data[[subject_var]])</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">418</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">270</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  subjects &lt;- droplevels(data[[subject_var]])</pre>
+                      <pre class="language-r">#' @param x (`numeric`)\cr number of visit levels.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">419</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">271</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_subjects &lt;- length(levels(subjects))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">420</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">272</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fit &lt;- stats::lm(formula = model_formula, data = data)</pre>
+                      <pre class="language-r">#' @return Logical value `TRUE`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">421</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">273</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  res &lt;- rep(NA, n_subjects * n_visits)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">422</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">274</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  res[</pre>
+                      <pre class="language-r">h_confirm_large_levels &lt;- function(x) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">423</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">275</td>
+                    <td class="coverage">297<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_visits * as.integer(subjects) - n_visits + as.integer(data[[visit_var]])</pre>
+                      <pre class="language-r">  assert_count(x)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">424</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">276</td>
+                    <td class="coverage">297<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ] &lt;- residuals(fit)</pre>
+                      <pre class="language-r">  allowed_lvls &lt;- x &lt;= getOption("mmrm.max_visits", 100)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">425</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">277</td>
+                    <td class="coverage">297<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  res_mat &lt;- matrix(res, ncol = n_visits, nrow = n_subjects, byrow = TRUE)</pre>
+                      <pre class="language-r">  if (allowed_lvls) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">426</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">278</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  emp_covs &lt;- lapply(</pre>
+                      <pre class="language-r">    return(TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">427</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">279</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    unname(split(seq_len(n_subjects), subject_groups)),</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">428</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">280</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    function(x) {</pre>
+                      <pre class="language-r">  if (!interactive()) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">429</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">281</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      stats::cov(res_mat[x, , drop = FALSE], use = "pairwise.complete.obs")</pre>
+                      <pre class="language-r">    stop("Visit levels too large!", call. = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">430</td>
+                    <td class="num">282</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">431</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">283</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  proceed &lt;- utils::askYesNo(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">432</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">284</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  unlist(lapply(emp_covs, h_get_theta_from_cov))</pre>
+                      <pre class="language-r">    paste(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">433</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">285</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      "Visit levels is possibly too large.",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">434</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">286</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain Theta from Covariance Matrix</pre>
+                      <pre class="language-r">      "This requires large memory. Are you sure to continue?",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">435</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">287</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      collapse = " "</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">436</td>
+                    <td class="num">288</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtain unstructured theta from covariance matrix.</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">437</td>
+                    <td class="num">289</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">438</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">290</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param covariance (`matrix`) of covariance matrix values.</pre>
+                      <pre class="language-r">  if (!identical(proceed, TRUE)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">439</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">291</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    stop("Visit levels too large!", call. = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">440</td>
+                    <td class="num">292</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">441</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">293</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' If the covariance matrix has `NA` in some of the elements, they will be replaced by</pre>
+                      <pre class="language-r">  return(TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">442</td>
+                    <td class="num">294</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 0 (non-diagonal) and 1 (diagonal). This ensures that the matrix is positive definite.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">443</td>
+                    <td class="num">295</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">444</td>
+                    <td class="num">296</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Numeric vector of the theta values.</pre>
+                      <pre class="language-r">#' Default Value on NULL</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">445</td>
+                    <td class="num">297</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' Return default value when first argument is NULL.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">446</td>
+                    <td class="num">298</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_theta_from_cov &lt;- function(covariance) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">447</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(covariance, mode = "numeric", ncols = nrow(covariance))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">448</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">299</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  covariance[is.na(covariance)] &lt;- 0</pre>
+                      <pre class="language-r">#' @param x Object.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">449</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">300</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  diag(covariance)[diag(covariance) == 0] &lt;- 1</pre>
+                      <pre class="language-r">#' @param y Object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">450</td>
+                    <td class="num">301</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # empirical is not always positive definite in some special cases of numeric singularity.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">451</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">302</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  qr_res &lt;- qr(covariance)</pre>
+                      <pre class="language-r">#' @details If `x` is NULL, returns `y`. Otherwise return `x`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">452</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">303</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (qr_res$rank &lt; ncol(covariance)) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">453</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">304</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    covariance &lt;- Matrix::nearPD(covariance)$mat</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">454</td>
+                    <td class="num">305</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">h_default_value &lt;- function(x, y) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">455</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">306</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  emp_chol &lt;- t(chol(covariance))</pre>
+                      <pre class="language-r">  if (is.null(x)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">456</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">307</td>
+                    <td class="coverage">277<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  mat &lt;- t(solve(diag(diag(emp_chol)), emp_chol))</pre>
+                      <pre class="language-r">    y</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">457</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">308</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret &lt;- c(log(diag(emp_chol)), mat[upper.tri(mat)])</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">458</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">309</td>
+                    <td class="coverage">35<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  unname(ret)</pre>
+                      <pre class="language-r">    x</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">459</td>
+                    <td class="num">310</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">460</td>
+                    <td class="num">311</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">461</td>
+                    <td class="num">312</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Register S3 Method</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">462</td>
+                    <td class="num">313</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Register S3 method to a generic.</pre>
+                      <pre class="language-r">#' Warn on na.action</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">463</td>
+                    <td class="num">314</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">464</td>
+                    <td class="num">315</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param pkg (`string`) name of the package name.</pre>
+                      <pre class="language-r">h_warn_na_action &lt;- function() {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">465</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">316</td>
+                    <td class="coverage">260<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param generic (`string`) name of the generic.</pre>
+                      <pre class="language-r">  if (!identical(getOption("na.action"), "na.omit")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">466</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">317</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param class (`string`) class name the function want to dispatch.</pre>
+                      <pre class="language-r">    warning("na.action is always set to `na.omit` for `mmrm` fit!")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">467</td>
+                    <td class="num">318</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param envir (`environment`) the location the method is defined.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">468</td>
+                    <td class="num">319</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">469</td>
+                    <td class="num">320</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details This function is adapted from `emmeans:::register_s3_method()`.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">470</td>
+                    <td class="num">321</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Obtain `na.action` as Function</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">471</td>
+                    <td class="num">322</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">472</td>
+                    <td class="num">323</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_register_s3 &lt;- function(pkg, generic, class, envir = parent.frame()) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">473</td>
-                    <td class="coverage">1<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(pkg)</pre>
+                      <pre class="language-r">h_get_na_action &lt;- function(na_action) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">474</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">324</td>
+                    <td class="coverage">56<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(generic)</pre>
+                      <pre class="language-r">  if (is.function(na_action) &amp;&amp; identical(methods::formalArgs(na_action), c("object", "..."))) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">475</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">325</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(class)</pre>
+                      <pre class="language-r">    return(na_action)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">476</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">326</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_environment(envir)</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">477</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">327</td>
+                    <td class="coverage">51<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fun &lt;- get(paste0(generic, ".", class), envir = envir)</pre>
+                      <pre class="language-r">  if (is.character(na_action) &amp;&amp; length(na_action) == 1L) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">478</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">328</td>
+                    <td class="coverage">51<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (isNamespaceLoaded(pkg)) {</pre>
+                      <pre class="language-r">    assert_subset(na_action, c("na.omit", "na.exclude", "na.fail", "na.pass", "na.contiguous"))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">479</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">329</td>
+                    <td class="coverage">51<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    registerS3method(generic, class, fun, envir = asNamespace(pkg))</pre>
+                      <pre class="language-r">    return(get(na_action, mode = "function", pos = "package:stats"))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">480</td>
+                    <td class="num">330</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">481</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">331</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  setHook(packageEvent(pkg, "onLoad"), function(...) {</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">482</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">332</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    registerS3method(generic, class, fun, envir = asNamespace(pkg))</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">483</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">484</td>
+                    <td class="num">333</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' Validate mmrm Formula</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">485</td>
+                    <td class="num">334</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @param formula (`formula`)\cr to check.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">486</td>
+                    <td class="num">335</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Check if a Factor Should Drop Levels</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">487</td>
+                    <td class="num">336</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @details In mmrm models, `.` is not allowed as it introduces ambiguity of covariates</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">488</td>
+                    <td class="num">337</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`vector`) vector to check.</pre>
+                      <pre class="language-r">#' to be used, so it is not allowed to be in formula.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">489</td>
+                    <td class="num">338</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">490</td>
+                    <td class="num">339</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">491</td>
+                    <td class="num">340</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_extra_levels &lt;- function(x) {</pre>
+                      <pre class="language-r">h_valid_formula &lt;- function(formula) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">492</td>
-                    <td class="coverage">1629<em>x</em></td>
+                    <td class="num">341</td>
+                    <td class="coverage">183<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  is.factor(x) &amp;&amp; length(levels(x)) &gt; length(unique(x))</pre>
+                      <pre class="language-r">  assert_formula(formula)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">493</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">342</td>
+                    <td class="coverage">183<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  if ("." %in% all.vars(formula)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">494</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">343</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    stop("`.` is not allowed in mmrm models!")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">495</td>
+                    <td class="num">344</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Drop Levels from Dataset</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">496</td>
+                    <td class="num">345</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param data (`data.frame`) data to drop levels.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">497</td>
+                    <td class="num">346</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param subject_var (`character`) subject variable.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">498</td>
+                    <td class="num">347</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param visit_var (`character`) visit variable.</pre>
+                      <pre class="language-r">#' Standard Starting Value</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">499</td>
+                    <td class="num">348</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param except (`character`) variables to exclude from dropping.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">500</td>
+                    <td class="num">349</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @description Obtain standard start values.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">501</td>
+                    <td class="num">350</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_drop_levels &lt;- function(data, subject_var, visit_var, except) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">502</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_data_frame(data)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">503</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_character(subject_var)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">504</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_character(visit_var)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">505</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_character(except, null.ok = TRUE)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">506</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  all_cols &lt;- colnames(data)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">507</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  to_drop &lt;- vapply(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">508</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    data,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">509</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    h_extra_levels,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">510</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    logical(1L)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">511</td>
+                    <td class="num">351</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">512</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  to_drop &lt;- all_cols[to_drop]</pre>
+                      <pre class="language-r">#' @param cov_type (`string`)\cr name of the covariance structure.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">513</td>
+                    <td class="num">352</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # only drop levels for those not defined in excep and not in visit_var.</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">514</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  to_drop &lt;- setdiff(to_drop, c(visit_var, except))</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">515</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  data[to_drop] &lt;- lapply(data[to_drop], droplevels)</pre>
+                      <pre class="language-r">#' @param n_visits (`int`)\cr number of visits.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">516</td>
+                    <td class="num">353</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # subject var are always dropped and no message given.</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">517</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  dropped &lt;- setdiff(to_drop, subject_var)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">518</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  if (length(dropped) &gt; 0) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">519</td>
-                    <td class="coverage">3<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    message(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">520</td>
-                    <td class="coverage">3<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      "Some factor levels are dropped due to singular design matrix: ",</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">521</td>
-                    <td class="coverage">3<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      toString(dropped)</pre>
+                      <pre class="language-r">#' @param n_groups (`int`)\cr number of groups.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">522</td>
+                    <td class="num">354</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' @param ... not used.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">523</td>
+                    <td class="num">355</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">524</td>
-                    <td class="coverage">263<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  data</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">525</td>
+                    <td class="num">356</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">526</td>
+                    <td class="num">357</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' `std_start` will try to provide variance parameter from identity matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">527</td>
+                    <td class="num">358</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Warn if TMB is Configured to Use Non-Deterministic Hash for Tape Optimizer</pre>
+                      <pre class="language-r">#' However, for `ar1` and `ar1h` the corresponding values are not ideal because the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">528</td>
+                    <td class="num">359</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' \eqn{\rho} is usually a positive number thus using 0 as starting value can lead to</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">529</td>
+                    <td class="num">360</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This function checks the TMB configuration for the `tmbad_deterministic_hash` setting</pre>
+                      <pre class="language-r">#' incorrect optimization result, and we use 0.5 as the initial value of \eqn{\rho}.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">530</td>
+                    <td class="num">361</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' If it is set to `FALSE`, a warning is issued indicating that this may lead to</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">531</td>
+                    <td class="num">362</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' unreproducible results.</pre>
+                      <pre class="language-r">#' @return A numeric vector of starting values.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">532</td>
+                    <td class="num">363</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">533</td>
+                    <td class="num">364</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return No return value, called for side effects.</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">534</td>
+                    <td class="num">365</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">std_start &lt;- function(cov_type, n_visits, n_groups, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">535</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">366</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_tmb_warn_non_deterministic &lt;- function() {</pre>
+                      <pre class="language-r">  assert_string(cov_type)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">536</td>
-                    <td class="coverage">169<em>x</em></td>
+                    <td class="num">367</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (utils::packageVersion("TMB") &lt; "1.9.15") {</pre>
+                      <pre class="language-r">  assert_subset(cov_type, cov_types(c("abbr", "habbr")))</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">537</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">368</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return()</pre>
+                      <pre class="language-r">  assert_int(n_visits, lower = 1L)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">538</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">369</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  assert_int(n_groups, lower = 1L)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">539</td>
-                    <td class="coverage">169<em>x</em></td>
+                    <td class="num">370</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tmb_config &lt;- TMB::config(DLL = "mmrm")</pre>
+                      <pre class="language-r">  start_value &lt;- switch(cov_type,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">540</td>
-                    <td class="coverage">169<em>x</em></td>
+                    <td class="num">371</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tape_deterministic &lt;- tmb_config$tmbad_deterministic_hash</pre>
+                      <pre class="language-r">    us = rep(0, n_visits * (n_visits + 1) / 2),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">541</td>
-                    <td class="coverage">169<em>x</em></td>
+                    <td class="num">372</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!tape_deterministic) {</pre>
+                      <pre class="language-r">    toep = rep(0, n_visits),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">542</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">373</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    msg &lt;- paste(</pre>
+                      <pre class="language-r">    toeph = rep(0, 2 * n_visits - 1),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">543</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">374</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "TMB is configured to use a non-deterministic hash for its tape optimizer,",</pre>
+                      <pre class="language-r">    ar1 = c(0, 0.5),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">544</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">375</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "and this may lead to unreproducible results.",</pre>
+                      <pre class="language-r">    ar1h = c(rep(0, n_visits), 0.5),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">545</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">376</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "To disable this behavior, use `TMB::config(tmbad_deterministic_hash = 1)`.",</pre>
+                      <pre class="language-r">    ad = rep(0, n_visits),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">546</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">377</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      sep = "\n"</pre>
+                      <pre class="language-r">    adh = rep(0, 2 * n_visits - 1),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">547</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">378</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">    cs = rep(0, 2),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">548</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">379</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning(msg)</pre>
+                      <pre class="language-r">    csh = rep(0, n_visits + 1),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">549</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">380</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    sp_exp = rep(0, 2)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">550</td>
+                    <td class="num">381</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/utils-formula.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
-                  <tr class="never">
-                    <td class="num">1</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">382</td>
+                    <td class="coverage">502<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Extract Formula Terms used for Covariance Structure Definition</pre>
+                      <pre class="language-r">  rep(start_value, n_groups)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">383</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">384</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param f (`formula`)\cr a formula from which covariance terms should be</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">385</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   extracted.</pre>
+                      <pre class="language-r">#' Empirical Starting Value</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">386</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">387</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A list of covariance structure expressions found in `f`.</pre>
+                      <pre class="language-r">#' @description Obtain empirical start value for unstructured covariance</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">388</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">389</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats terms</pre>
+                      <pre class="language-r">#' @param data (`data.frame`)\cr data used for model fitting.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">390</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @param model_formula (`formula`)\cr the formula in mmrm model without covariance structure part.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">391</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_extract_covariance_terms &lt;- function(f) {</pre>
+                      <pre class="language-r">#' @param visit_var (`string`)\cr visit variable.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">11</td>
-                    <td class="coverage">291<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">392</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  specials &lt;- cov_types(c("abbr", "habbr"))</pre>
+                      <pre class="language-r">#' @param subject_var (`string`)\cr subject id variable.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">12</td>
-                    <td class="coverage">291<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">393</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  terms &lt;- stats::terms(formula_rhs(f), specials = specials)</pre>
+                      <pre class="language-r">#' @param subject_groups (`factor`)\cr subject group assignment.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">13</td>
-                    <td class="coverage">291<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">394</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  covariance_terms &lt;- Filter(length, attr(terms, "specials"))</pre>
+                      <pre class="language-r">#' @param ... not used.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">14</td>
-                    <td class="coverage">291<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">395</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  variables &lt;- attr(terms, "variables")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">15</td>
-                    <td class="coverage">291<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">396</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  lapply(covariance_terms, function(i) variables[[i + 1]])</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">397</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' This `emp_start` only works for unstructured covariance structure.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">17</td>
+                    <td class="num">398</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' It uses linear regression to first obtain the coefficients and use the residuals</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">18</td>
+                    <td class="num">399</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Drop Formula Terms used for Covariance Structure Definition</pre>
+                      <pre class="language-r">#' to obtain the empirical variance-covariance, and it is then used to obtain the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">19</td>
+                    <td class="num">400</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' starting values.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">20</td>
+                    <td class="num">401</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param f (`formula`)\cr a formula from which covariance terms should be</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">21</td>
+                    <td class="num">402</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   dropped.</pre>
+                      <pre class="language-r">#' @note `data` is used instead of `full_frame` because `full_frame` is already</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">22</td>
+                    <td class="num">403</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' transformed if model contains transformations, e.g. `log(FEV1) ~ exp(FEV1_BL)` will</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">23</td>
+                    <td class="num">404</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The formula without accepted covariance terms.</pre>
+                      <pre class="language-r">#' drop `FEV1` and `FEV1_BL` but add `log(FEV1)` and `exp(FEV1_BL)` in `full_frame`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">24</td>
+                    <td class="num">405</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">25</td>
+                    <td class="num">406</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details `terms` is used and it will preserve the environment attribute.</pre>
+                      <pre class="language-r">#' @return A numeric vector of starting values.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">26</td>
+                    <td class="num">407</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This ensures the returned formula and the input formula have the same environment.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">27</td>
+                    <td class="num">408</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats terms drop.terms</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">28</td>
+                    <td class="num">409</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">emp_start &lt;- function(data, model_formula, visit_var, subject_var, subject_groups, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">29</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">h_drop_covariance_terms &lt;- function(f) {</pre>
+                  <tr class="covered">
+                    <td class="num">410</td>
+                    <td class="coverage">4<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_formula(model_formula)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">30</td>
-                    <td class="coverage">274<em>x</em></td>
+                    <td class="num">411</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  specials &lt;- cov_types(c("abbr", "habbr"))</pre>
+                      <pre class="language-r">  assert_data_frame(data)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">31</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">412</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  assert_subset(all.vars(model_formula), colnames(data))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">32</td>
-                    <td class="coverage">274<em>x</em></td>
+                    <td class="num">413</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  terms &lt;- stats::terms(f, specials = specials)</pre>
+                      <pre class="language-r">  assert_string(visit_var)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">33</td>
-                    <td class="coverage">274<em>x</em></td>
+                    <td class="num">414</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  covariance_terms &lt;- Filter(Negate(is.null), attr(terms, "specials"))</pre>
+                      <pre class="language-r">  assert_string(subject_var)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">34</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">415</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  assert_factor(data[[visit_var]])</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">35</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">416</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # if no covariance terms were found, return original formula</pre>
+                      <pre class="language-r">  n_visits &lt;- length(levels(data[[visit_var]]))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">36</td>
-                    <td class="coverage">274<em>x</em></td>
+                    <td class="num">417</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (length(covariance_terms) == 0) {</pre>
+                      <pre class="language-r">  assert_factor(data[[subject_var]])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">37</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">418</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(f)</pre>
+                      <pre class="language-r">  subjects &lt;- droplevels(data[[subject_var]])</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">38</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">419</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  n_subjects &lt;- length(levels(subjects))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">39</td>
-                    <td class="coverage">268<em>x</em></td>
+                    <td class="num">420</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (length(f) != 3) {</pre>
+                      <pre class="language-r">  fit &lt;- stats::lm(formula = model_formula, data = data)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">40</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">421</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    update_str &lt;- "~ . -"</pre>
+                      <pre class="language-r">  res &lt;- rep(NA, n_subjects * n_visits)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">41</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">422</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">  res[</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">42</td>
-                    <td class="coverage">267<em>x</em></td>
+                    <td class="num">423</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    update_str &lt;- ". ~ . -"</pre>
+                      <pre class="language-r">    n_visits * as.integer(subjects) - n_visits + as.integer(data[[visit_var]])</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">43</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">424</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  ] &lt;- residuals(fit)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">44</td>
-                    <td class="coverage">268<em>x</em></td>
+                    <td class="num">425</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  stats::update(</pre>
+                      <pre class="language-r">  res_mat &lt;- matrix(res, ncol = n_visits, nrow = n_subjects, byrow = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">45</td>
-                    <td class="coverage">268<em>x</em></td>
+                    <td class="num">426</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    f,</pre>
+                      <pre class="language-r">  emp_covs &lt;- lapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">46</td>
-                    <td class="coverage">268<em>x</em></td>
+                    <td class="num">427</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stats::as.formula(paste(update_str, deparse(attr(terms, "variables")[[covariance_terms[[1]] + 1]])))</pre>
+                      <pre class="language-r">    unname(split(seq_len(n_subjects), subject_groups)),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">47</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">428</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    function(x) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">48</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">429</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      stats::cov(res_mat[x, , drop = FALSE], use = "pairwise.complete.obs")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">49</td>
+                    <td class="num">430</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">431</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Add Individual Covariance Variables As Terms to Formula</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">51</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">432</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  unlist(lapply(emp_covs, h_get_theta_from_cov))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">52</td>
+                    <td class="num">433</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param f (`formula`)\cr a formula to which covariance structure terms should</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">53</td>
+                    <td class="num">434</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   be added.</pre>
+                      <pre class="language-r">#' Obtain Theta from Covariance Matrix</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">435</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param covariance (`cov_struct`)\cr a covariance structure object from which</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">55</td>
+                    <td class="num">436</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   additional variables should be sourced.</pre>
+                      <pre class="language-r">#' @description Obtain unstructured theta from covariance matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">56</td>
+                    <td class="num">437</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">57</td>
+                    <td class="num">438</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A new formula with included covariance terms.</pre>
+                      <pre class="language-r">#' @param covariance (`matrix`) of covariance matrix values.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">58</td>
+                    <td class="num">439</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">59</td>
+                    <td class="num">440</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details [stats::update()] is used to append the covariance structure and the environment</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">60</td>
+                    <td class="num">441</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' attribute will not be changed. This ensures the returned formula and the input formula</pre>
+                      <pre class="language-r">#' If the covariance matrix has `NA` in some of the elements, they will be replaced by</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">61</td>
+                    <td class="num">442</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' have the same environment.</pre>
+                      <pre class="language-r">#' 0 (non-diagonal) and 1 (diagonal). This ensures that the matrix is positive definite.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">62</td>
+                    <td class="num">443</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">63</td>
+                    <td class="num">444</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @return Numeric vector of the theta values.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">445</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">64</td>
+                    <td class="num">446</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_add_covariance_terms &lt;- function(f, covariance) {</pre>
+                      <pre class="language-r">h_get_theta_from_cov &lt;- function(covariance) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">65</td>
-                    <td class="coverage">272<em>x</em></td>
+                    <td class="num">447</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cov_terms &lt;- with(covariance, c(subject, visits, group))</pre>
+                      <pre class="language-r">  assert_matrix(covariance, mode = "numeric", ncols = nrow(covariance))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">66</td>
-                    <td class="coverage">266<em>x</em></td>
+                    <td class="num">448</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cov_terms &lt;- paste(cov_terms, collapse = " + ")</pre>
+                      <pre class="language-r">  covariance[is.na(covariance)] &lt;- 0</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">67</td>
-                    <td class="coverage">266<em>x</em></td>
+                    <td class="num">449</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  stats::update(f, stats::as.formula(paste(". ~ . + ", cov_terms)))</pre>
+                      <pre class="language-r">  diag(covariance)[diag(covariance) == 0] &lt;- 1</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">68</td>
+                    <td class="num">450</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  # empirical is not always positive definite in some special cases of numeric singularity.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">69</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">451</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  qr_res &lt;- qr(covariance)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">452</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  if (qr_res$rank &lt; ncol(covariance)) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="missed">
+                    <td class="num">453</td>
+                    <td class="coverage">!</td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    covariance &lt;- Matrix::nearPD(covariance)$mat</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">70</td>
+                    <td class="num">454</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Add Formula Terms with Character</pre>
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">455</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  emp_chol &lt;- t(chol(covariance))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">456</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  mat &lt;- t(solve(diag(diag(emp_chol)), emp_chol))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">457</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  ret &lt;- c(log(diag(emp_chol)), mat[upper.tri(mat)])</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">458</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  unname(ret)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">71</td>
+                    <td class="num">459</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">72</td>
+                    <td class="num">460</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Add formula terms from the original formula with character representation.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">73</td>
+                    <td class="num">461</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Register S3 Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">74</td>
+                    <td class="num">462</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param f (`formula`)\cr a formula to be updated.</pre>
+                      <pre class="language-r">#' Register S3 method to a generic.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">75</td>
+                    <td class="num">463</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param adds (`character`)\cr representation of elements to be added.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">76</td>
+                    <td class="num">464</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param drop_response (`flag`)\cr whether response should be dropped.</pre>
+                      <pre class="language-r">#' @param pkg (`string`) name of the package name.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">77</td>
+                    <td class="num">465</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param generic (`string`) name of the generic.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">78</td>
+                    <td class="num">466</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details Elements in `adds` will be added from the formula, while the environment</pre>
+                      <pre class="language-r">#' @param class (`string`) class name the function want to dispatch.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">79</td>
+                    <td class="num">467</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' of the formula is unchanged. If `adds` is `NULL` or `character(0)`, the formula is</pre>
+                      <pre class="language-r">#' @param envir (`environment`) the location the method is defined.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">80</td>
+                    <td class="num">468</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' unchanged.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">81</td>
+                    <td class="num">469</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A new formula with elements in `drops` removed.</pre>
+                      <pre class="language-r">#' @details This function is adapted from `emmeans:::register_s3_method()`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">82</td>
+                    <td class="num">470</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">83</td>
+                    <td class="num">471</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">84</td>
+                    <td class="num">472</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_add_terms &lt;- function(f, adds, drop_response = FALSE) {</pre>
+                      <pre class="language-r">h_register_s3 &lt;- function(pkg, generic, class, envir = parent.frame()) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">85</td>
-                    <td class="coverage">599<em>x</em></td>
+                    <td class="num">473</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_character(adds, null.ok = TRUE)</pre>
+                      <pre class="language-r">  assert_string(pkg)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">86</td>
-                    <td class="coverage">599<em>x</em></td>
+                    <td class="num">474</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (length(adds) &gt; 0L) {</pre>
+                      <pre class="language-r">  assert_string(generic)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">87</td>
-                    <td class="coverage">321<em>x</em></td>
+                    <td class="num">475</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    add_terms &lt;- stats::as.formula(sprintf(". ~ . + %s", paste(adds, collapse = "+")))</pre>
+                      <pre class="language-r">  assert_string(class)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">88</td>
-                    <td class="coverage">321<em>x</em></td>
+                    <td class="num">476</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    f &lt;- stats::update(f, add_terms)</pre>
+                      <pre class="language-r">  assert_environment(envir)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">89</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">477</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  fun &lt;- get(paste0(generic, ".", class), envir = envir)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">599<em>x</em></td>
+                    <td class="num">478</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (drop_response &amp;&amp; length(f) == 3L) {</pre>
+                      <pre class="language-r">  if (isNamespaceLoaded(pkg)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">35<em>x</em></td>
+                    <td class="num">479</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    f[[2]] &lt;- NULL</pre>
+                      <pre class="language-r">    registerS3method(generic, class, fun, envir = asNamespace(pkg))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">92</td>
+                    <td class="num">480</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">599<em>x</em></td>
+                    <td class="num">481</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  f</pre>
+                      <pre class="language-r">  setHook(packageEvent(pkg, "onLoad"), function(...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">94</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">482</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    registerS3method(generic, class, fun, envir = asNamespace(pkg))</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/tmb-methods.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">483</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Methods for `mmrm_tmb` Objects</pre>
+                      <pre class="language-r">  })</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">484</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">485</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">486</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Check if a Factor Should Drop Levels</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">487</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM object.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">488</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`mmrm_tmb`)\cr same as `object`.</pre>
+                      <pre class="language-r">#' @param x (`vector`) vector to check.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">489</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula (`mmrm_tmb`)\cr same as `object`.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">490</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param complete (`flag`)\cr whether to include potential non-estimable</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">491</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   coefficients.</pre>
+                      <pre class="language-r">h_extra_levels &lt;- function(x) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">10</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">492</td>
+                    <td class="coverage">1629<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... mostly not used;</pre>
+                      <pre class="language-r">  is.factor(x) &amp;&amp; length(levels(x)) &gt; length(unique(x))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">493</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   Exception is `model.matrix()` passing `...` to the default method.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">494</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Depends on the method, see Functions.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">13</td>
+                    <td class="num">495</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Drop Levels from Dataset</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">14</td>
+                    <td class="num">496</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @name mmrm_tmb_methods</pre>
+                      <pre class="language-r">#' @param data (`data.frame`) data to drop levels.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">15</td>
+                    <td class="num">497</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param subject_var (`character`) subject variable.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">498</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @seealso [`mmrm_methods`], [`mmrm_tidiers`] for additional methods.</pre>
+                      <pre class="language-r">#' @param visit_var (`character`) visit variable.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">17</td>
+                    <td class="num">499</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param except (`character`) variables to exclude from dropping.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">18</td>
+                    <td class="num">500</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">19</td>
+                    <td class="num">501</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' formula &lt;- FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)</pre>
+                      <pre class="language-r">h_drop_levels &lt;- function(data, subject_var, visit_var, except) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">20</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">502</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' object &lt;- fit_mmrm(formula, fev_data, weights = rep(1, nrow(fev_data)))</pre>
+                      <pre class="language-r">  assert_data_frame(data)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">21</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">503</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">NULL</pre>
+                      <pre class="language-r">  assert_character(subject_var)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">22</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">504</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  assert_character(visit_var)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">23</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">505</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the estimated coefficients.</pre>
+                      <pre class="language-r">  assert_character(except, null.ok = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">24</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats coef</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">25</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">26</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">27</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' # Estimated coefficients:</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">28</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' coef(object)</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">29</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">506</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">coef.mmrm_tmb &lt;- function(object, complete = TRUE, ...) {</pre>
+                      <pre class="language-r">  all_cols &lt;- colnames(data)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">30</td>
-                    <td class="coverage">58<em>x</em></td>
+                    <td class="num">507</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(complete)</pre>
+                      <pre class="language-r">  to_drop &lt;- vapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">31</td>
-                    <td class="coverage">58<em>x</em></td>
+                    <td class="num">508</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  nm &lt;- if (complete) "beta_est_complete" else "beta_est"</pre>
+                      <pre class="language-r">    data,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">32</td>
-                    <td class="coverage">58<em>x</em></td>
+                    <td class="num">509</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  component(object, name = nm)</pre>
+                      <pre class="language-r">    h_extra_levels,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">33</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">510</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    logical(1L)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">34</td>
+                    <td class="num">511</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">35</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">512</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the fitted values.</pre>
+                      <pre class="language-r">  to_drop &lt;- all_cols[to_drop]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">36</td>
+                    <td class="num">513</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats fitted</pre>
+                      <pre class="language-r">  # only drop levels for those not defined in excep and not in visit_var.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">37</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">514</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">  to_drop &lt;- setdiff(to_drop, c(visit_var, except))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">38</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">515</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">  data[to_drop] &lt;- lapply(data[to_drop], droplevels)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">39</td>
+                    <td class="num">516</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Fitted values:</pre>
+                      <pre class="language-r">  # subject var are always dropped and no message given.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">40</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">517</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fitted(object)</pre>
+                      <pre class="language-r">  dropped &lt;- setdiff(to_drop, subject_var)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">41</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">518</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">fitted.mmrm_tmb &lt;- function(object, ...) {</pre>
+                      <pre class="language-r">  if (length(dropped) &gt; 0) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">42</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">519</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fitted_col &lt;- component(object, "x_matrix") %*% component(object, "beta_est")</pre>
+                      <pre class="language-r">    message(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">43</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">520</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fitted_col[, 1L, drop = TRUE]</pre>
+                      <pre class="language-r">      "Some factor levels are dropped due to singular design matrix: ",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">44</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">521</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      toString(dropped)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">45</td>
+                    <td class="num">522</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">46</td>
+                    <td class="num">523</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods predict conditional means for new data;</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">47</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">524</td>
+                    <td class="coverage">263<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'  optionally with standard errors and confidence or prediction intervals.</pre>
+                      <pre class="language-r">  data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">48</td>
+                    <td class="num">525</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'  Returns a vector of predictions if `se.fit == FALSE` and</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">49</td>
+                    <td class="num">526</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'  `interval == "none"`; otherwise it returns a data.frame with multiple</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">527</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'  columns and one row per input data row.</pre>
+                      <pre class="language-r">#' Predicate if the TMB Version Used to Compile the Package is Sufficient</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">51</td>
+                    <td class="num">528</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">52</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @param newdata (`data.frame`)\cr optional new data, otherwise data from `object` is used.</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">53</td>
+                    <td class="num">529</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param se.fit (`flag`)\cr indicator if standard errors are required.</pre>
+                      <pre class="language-r">#' @return Flag whether the TMB version is sufficient.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">530</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param interval (`string`)\cr type of interval calculation. Can be abbreviated.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">55</td>
+                    <td class="num">531</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param level (`number`)\cr tolerance/confidence level.</pre>
+                      <pre class="language-r">h_tmb_version_sufficient &lt;- function() {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">56</td>
+                    <td class="num">532</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param nsim (`count`)\cr number of simulations to use.</pre>
+                      <pre class="language-r">  # Note: There is no version information saved in the dynamic library, but</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">57</td>
+                    <td class="num">533</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param conditional (`flag`)\cr indicator if the prediction is conditional on the observation or not.</pre>
+                      <pre class="language-r">  # we can check like this:</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">58</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">534</td>
+                    <td class="coverage">170<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  tmb_config &lt;- TMB::config(DLL = "mmrm")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">59</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">535</td>
+                    <td class="coverage">170<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats predict</pre>
+                      <pre class="language-r">  tape_deterministic &lt;- tmb_config$tmbad_deterministic_hash</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">60</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">536</td>
+                    <td class="coverage">170<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">  !is.null(tape_deterministic)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">61</td>
+                    <td class="num">537</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">62</td>
+                    <td class="num">538</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">63</td>
+                    <td class="num">539</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' predict(object, newdata = fev_data)</pre>
+                      <pre class="language-r">#' Warn if TMB is Configured to Use Non-Deterministic Hash for Tape Optimizer</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">64</td>
+                    <td class="num">540</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">predict.mmrm_tmb &lt;- function(object,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">65</td>
+                    <td class="num">541</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             newdata,</pre>
+                      <pre class="language-r">#' This function checks the TMB configuration for the `tmbad_deterministic_hash` setting</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">66</td>
+                    <td class="num">542</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             se.fit = FALSE, # nolint</pre>
+                      <pre class="language-r">#' If it is set to `FALSE`, a warning is issued indicating that this may lead to</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">67</td>
+                    <td class="num">543</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             interval = c("none", "confidence", "prediction"),</pre>
+                      <pre class="language-r">#' unreproducible results.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">68</td>
+                    <td class="num">544</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             level = 0.95,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">69</td>
+                    <td class="num">545</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             nsim = 1000L,</pre>
+                      <pre class="language-r">#' @return No return value, called for side effects.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">70</td>
+                    <td class="num">546</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             conditional = FALSE,</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">71</td>
+                    <td class="num">547</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             ...) {</pre>
+                      <pre class="language-r">h_tmb_warn_non_deterministic &lt;- function() {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">72</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">548</td>
+                    <td class="coverage">169<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(newdata)) {</pre>
+                      <pre class="language-r">  if (!h_tmb_version_sufficient()) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">73</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">549</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    newdata &lt;- object$data</pre>
+                      <pre class="language-r">    return()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">74</td>
+                    <td class="num">550</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">75</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">551</td>
+                    <td class="coverage">169<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_data_frame(newdata)</pre>
+                      <pre class="language-r">  tmb_config &lt;- TMB::config(DLL = "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">76</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">552</td>
+                    <td class="coverage">169<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  orig_row_names &lt;- row.names(newdata)</pre>
+                      <pre class="language-r">  tape_deterministic &lt;- tmb_config$tmbad_deterministic_hash</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">77</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">553</td>
+                    <td class="coverage">169<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(se.fit)</pre>
+                      <pre class="language-r">  if (!tape_deterministic) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">78</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">554</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(level, lower = 0, upper = 1)</pre>
+                      <pre class="language-r">    msg &lt;- paste(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">79</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">555</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
+                      <pre class="language-r">      "TMB is configured to use a non-deterministic hash for its tape optimizer,",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">80</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">556</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(conditional)</pre>
+                      <pre class="language-r">      "and this may lead to unreproducible results.",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">81</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">557</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  interval &lt;- match.arg(interval)</pre>
+                      <pre class="language-r">      "To disable this behavior, use `TMB::config(tmbad_deterministic_hash = 1)`.",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">82</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">558</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  formula_parts &lt;- object$formula_parts</pre>
+                      <pre class="language-r">      sep = "\n"</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">83</td>
-                    <td class="coverage">45<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">559</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (any(object$tmb_data$x_cols_aliased)) {</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">84</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">560</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning(</pre>
+                      <pre class="language-r">    warning(msg)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">85</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">561</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "In fitted object there are co-linear variables and therefore dropped terms, ",</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">86</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">562</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "and this could lead to incorrect prediction on new data."</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/tmb.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">87</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' Processing the Formula for `TMB` Fit</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">88</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">89</td>
-                    <td class="coverage">45<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">3</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  colnames &lt;- names(Filter(isFALSE, object$tmb_data$x_cols_aliased))</pre>
+                      <pre class="language-r">#' @param formula (`formula`)\cr Original formula.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">45<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">4</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!conditional &amp;&amp; interval %in% c("none", "confidence")) {</pre>
+                      <pre class="language-r">#' @param covariance (`cov_struct`)\cr A covariance structure from which</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">91</td>
+                    <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # model.matrix always return a complete matrix (no NA allowed)</pre>
+                      <pre class="language-r">#'   additional formula parts should be added.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">6</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x_mat &lt;- stats::model.matrix(object, data = newdata, use_response = FALSE)[, colnames, drop = FALSE]</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">7</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x_mat_full &lt;- matrix(</pre>
+                      <pre class="language-r">#' @return List of class `mmrm_tmb_formula_parts` with elements:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">94</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">8</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      NA,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">95</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">9</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      nrow = nrow(newdata), ncol = ncol(x_mat),</pre>
+                      <pre class="language-r">#' - `formula`: the original input.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">96</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">10</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      dimnames = list(row.names(newdata), colnames(x_mat))</pre>
+                      <pre class="language-r">#' - `model_formula`: `formula` with the covariance term is removed.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">97</td>
+                    <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' - `model_formula`: `formula` with the covariance term removed.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">98</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">12</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x_mat_full[row.names(x_mat), ] &lt;- x_mat</pre>
+                      <pre class="language-r">#' - `full_formula`: same as `model_formula` but includes the covariance</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">99</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">13</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    predictions &lt;- (x_mat_full %*% component(object, "beta_est"))[, 1]</pre>
+                      <pre class="language-r">#'   structure's subject, visit and (optionally) group variables.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">100</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">14</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    predictions_raw &lt;- stats::setNames(rep(NA_real_, nrow(newdata)), row.names(newdata))</pre>
+                      <pre class="language-r">#' - `cov_type`: `string` with covariance term type (e.g. `"us"`).</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">101</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">15</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    predictions_raw[names(predictions)] &lt;- predictions</pre>
+                      <pre class="language-r">#' - `is_spatial`: `flag` indicator of whether the covariance structure is</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">102</td>
-                    <td class="coverage">27<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">16</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (identical(interval, "none")) {</pre>
+                      <pre class="language-r">#'   spatial</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">103</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">17</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      return(predictions_raw)</pre>
+                      <pre class="language-r">#' - `visit_var`: `character` with the visit variable name.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">104</td>
+                    <td class="num">18</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#' - `subject_var`: `string` with the subject variable name.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">105</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">19</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    se &lt;- switch(interval,</pre>
+                      <pre class="language-r">#' - `group_var`: `string` with the group variable name. If no group specified,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">106</td>
+                    <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # can be NA if there are aliased cols</pre>
+                      <pre class="language-r">#'   this element is `NULL`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">107</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">21</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "confidence" = diag(x_mat_full %*% component(object, "beta_vcov") %*% t(x_mat_full)),</pre>
+                      <pre class="language-r">#' - `model_var`: `character` with the variables names of the formula, except `subject_var`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">108</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">22</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "none" = NA_real_</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">109</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">110</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    res &lt;- cbind(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">111</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      fit = predictions, se = se,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">112</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      lwr = predictions - stats::qnorm(1 - level / 2) * se, upr = predictions + stats::qnorm(1 - level / 2) * se</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">113</td>
+                    <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">114</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    if (!se.fit) {</pre>
+                      <pre class="language-r">h_mmrm_tmb_formula_parts &lt;- function(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">115</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">25</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      res &lt;- res[, setdiff(colnames(res), "se")]</pre>
+                      <pre class="language-r">    formula,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">116</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">    covariance = as.cov_struct(formula, warn_partial = FALSE)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">117</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">27</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    res_raw &lt;- matrix(</pre>
+                      <pre class="language-r">  assert_formula(formula)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">118</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">28</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      NA_real_,</pre>
+                      <pre class="language-r">  assert_true(identical(length(formula), 3L))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">119</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">29</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ncol = ncol(res), nrow = nrow(newdata),</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">120</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">30</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      dimnames = list(row.names(newdata), colnames(res))</pre>
+                      <pre class="language-r">  model_formula &lt;- h_drop_covariance_terms(formula)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">121</td>
+                    <td class="num">31</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">122</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">32</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    res_raw[row.names(res), ] &lt;- res</pre>
+                      <pre class="language-r">  structure(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">123</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">33</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(res_raw)</pre>
+                      <pre class="language-r">    list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">124</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">34</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">      formula = formula,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">125</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">35</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tmb_data &lt;- h_mmrm_tmb_data(</pre>
+                      <pre class="language-r">      model_formula = model_formula,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">126</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">36</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    formula_parts, newdata,</pre>
+                      <pre class="language-r">      full_formula = h_add_covariance_terms(model_formula, covariance),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">127</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">37</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    weights = rep(1, nrow(newdata)),</pre>
+                      <pre class="language-r">      cov_type = tmb_cov_type(covariance),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">128</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">38</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    reml = TRUE,</pre>
+                      <pre class="language-r">      is_spatial = covariance$type == "sp_exp",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">129</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">39</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    singular = "keep",</pre>
+                      <pre class="language-r">      visit_var = covariance$visits,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">130</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">40</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    drop_visit_levels = FALSE,</pre>
+                      <pre class="language-r">      subject_var = covariance$subject,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">131</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">41</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    allow_na_response = TRUE,</pre>
+                      <pre class="language-r">      group_var = if (length(covariance$group) &lt; 1) NULL else covariance$group,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">132</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">42</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    drop_levels = FALSE,</pre>
+                      <pre class="language-r">      model_var = setdiff(all.vars(formula[[3]]), covariance$subject)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">133</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">43</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    xlev = component(object, "xlev"),</pre>
+                      <pre class="language-r">    ),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">134</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">44</td>
+                    <td class="coverage">270<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    contrasts = component(object, "contrasts")</pre>
+                      <pre class="language-r">    class = "mmrm_tmb_formula_parts"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">135</td>
+                    <td class="num">45</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">136</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">46</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tmb_data$x_matrix &lt;- tmb_data$x_matrix[, colnames, drop = FALSE]</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">137</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">47</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  predictions &lt;- h_get_prediction(</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">138</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">48</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tmb_data, object$theta_est, object$beta_est, component(object, "beta_vcov")</pre>
+                      <pre class="language-r">#' Data for `TMB` Fit</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">139</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">49</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )$prediction</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">140</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">50</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  res &lt;- cbind(fit = rep(NA_real_, nrow(newdata)))</pre>
+                      <pre class="language-r">#' @param formula_parts (`mmrm_tmb_formula_parts`)\cr list with formula parts</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">141</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">51</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  new_order &lt;- match(row.names(tmb_data$full_frame), orig_row_names)</pre>
+                      <pre class="language-r">#'   from [h_mmrm_tmb_formula_parts()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">142</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">52</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  res[new_order, "fit"] &lt;- predictions[, "fit"]</pre>
+                      <pre class="language-r">#' @param data (`data.frame`)\cr which contains variables used in `formula_parts`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">143</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">53</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  se &lt;- switch(interval,</pre>
+                      <pre class="language-r">#' @param weights (`vector`)\cr weights to be used in the fitting process.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">144</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">54</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "confidence" = sqrt(predictions[, "conf_var"]),</pre>
+                      <pre class="language-r">#' @param reml (`flag`)\cr whether restricted maximum likelihood (REML) estimation is used,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">145</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">55</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "prediction" = sqrt(h_get_prediction_variance(object, nsim, tmb_data)),</pre>
+                      <pre class="language-r">#'   otherwise maximum likelihood (ML) is used.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">146</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">56</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "none" = NULL</pre>
+                      <pre class="language-r">#' @param singular (`string`)\cr choices of method deal with rank-deficient matrices. "error" to</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">147</td>
+                    <td class="num">57</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'   stop the function return the error, "drop" to drop these columns, and "keep" to keep all the columns.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">148</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">58</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (interval != "none") {</pre>
+                      <pre class="language-r">#' @param drop_visit_levels (`flag`)\cr whether to drop levels for visit variable, if visit variable is a factor.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">149</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">59</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    res &lt;- cbind(</pre>
+                      <pre class="language-r">#' @param allow_na_response (`flag`)\cr whether NA in response is allowed.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">150</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">60</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      res,</pre>
+                      <pre class="language-r">#' @param drop_levels (`flag`)\cr whether drop levels for covariates. If not dropped could lead to singular matrix.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">151</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">61</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      se = NA_real_</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">152</td>
+                    <td class="num">62</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' @return List of class `mmrm_tmb_data` with elements:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">153</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">63</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    res[new_order, "se"] &lt;- se</pre>
+                      <pre class="language-r">#' - `full_frame`: `data.frame` with `n` rows containing all variables needed in the model.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">154</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">64</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    alpha &lt;- 1 - level</pre>
+                      <pre class="language-r">#' - `data`: `data.frame` of input dataset.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">155</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">65</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    z &lt;- stats::qnorm(1 - alpha / 2) * res[, "se"]</pre>
+                      <pre class="language-r">#' - `x_matrix`: `matrix` with `n` rows and `p` columns specifying the overall design matrix.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">156</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">66</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    res &lt;- cbind(</pre>
+                      <pre class="language-r">#' - `x_cols_aliased`: `logical` with potentially more than `p` elements indicating which</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">157</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">67</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      res,</pre>
+                      <pre class="language-r">#'      columns in the original design matrix have been left out to obtain a full rank</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">158</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">68</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      lwr = res[, "fit"] - z,</pre>
+                      <pre class="language-r">#'      `x_matrix`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">159</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">69</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      upr = res[, "fit"] + z</pre>
+                      <pre class="language-r">#' - `y_vector`: length `n` `numeric` specifying the overall response vector.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">160</td>
+                    <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' - `weights_vector`: length `n` `numeric` specifying the weights vector.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">161</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">71</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (!se.fit) {</pre>
+                      <pre class="language-r">#' - `n_visits`: `int` with the number of visits, which is the dimension of the</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">162</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">72</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      res &lt;- res[, setdiff(colnames(res), "se")]</pre>
+                      <pre class="language-r">#'      covariance matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">163</td>
+                    <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#' - `n_subjects`: `int` with the number of subjects.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">164</td>
+                    <td class="num">74</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' - `subject_zero_inds`: length `n_subjects` `integer` containing the zero-based start</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">165</td>
+                    <td class="num">75</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Use original names.</pre>
+                      <pre class="language-r">#'     indices for each subject.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">166</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">76</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  row.names(res) &lt;- orig_row_names</pre>
+                      <pre class="language-r">#' - `subject_n_visits`: length `n_subjects` `integer` containing the number of</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">167</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">77</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (ncol(res) == 1) {</pre>
+                      <pre class="language-r">#'     observed visits for each subjects. So the sum of this vector equals `n`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">168</td>
-                    <td class="coverage">11<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">78</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    res &lt;- res[, "fit"]</pre>
+                      <pre class="language-r">#' - `cov_type`: `string` value specifying the covariance type.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">169</td>
+                    <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' - `is_spatial_int`: `int` specifying whether the covariance structure is spatial(1) or not(0).</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">170</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">80</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  return(res)</pre>
+                      <pre class="language-r">#' - `reml`: `int` specifying whether REML estimation is used (1), otherwise ML (0).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">171</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' - `subject_groups`: `factor` specifying the grouping for each subject.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">172</td>
+                    <td class="num">82</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' - `n_groups`: `int` with the number of total groups</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">173</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Get Prediction</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">174</td>
+                    <td class="num">84</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @details Note that the `subject_var` must not be factor but can also be character.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">175</td>
+                    <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Get predictions with given `data`, `theta`, `beta`, `beta_vcov`.</pre>
+                      <pre class="language-r">#'   If it is character, then it will be converted to factor internally. Here</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">176</td>
+                    <td class="num">86</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   the levels will be the unique values, sorted alphabetically and numerically if there</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">177</td>
+                    <td class="num">87</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details See `predict` function in `predict.cpp` which is called internally.</pre>
+                      <pre class="language-r">#'   is a common string prefix of numbers in the character elements. For full control</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">178</td>
+                    <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   on the order please use a factor.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">179</td>
+                    <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr object.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">180</td>
+                    <td class="num">90</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param theta (`numeric`)\cr theta value.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">181</td>
+                    <td class="num">91</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param beta (`numeric`)\cr beta value.</pre>
+                      <pre class="language-r">h_mmrm_tmb_data &lt;- function(formula_parts,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">182</td>
+                    <td class="num">92</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param beta_vcov (`matrix`)\cr beta_vcov matrix.</pre>
+                      <pre class="language-r">                            data,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">183</td>
+                    <td class="num">93</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                            weights,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">184</td>
+                    <td class="num">94</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with:</pre>
+                      <pre class="language-r">                            reml,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">185</td>
+                    <td class="num">95</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `prediction`: Matrix with columns `fit`, `conf_var`, and `var`.</pre>
+                      <pre class="language-r">                            singular = c("drop", "error", "keep"),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">186</td>
+                    <td class="num">96</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `covariance`: List with subject specific covariance matrices.</pre>
+                      <pre class="language-r">                            drop_visit_levels,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">187</td>
+                    <td class="num">97</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `index`: List of zero-based subject indices.</pre>
+                      <pre class="language-r">                            allow_na_response = FALSE,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">188</td>
+                    <td class="num">98</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                            drop_levels = TRUE,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">189</td>
+                    <td class="num">99</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">                            xlev = NULL,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">190</td>
+                    <td class="num">100</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_prediction &lt;- function(tmb_data, theta, beta, beta_vcov) {</pre>
+                      <pre class="language-r">                            contrasts = NULL) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">191</td>
-                    <td class="coverage">1696<em>x</em></td>
+                    <td class="num">101</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
+                      <pre class="language-r">  assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">192</td>
-                    <td class="coverage">1696<em>x</em></td>
+                    <td class="num">102</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(theta)</pre>
+                      <pre class="language-r">  assert_data_frame(data)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">193</td>
-                    <td class="coverage">1696<em>x</em></td>
+                    <td class="num">103</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_beta &lt;- ncol(tmb_data$x_matrix)</pre>
+                      <pre class="language-r">  varname &lt;- formula_parts[grepl("_var", names(formula_parts))]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">194</td>
-                    <td class="coverage">1696<em>x</em></td>
+                    <td class="num">104</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(beta, finite = TRUE, any.missing = FALSE, len = n_beta)</pre>
+                      <pre class="language-r">  assert_names(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">195</td>
-                    <td class="coverage">1696<em>x</em></td>
+                    <td class="num">105</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(beta_vcov, mode = "numeric", any.missing = FALSE, nrows = n_beta, ncols = n_beta)</pre>
+                      <pre class="language-r">    names(data),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">196</td>
-                    <td class="coverage">1696<em>x</em></td>
+                    <td class="num">106</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  .Call(`_mmrm_predict`, PACKAGE = "mmrm", tmb_data, theta, beta, beta_vcov)</pre>
+                      <pre class="language-r">    must.include = unlist(varname, use.names = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">197</td>
+                    <td class="num">107</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">198</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">108</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  assert_true(is.factor(data[[formula_parts$subject_var]]) || is.character(data[[formula_parts$subject_var]]))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">199</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">109</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Get Prediction Variance</pre>
+                      <pre class="language-r">  assert_numeric(weights, len = nrow(data))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">200</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">110</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_flag(reml)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">201</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">111</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Get prediction variance with given fit, `tmb_data` with the Monte Carlo sampling method.</pre>
+                      <pre class="language-r">  singular &lt;- match.arg(singular)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">202</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">112</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_flag(drop_visit_levels)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">203</td>
+                    <td class="num">113</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">204</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">114</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param nsim (`count`)\cr number of samples.</pre>
+                      <pre class="language-r">  if (is.character(data[[formula_parts$subject_var]])) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">205</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">115</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr object.</pre>
+                      <pre class="language-r">    data[[formula_parts$subject_var]] &lt;- factor(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">206</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">116</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      data[[formula_parts$subject_var]],</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">117</td>
+                    <td class="coverage">5<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      levels = stringr::str_sort(unique(data[[formula_parts$subject_var]]), numeric = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">207</td>
+                    <td class="num">118</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">208</td>
+                    <td class="num">119</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_prediction_variance &lt;- function(object, nsim, tmb_data) {</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">209</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">120</td>
+                    <td class="coverage">312<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
+                      <pre class="language-r">  data_order &lt;- if (formula_parts$is_spatial) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">210</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">121</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
+                      <pre class="language-r">    order(data[[formula_parts$subject_var]])</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">211</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">122</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">212</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">123</td>
+                    <td class="coverage">296<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  theta_chol &lt;- chol(object$theta_vcov)</pre>
+                      <pre class="language-r">    subject_visit_data &lt;- data[, c(formula_parts$subject_var, formula_parts$visit_var)]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">213</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">124</td>
+                    <td class="coverage">296<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_theta &lt;- length(object$theta_est)</pre>
+                      <pre class="language-r">    is_duplicated &lt;- duplicated(subject_visit_data)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">214</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">125</td>
+                    <td class="coverage">296<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  res &lt;- replicate(nsim, {</pre>
+                      <pre class="language-r">    if (any(is_duplicated)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">215</td>
-                    <td class="coverage">1150<em>x</em></td>
+                    <td class="num">126</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    z &lt;- stats::rnorm(n = n_theta)</pre>
+                      <pre class="language-r">      stop(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">216</td>
-                    <td class="coverage">1150<em>x</em></td>
+                    <td class="num">127</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    theta_sample &lt;- object$theta_est + z %*% theta_chol</pre>
+                      <pre class="language-r">        "time points have to be unique for each subject, detected following duplicates in data:\n",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">217</td>
-                    <td class="coverage">1150<em>x</em></td>
+                    <td class="num">128</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cond_beta_results &lt;- object$tmb_object$report(theta_sample)</pre>
+                      <pre class="language-r">        paste(utils::capture.output(print(subject_visit_data[is_duplicated, ])), collapse = "\n")</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">218</td>
-                    <td class="coverage">1150<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">129</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    beta_mean &lt;- cond_beta_results$beta</pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">219</td>
-                    <td class="coverage">1150<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">130</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    beta_cov &lt;- cond_beta_results$beta_vcov</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">220</td>
-                    <td class="coverage">1150<em>x</em></td>
+                    <td class="num">131</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_get_prediction(tmb_data, theta_sample, beta_mean, beta_cov)$prediction</pre>
+                      <pre class="language-r">    order(data[[formula_parts$subject_var]], data[[formula_parts$visit_var]])</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">221</td>
+                    <td class="num">132</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">222</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  mean_of_var &lt;- rowMeans(res[, "var", ])</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">223</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">133</td>
+                    <td class="coverage">311<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  var_of_mean &lt;- apply(res[, "fit", ], 1, stats::var)</pre>
+                      <pre class="language-r">  if (identical(formula_parts$is_spatial, FALSE)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">224</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">134</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  mean_of_var + var_of_mean</pre>
+                      <pre class="language-r">    h_confirm_large_levels(length(levels(data[[formula_parts$visit_var]])))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">225</td>
+                    <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">226</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">136</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  data &lt;- data[data_order, ]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">227</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">137</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the model frame.</pre>
+                      <pre class="language-r">  weights &lt;- weights[data_order]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">228</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">138</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param data (`data.frame`)\cr object in which to construct the frame.</pre>
+                      <pre class="language-r">  data &lt;- data.frame(data, weights)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">229</td>
+                    <td class="num">139</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param include (`character`)\cr names of variable types to include.</pre>
+                      <pre class="language-r">  # Weights is always the last column.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">230</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">140</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   Must be `NULL` or one or more of `c("subject_var", "visit_var", "group_var", "response_var")`.</pre>
+                      <pre class="language-r">  weights_name &lt;- colnames(data)[ncol(data)]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">231</td>
+                    <td class="num">141</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param full (`flag`)\cr indicator whether to return full model frame (deprecated).</pre>
+                      <pre class="language-r">  # If `y` is allowed to be NA, then first replace y with 1:n, then replace it with original y.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">232</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">142</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param na.action (`string`)\cr na action.</pre>
+                      <pre class="language-r">  if (!allow_na_response) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">233</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">143</td>
+                    <td class="coverage">260<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats model.frame</pre>
+                      <pre class="language-r">    h_warn_na_action()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">234</td>
+                    <td class="num">144</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">235</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">145</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  full_frame &lt;- eval(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">236</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">146</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">    bquote(stats::model.frame(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">237</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">147</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `include` argument controls the variables the returned model frame will include.</pre>
+                      <pre class="language-r">      formula_parts$full_formula,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">238</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">148</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Possible options are "response_var", "subject_var", "visit_var" and "group_var", representing the</pre>
+                      <pre class="language-r">      data = data,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">239</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">149</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' response variable, subject variable, visit variable or group variable.</pre>
+                      <pre class="language-r">      weights = .(as.symbol(weights_name)),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">240</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">150</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `character` values in new data will always be factorized according to the data in the fit</pre>
+                      <pre class="language-r">      na.action = "na.pass",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">241</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">151</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' to avoid mismatched in levels or issues in `model.matrix`.</pre>
+                      <pre class="language-r">      xlev = xlev</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">242</td>
+                    <td class="num">152</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    ))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">243</td>
+                    <td class="num">153</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">244</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">154</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Model frame:</pre>
+                      <pre class="language-r">  if (drop_levels) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">245</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">155</td>
+                    <td class="coverage">262<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' model.frame(object)</pre>
+                      <pre class="language-r">    full_frame &lt;- h_drop_levels(full_frame, formula_parts$subject_var, formula_parts$visit_var, names(xlev))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">246</td>
+                    <td class="num">156</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' model.frame(object, include = "subject_var")</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">247</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">157</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">model.frame.mmrm_tmb &lt;- function(formula, data, include = c("subject_var", "visit_var", "group_var", "response_var"),</pre>
+                      <pre class="language-r">  has_response &lt;- !identical(attr(attr(full_frame, "terms"), "response"), 0L)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">248</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">158</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                 full, na.action = "na.omit", ...) { # nolint</pre>
+                      <pre class="language-r">  keep_ind &lt;- if (allow_na_response &amp;&amp; has_response) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">249</td>
+                    <td class="num">159</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Construct updated formula and data arguments.</pre>
+                      <pre class="language-r">    # Note that response is always the first column if there is response.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">250</td>
-                    <td class="coverage">46<em>x</em></td>
+                    <td class="num">160</td>
+                    <td class="coverage">50<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  lst_formula_and_data &lt;-</pre>
+                      <pre class="language-r">    stats::complete.cases(full_frame[, -1L, drop = FALSE])</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">251</td>
-                    <td class="coverage">46<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">161</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_construct_model_frame_inputs(</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">252</td>
-                    <td class="coverage">46<em>x</em></td>
+                    <td class="num">162</td>
+                    <td class="coverage">260<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      formula = formula,</pre>
+                      <pre class="language-r">    stats::complete.cases(full_frame)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">253</td>
-                    <td class="coverage">46<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">163</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      data = data,</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">254</td>
-                    <td class="coverage">46<em>x</em></td>
+                    <td class="num">164</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      include = include,</pre>
+                      <pre class="language-r">  full_frame &lt;- full_frame[keep_ind, ]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">255</td>
-                    <td class="coverage">46<em>x</em></td>
+                    <td class="num">165</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      full = full</pre>
+                      <pre class="language-r">  if (drop_visit_levels &amp;&amp; !formula_parts$is_spatial &amp;&amp; h_extra_levels(full_frame[[formula_parts$visit_var]])) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">256</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">166</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">    visit_vec &lt;- full_frame[[formula_parts$visit_var]]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">257</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">167</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Only if include is default (full) and also data is missing, and also na.action is na.omit we will</pre>
+                      <pre class="language-r">    old_levels &lt;- levels(visit_vec)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">258</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">168</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # use the model frame from the tmb_data.</pre>
+                      <pre class="language-r">    full_frame[[formula_parts$visit_var]] &lt;- droplevels(visit_vec)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">259</td>
-                    <td class="coverage">46<em>x</em></td>
+                    <td class="num">169</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  include_choice &lt;- c("subject_var", "visit_var", "group_var", "response_var")</pre>
+                      <pre class="language-r">    new_levels &lt;- levels(full_frame[[formula_parts$visit_var]])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">260</td>
-                    <td class="coverage">46<em>x</em></td>
+                    <td class="num">170</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(data) &amp;&amp; setequal(include, include_choice) &amp;&amp; identical(h_get_na_action(na.action), stats::na.omit)) {</pre>
+                      <pre class="language-r">    dropped &lt;- setdiff(old_levels, new_levels)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">261</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">171</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ret &lt;- formula$tmb_data$full_frame</pre>
+                      <pre class="language-r">    message(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">262</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">172</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Remove weights column.</pre>
+                      <pre class="language-r">      "In ", formula_parts$visit_var, " there are dropped visits: ", toString(dropped),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">263</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">173</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ret[, "(weights)"] &lt;- NULL</pre>
+                      <pre class="language-r">      ".\n Additional attributes including contrasts are lost.\n",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">264</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">174</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ret</pre>
+                      <pre class="language-r">      "To avoid this behavior, make sure use `drop_visit_levels = FALSE`."</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">265</td>
+                    <td class="num">175</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">266</td>
+                    <td class="num">176</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Construct data frame to return to users.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">267</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">177</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ret &lt;-</pre>
+                      <pre class="language-r">  is_factor_col &lt;- vapply(full_frame, is.factor, FUN.VALUE = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">268</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">178</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      stats::model.frame(</pre>
+                      <pre class="language-r">  is_factor_col &lt;- intersect(names(is_factor_col)[is_factor_col], all.vars(formula_parts$model_formula))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">269</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">179</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        formula = lst_formula_and_data$formula,</pre>
+                      <pre class="language-r">  x_matrix &lt;- stats::model.matrix(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">270</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">180</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        data = h_get_na_action(na.action)(lst_formula_and_data$data),</pre>
+                      <pre class="language-r">    formula_parts$model_formula,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">271</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">181</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        na.action = na.action,</pre>
+                      <pre class="language-r">    data = full_frame,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">272</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">182</td>
+                    <td class="coverage">310<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        xlev = stats::.getXlevels(terms(formula), formula$tmb_data$full_frame)</pre>
+                      <pre class="language-r">    contrasts.arg = h_default_value(contrasts, lapply(full_frame[is_factor_col], contrasts))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">273</td>
+                    <td class="num">183</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">274</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">184</td>
+                    <td class="coverage">309<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  x_cols_aliased &lt;- stats::setNames(rep(FALSE, ncol(x_matrix)), nm = colnames(x_matrix))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">275</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="num">185</td>
+                    <td class="coverage">309<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret</pre>
+                      <pre class="language-r">  qr_x_mat &lt;- qr(x_matrix)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">276</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">186</td>
+                    <td class="coverage">309<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  if (qr_x_mat$rank &lt; ncol(x_matrix)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">277</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">187</td>
+                    <td class="coverage">23<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    cols_to_drop &lt;- utils::tail(qr_x_mat$pivot, ncol(x_matrix) - qr_x_mat$rank)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">278</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">188</td>
+                    <td class="coverage">23<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    if (identical(singular, "error")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">279</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">189</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Construction of Model Frame Formula and Data Inputs</pre>
+                      <pre class="language-r">      stop(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">280</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">190</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        "design matrix only has rank ", qr_x_mat$rank, " and ", length(cols_to_drop),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">281</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">191</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description</pre>
+                      <pre class="language-r">        " columns (", toString(colnames(x_matrix)[cols_to_drop]), ") could be dropped",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">282</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">192</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Input formulas are converted from mmrm-style to a style compatible</pre>
+                      <pre class="language-r">        " to achieve full rank ", ncol(x_matrix), " by using `accept_singular = TRUE`"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">283</td>
+                    <td class="num">193</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' with default [stats::model.frame()] and [stats::model.matrix()] methods.</pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">284</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">194</td>
+                    <td class="coverage">22<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    } else if (identical(singular, "drop")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">285</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">195</td>
+                    <td class="coverage">11<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' The full formula is returned so we can construct, for example, the</pre>
+                      <pre class="language-r">      assign_attr &lt;- attr(x_matrix, "assign")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">286</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">196</td>
+                    <td class="coverage">11<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `model.frame()` including all columns as well as the requested subset.</pre>
+                      <pre class="language-r">      contrasts_attr &lt;- attr(x_matrix, "contrasts")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">287</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">197</td>
+                    <td class="coverage">11<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' The full set is used to identify rows to include in the reduced model frame.</pre>
+                      <pre class="language-r">      x_matrix &lt;- x_matrix[, -cols_to_drop, drop = FALSE]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">288</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">198</td>
+                    <td class="coverage">11<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      x_cols_aliased[cols_to_drop] &lt;- TRUE</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">289</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">199</td>
+                    <td class="coverage">11<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula (`mmrm`)\cr mmrm fit object.</pre>
+                      <pre class="language-r">      attr(x_matrix, "assign") &lt;- assign_attr[-cols_to_drop]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">290</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">200</td>
+                    <td class="coverage">11<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param data optional data frame that will be</pre>
+                      <pre class="language-r">      attr(x_matrix, "contrasts") &lt;- contrasts_attr</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">291</td>
+                    <td class="num">201</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   passed to `model.frame()` or `model.matrix()`</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">292</td>
+                    <td class="num">202</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param include (`character`)\cr names of variable to include</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">293</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">203</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param full (`flag`)\cr indicator whether to return full model frame (deprecated).</pre>
+                      <pre class="language-r">  y_vector &lt;- if (has_response) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">294</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">204</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    as.numeric(stats::model.response(full_frame))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">295</td>
+                    <td class="num">205</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return named list with four elements:</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">296</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">206</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `"formula"`: the formula including the columns requested in the `include=` argument.</pre>
+                      <pre class="language-r">    rep(NA_real_, nrow(full_frame))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">297</td>
+                    <td class="num">207</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `"data"`: a data frame including all columns needed in the formula.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">298</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">208</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   full formula are identical</pre>
+                      <pre class="language-r">  weights_vector &lt;- as.numeric(stats::model.weights(full_frame))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">299</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">209</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  n_subjects &lt;- length(unique(full_frame[[formula_parts$subject_var]]))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">300</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">210</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_construct_model_frame_inputs &lt;- function(formula,</pre>
+                      <pre class="language-r">  subject_zero_inds &lt;- which(!duplicated(full_frame[[formula_parts$subject_var]])) - 1L</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">301</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">211</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                           data,</pre>
+                      <pre class="language-r">  subject_n_visits &lt;- c(utils::tail(subject_zero_inds, -1L), nrow(full_frame)) - subject_zero_inds</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">302</td>
+                    <td class="num">212</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                           include,</pre>
+                      <pre class="language-r">  # It is possible that `subject_var` is factor with more levels (and this does not affect fit)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">303</td>
+                    <td class="num">213</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                           include_choice = c("subject_var", "visit_var", "group_var", "response_var"),</pre>
+                      <pre class="language-r">  # so no check is needed for `subject_visits`.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">304</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">214</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                           full) {</pre>
+                      <pre class="language-r">  assert_true(all(subject_n_visits &gt; 0))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">305</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">215</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!missing(full) &amp;&amp; identical(full, TRUE)) {</pre>
+                      <pre class="language-r">  if (!is.null(formula_parts$group_var)) {</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">306</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">216</td>
+                    <td class="coverage">41<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    lifecycle::deprecate_warn("0.3", "model.frame.mmrm_tmb(full)")</pre>
+                      <pre class="language-r">    assert_factor(data[[formula_parts$group_var]])</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">307</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">217</td>
+                    <td class="coverage">41<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    include &lt;- include_choice</pre>
+                      <pre class="language-r">    subject_groups &lt;- full_frame[[formula_parts$group_var]][subject_zero_inds + 1L]</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">218</td>
+                    <td class="coverage">41<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    n_groups &lt;- nlevels(subject_groups)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">308</td>
+                    <td class="num">219</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  } else {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">220</td>
+                    <td class="coverage">267<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    subject_groups &lt;- factor(rep(0L, n_subjects))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">221</td>
+                    <td class="coverage">267<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    n_groups &lt;- 1L</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">309</td>
+                    <td class="num">222</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">310</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">223</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(formula, classes = "mmrm_tmb")</pre>
+                      <pre class="language-r">  coordinates &lt;- full_frame[, formula_parts$visit_var, drop = FALSE]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">311</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">224</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(include, include_choice)</pre>
+                      <pre class="language-r">  if (formula_parts$is_spatial) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">312</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">225</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(data)) {</pre>
+                      <pre class="language-r">    lapply(coordinates, assert_numeric)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">313</td>
-                    <td class="coverage">256<em>x</em></td>
+                    <td class="num">226</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    data &lt;- formula$data</pre>
+                      <pre class="language-r">    coordinates_matrix &lt;- as.matrix(coordinates)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">227</td>
+                    <td class="coverage">16<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    n_visits &lt;- max(subject_n_visits)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">314</td>
+                    <td class="num">228</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">315</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">229</td>
+                    <td class="coverage">292<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_data_frame(data)</pre>
+                      <pre class="language-r">    assert(identical(ncol(coordinates), 1L))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">316</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">230</td>
+                    <td class="coverage">292<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    assert_factor(coordinates[[1L]])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">317</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">231</td>
+                    <td class="coverage">292<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  drop_response &lt;- !"response_var" %in% include</pre>
+                      <pre class="language-r">    coordinates_matrix &lt;- as.matrix(as.integer(coordinates[[1L]]) - 1, ncol = 1)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">318</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">232</td>
+                    <td class="coverage">292<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  add_vars &lt;- unlist(formula$formula_parts[include])</pre>
+                      <pre class="language-r">    n_visits &lt;- nlevels(coordinates[[1L]])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">319</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">233</td>
+                    <td class="coverage">292<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  new_formula &lt;- h_add_terms(formula$formula_parts$model_formula, add_vars, drop_response)</pre>
+                      <pre class="language-r">    assert_true(all(subject_n_visits &lt;= n_visits))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">320</td>
+                    <td class="num">234</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">321</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">235</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  drop_response_full &lt;- !"response_var" %in% include_choice</pre>
+                      <pre class="language-r">  structure(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">322</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">236</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  add_vars_full &lt;- unlist(formula$formula_parts[include_choice])</pre>
+                      <pre class="language-r">    list(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">323</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">237</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  new_formula_full &lt;-</pre>
+                      <pre class="language-r">      full_frame = full_frame,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">324</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">238</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_add_terms(formula$formula_parts$model_formula, add_vars_full, drop_response_full)</pre>
+                      <pre class="language-r">      data = data,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">325</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">239</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      x_matrix = x_matrix,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">326</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">240</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Update data based on the columns in the full formula return.</pre>
+                      <pre class="language-r">      x_cols_aliased = x_cols_aliased,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">327</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">241</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  all_vars &lt;- all.vars(new_formula_full)</pre>
+                      <pre class="language-r">      coordinates = coordinates_matrix,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">328</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">242</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_names(colnames(data), must.include = all_vars)</pre>
+                      <pre class="language-r">      y_vector = y_vector,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">329</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">243</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  data &lt;- data[, all_vars, drop = FALSE]</pre>
+                      <pre class="language-r">      weights_vector = weights_vector,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">330</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">244</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      n_visits = n_visits,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">331</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">245</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Return list with updated formula, data.</pre>
+                      <pre class="language-r">      n_subjects = n_subjects,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">332</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">246</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">      subject_zero_inds = subject_zero_inds,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">333</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">247</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    formula = new_formula,</pre>
+                      <pre class="language-r">      subject_n_visits = subject_n_visits,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">334</td>
-                    <td class="coverage">280<em>x</em></td>
+                    <td class="num">248</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    data = data</pre>
+                      <pre class="language-r">      cov_type = formula_parts$cov_type,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">335</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">249</td>
+                    <td class="coverage">308<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">      is_spatial_int = as.integer(formula_parts$is_spatial),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">250</td>
+                    <td class="coverage">308<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      reml = as.integer(reml),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">251</td>
+                    <td class="coverage">308<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      subject_groups = subject_groups,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">252</td>
+                    <td class="coverage">308<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      n_groups = n_groups</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">336</td>
+                    <td class="num">253</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    ),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">254</td>
+                    <td class="coverage">308<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    class = "mmrm_tmb_data"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">337</td>
+                    <td class="num">255</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">338</td>
+                    <td class="num">256</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the model matrix.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">339</td>
+                    <td class="num">257</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">340</td>
+                    <td class="num">258</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param use_response (`flag`)\cr whether to use the response for complete rows.</pre>
+                      <pre class="language-r">#' Start Parameters for `TMB` Fit</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">341</td>
+                    <td class="num">259</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">342</td>
+                    <td class="num">260</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @param formula_parts (`mmrm_tmb_formula_parts`)\cr produced by</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">343</td>
+                    <td class="num">261</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Model matrix:</pre>
+                      <pre class="language-r">#'  [h_mmrm_tmb_formula_parts()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">344</td>
+                    <td class="num">262</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' model.matrix(object)</pre>
+                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">345</td>
+                    <td class="num">263</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">model.matrix.mmrm_tmb &lt;- function(object, data, use_response = TRUE, ...) { # nolint</pre>
+                      <pre class="language-r">#' @param start (`numeric` or `NULL`)\cr optional start values for variance</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">346</td>
+                    <td class="num">264</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Always return the utilized model matrix if data not provided.</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">347</td>
-                    <td class="coverage">37<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(data)) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">348</td>
-                    <td class="coverage">3<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    return(object$tmb_data$x_matrix)</pre>
+                      <pre class="language-r">#'   parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">349</td>
+                    <td class="num">265</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">350</td>
-                    <td class="coverage">34<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  stats::model.matrix(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">351</td>
-                    <td class="coverage">34<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    h_add_terms(object$formula_parts$model_formula, NULL, drop_response = !use_response),</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">352</td>
-                    <td class="coverage">34<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    data = data,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">353</td>
-                    <td class="coverage">34<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    contrasts.arg = attr(object$tmb_data$x_matrix, "contrasts"),</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">354</td>
-                    <td class="coverage">34<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    xlev = component(object, "xlev"),</pre>
+                      <pre class="language-r">#' @param n_groups (`int`)\cr number of groups.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">355</td>
+                    <td class="num">266</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ...</pre>
+                      <pre class="language-r">#' @return List with element `theta` containing the start values for the variance</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">356</td>
+                    <td class="num">267</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'   parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">357</td>
+                    <td class="num">268</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">358</td>
+                    <td class="num">269</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">359</td>
+                    <td class="num">270</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the terms object.</pre>
+                      <pre class="language-r">h_mmrm_tmb_parameters &lt;- function(formula_parts,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">360</td>
+                    <td class="num">271</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats model.frame</pre>
+                      <pre class="language-r">                                  tmb_data,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">361</td>
+                    <td class="num">272</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">                                  start,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">362</td>
+                    <td class="num">273</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                                  n_groups = 1L) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">363</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">274</td>
+                    <td class="coverage">265<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">  assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">364</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">275</td>
+                    <td class="coverage">265<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # terms:</pre>
+                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">365</td>
+                    <td class="num">276</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' terms(object)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">366</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">277</td>
+                    <td class="coverage">265<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' terms(object, include = "subject_var")</pre>
+                      <pre class="language-r">  m &lt;- tmb_data$n_visits</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">367</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">278</td>
+                    <td class="coverage">265<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">terms.mmrm_tmb &lt;- function(x, include = "response_var", ...) { # nolint</pre>
+                      <pre class="language-r">  start_value0 &lt;- std_start(formula_parts$cov_type, m, n_groups)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">368</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">279</td>
+                    <td class="coverage">265<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Construct updated formula and data arguments.</pre>
+                      <pre class="language-r">  theta_dim &lt;- length(start_value0)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">369</td>
-                    <td class="coverage">231<em>x</em></td>
+                    <td class="num">280</td>
+                    <td class="coverage">265<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  lst_formula_and_data &lt;-</pre>
+                      <pre class="language-r">  start_values &lt;- if (is.null(start)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">370</td>
-                    <td class="coverage">231<em>x</em></td>
+                    <td class="num">281</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_construct_model_frame_inputs(</pre>
+                      <pre class="language-r">    start_value0</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">371</td>
-                    <td class="coverage">231<em>x</em></td>
+                    <td class="num">282</td>
+                    <td class="coverage">265<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      formula = x,</pre>
+                      <pre class="language-r">  } else if (test_function(start)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">372</td>
-                    <td class="coverage">231<em>x</em></td>
+                    <td class="num">283</td>
+                    <td class="coverage">233<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      include = include</pre>
+                      <pre class="language-r">    do.call(start, utils::modifyList(formula_parts, tmb_data))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">373</td>
+                    <td class="num">284</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">374</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">285</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    start</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">375</td>
+                    <td class="num">286</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Use formula method for `terms()` to construct the mmrm terms object.</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">376</td>
-                    <td class="coverage">231<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  stats::terms(</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">377</td>
-                    <td class="coverage">231<em>x</em></td>
+                    <td class="num">287</td>
+                    <td class="coverage">264<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x = lst_formula_and_data$formula,</pre>
+                      <pre class="language-r">  assert_numeric(start_values, len = theta_dim, any.missing = FALSE, finite = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">378</td>
-                    <td class="coverage">231<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    data = lst_formula_and_data$data</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">379</td>
-                    <td class="coverage"></td>
+                    <td class="num">288</td>
+                    <td class="coverage">262<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  list(theta = start_values)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">380</td>
+                    <td class="num">289</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">381</td>
+                    <td class="num">290</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">382</td>
+                    <td class="num">291</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' Asserting Sane Start Values for `TMB` Fit</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">383</td>
+                    <td class="num">292</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the attained log likelihood value.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">384</td>
+                    <td class="num">293</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats logLik</pre>
+                      <pre class="language-r">#' @param tmb_object (`list`)\cr created with [TMB::MakeADFun()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">385</td>
+                    <td class="num">294</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">386</td>
+                    <td class="num">295</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @return Nothing, only used for assertions.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">387</td>
+                    <td class="num">296</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Log likelihood given the estimated parameters:</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">388</td>
+                    <td class="num">297</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' logLik(object)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">389</td>
+                    <td class="num">298</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">logLik.mmrm_tmb &lt;- function(object, ...) {</pre>
+                      <pre class="language-r">h_mmrm_tmb_assert_start &lt;- function(tmb_object) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">390</td>
-                    <td class="coverage">50<em>x</em></td>
+                    <td class="num">299</td>
+                    <td class="coverage">249<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  -component(object, "neg_log_lik")</pre>
+                      <pre class="language-r">  assert_list(tmb_object)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">391</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">300</td>
+                    <td class="coverage">249<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  assert_subset(c("fn", "gr", "par"), names(tmb_object))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">392</td>
+                    <td class="num">301</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">393</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">302</td>
+                    <td class="coverage">249<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the used formula.</pre>
+                      <pre class="language-r">  if (is.na(tmb_object$fn(tmb_object$par))) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">394</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">303</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats formula</pre>
+                      <pre class="language-r">    stop("negative log-likelihood is NaN at starting parameter values")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">395</td>
+                    <td class="num">304</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">396</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">305</td>
+                    <td class="coverage">248<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">  if (any(is.na(tmb_object$gr(tmb_object$par)))) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">397</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">306</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Formula which was used:</pre>
+                      <pre class="language-r">    stop("some elements of gradient are NaN at starting parameter values")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">398</td>
+                    <td class="num">307</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' formula(object)</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">399</td>
+                    <td class="num">308</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">formula.mmrm_tmb &lt;- function(x, ...) {</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">400</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">309</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x$formula_parts$formula</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">401</td>
+                    <td class="num">310</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' Checking the `TMB` Optimization Result</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">402</td>
+                    <td class="num">311</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">403</td>
+                    <td class="num">312</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the variance-covariance matrix estimate</pre>
+                      <pre class="language-r">#' @param tmb_opt (`list`)\cr optimization result.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">404</td>
+                    <td class="num">313</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   for the coefficients.</pre>
+                      <pre class="language-r">#' @param mmrm_tmb (`mmrm_tmb`)\cr result from [h_mmrm_tmb_fit()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">405</td>
+                    <td class="num">314</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats vcov</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">406</td>
+                    <td class="num">315</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">#' @return Nothing, only used to generate warnings in case that the model</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">407</td>
+                    <td class="num">316</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' did not converge.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">408</td>
+                    <td class="num">317</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Variance-covariance matrix estimate for coefficients:</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">409</td>
+                    <td class="num">318</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' vcov(object)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">410</td>
+                    <td class="num">319</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">vcov.mmrm_tmb &lt;- function(object, complete = TRUE, ...) {</pre>
+                      <pre class="language-r">h_mmrm_tmb_check_conv &lt;- function(tmb_opt, mmrm_tmb) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">411</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">320</td>
+                    <td class="coverage">245<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(complete)</pre>
+                      <pre class="language-r">  assert_list(tmb_opt)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">412</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">321</td>
+                    <td class="coverage">245<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  nm &lt;- if (complete) "beta_vcov_complete" else "beta_vcov"</pre>
+                      <pre class="language-r">  assert_subset(c("par", "objective", "convergence", "message"), names(tmb_opt))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">413</td>
-                    <td class="coverage">3<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  component(object, name = nm)</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">414</td>
-                    <td class="coverage"></td>
+                    <td class="num">322</td>
+                    <td class="coverage">245<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  assert_class(mmrm_tmb, "mmrm_tmb")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">415</td>
+                    <td class="num">323</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">416</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">324</td>
+                    <td class="coverage">245<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the variance-covariance matrix estimate</pre>
+                      <pre class="language-r">  if (!is.null(tmb_opt$convergence) &amp;&amp; tmb_opt$convergence != 0) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">417</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">325</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   for the residuals.</pre>
+                      <pre class="language-r">    warning("Model convergence problem: ", tmb_opt$message, ".")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">418</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">326</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param sigma cannot be used (this parameter does not exist in MMRM).</pre>
+                      <pre class="language-r">    return()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">419</td>
+                    <td class="num">327</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom nlme VarCorr</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">420</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">328</td>
+                    <td class="coverage">242<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export VarCorr</pre>
+                      <pre class="language-r">  theta_vcov &lt;- mmrm_tmb$theta_vcov</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">421</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">329</td>
+                    <td class="coverage">242<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @aliases VarCorr</pre>
+                      <pre class="language-r">  if (is(theta_vcov, "try-error")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">422</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">330</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">    warning("Model convergence problem: hessian is singular, theta_vcov not available.")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">423</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">331</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">    return()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">424</td>
+                    <td class="num">332</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Variance-covariance matrix estimate for residuals:</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">425</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">333</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' VarCorr(object)</pre>
+                      <pre class="language-r">  if (!all(is.finite(theta_vcov))) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">426</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">334</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">VarCorr.mmrm_tmb &lt;- function(x, sigma = NA, ...) { # nolint</pre>
+                      <pre class="language-r">    warning("Model convergence problem: theta_vcov contains non-finite values.")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">427</td>
-                    <td class="coverage">10<em>x</em></td>
+                    <td class="num">335</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_scalar_na(sigma)</pre>
+                      <pre class="language-r">    return()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">428</td>
+                    <td class="num">336</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">429</td>
-                    <td class="coverage">10<em>x</em></td>
+                    <td class="num">337</td>
+                    <td class="coverage">236<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  component(x, name = "varcor")</pre>
+                      <pre class="language-r">  eigen_vals &lt;- eigen(theta_vcov, only.values = TRUE)$values</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">430</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">338</td>
+                    <td class="coverage">236<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  if (mode(eigen_vals) == "complex" || any(eigen_vals &lt;= 0)) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">431</td>
+                    <td class="num">339</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    # Note: complex eigen values signal that the matrix is not symmetric, therefore not positive definite.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">432</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">340</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the deviance, which is defined here</pre>
+                      <pre class="language-r">    warning("Model convergence problem: theta_vcov is not positive definite.")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">433</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">341</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   as twice the negative log likelihood, which can either be integrated</pre>
+                      <pre class="language-r">    return()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">434</td>
+                    <td class="num">342</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   over the coefficients for REML fits or the usual one for ML fits.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">435</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">343</td>
+                    <td class="coverage">233<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats deviance</pre>
+                      <pre class="language-r">  qr_rank &lt;- qr(theta_vcov)$rank</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">436</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">344</td>
+                    <td class="coverage">233<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">  if (qr_rank &lt; ncol(theta_vcov)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">437</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">345</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">    warning("Model convergence problem: theta_vcov is numerically singular.")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">438</td>
+                    <td class="num">346</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # REML criterion (twice the negative log likelihood):</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">439</td>
+                    <td class="num">347</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' deviance(object)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">440</td>
+                    <td class="num">348</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">deviance.mmrm_tmb &lt;- function(object, ...) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">441</td>
-                    <td class="coverage">74<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  2 * component(object, "neg_log_lik")</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">442</td>
+                    <td class="num">349</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' Extract covariance matrix from `TMB` report and input data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">443</td>
+                    <td class="num">350</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">444</td>
+                    <td class="num">351</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the Akaike Information Criterion,</pre>
+                      <pre class="language-r">#' This helper does some simple post-processing to extract covariance matrix or named</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">445</td>
+                    <td class="num">352</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   where the degrees of freedom are the number of variance parameters (`n_theta`).</pre>
+                      <pre class="language-r">#' list of covariance matrices if the fitting is using grouped covariance matrices.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">446</td>
+                    <td class="num">353</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   If `corrected`, then this is multiplied with `m / (m - n_theta - 1)` where</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">447</td>
+                    <td class="num">354</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `m` is the number of observations minus the number of coefficients, or</pre>
+                      <pre class="language-r">#' @param tmb_report (`list`)\cr report created with [TMB::MakeADFun()] report function.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">448</td>
+                    <td class="num">355</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `n_theta + 2` if it is smaller than that \insertCite{hurvich1989regression,burnham1998practical}{mmrm}.</pre>
+                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">449</td>
+                    <td class="num">356</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param corrected (`flag`)\cr whether corrected AIC should be calculated.</pre>
+                      <pre class="language-r">#' @param visit_var (`character`)\cr character vector of the visit variable</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">450</td>
+                    <td class="num">357</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param k (`number`)\cr the penalty per parameter to be used; default `k = 2`</pre>
+                      <pre class="language-r">#' @param is_spatial (`flag`)\cr indicator whether the covariance structure is spatial.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">451</td>
+                    <td class="num">358</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   is the classical AIC.</pre>
+                      <pre class="language-r">#' @return Return a simple covariance matrix if there is no grouping, or a named</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">452</td>
+                    <td class="num">359</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats AIC</pre>
+                      <pre class="language-r">#' list of estimated grouped covariance matrices,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">453</td>
+                    <td class="num">360</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">#' with its name equal to the group levels.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">454</td>
+                    <td class="num">361</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">455</td>
+                    <td class="num">362</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # AIC:</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">456</td>
+                    <td class="num">363</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' AIC(object)</pre>
+                      <pre class="language-r">h_mmrm_tmb_extract_cov &lt;- function(tmb_report, tmb_data, visit_var, is_spatial) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">457</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">364</td>
+                    <td class="coverage">241<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' AIC(object, corrected = TRUE)</pre>
+                      <pre class="language-r">  d &lt;- dim(tmb_report$covariance_lower_chol)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">458</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">365</td>
+                    <td class="coverage">241<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @references</pre>
+                      <pre class="language-r">  visit_names &lt;- if (!is_spatial) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">459</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">366</td>
+                    <td class="coverage">228<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - \insertRef{hurvich1989regression}{mmrm}</pre>
+                      <pre class="language-r">    levels(tmb_data$full_frame[[visit_var]])</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">460</td>
+                    <td class="num">367</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - \insertRef{burnham1998practical}{mmrm}</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">461</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">368</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">AIC.mmrm_tmb &lt;- function(object, corrected = FALSE, ..., k = 2) {</pre>
+                      <pre class="language-r">    c(0, 1)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">462</td>
+                    <td class="num">369</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # nolint</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">463</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">370</td>
+                    <td class="coverage">241<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(corrected)</pre>
+                      <pre class="language-r">  cov &lt;- lapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">464</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">371</td>
+                    <td class="coverage">241<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(k, lower = 1)</pre>
+                      <pre class="language-r">    seq_len(d[1] / d[2]),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">465</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">372</td>
+                    <td class="coverage">241<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    function(i) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">466</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">373</td>
+                    <td class="coverage">278<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_theta &lt;- length(component(object, "theta_est"))</pre>
+                      <pre class="language-r">      ret &lt;- tcrossprod(tmb_report$covariance_lower_chol[seq(1 + (i - 1) * d[2], i * d[2]), ])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">467</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">374</td>
+                    <td class="coverage">278<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- if (!corrected) {</pre>
+                      <pre class="language-r">      dimnames(ret) &lt;- list(visit_names, visit_names)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">468</td>
-                    <td class="coverage">43<em>x</em></td>
+                    <td class="num">375</td>
+                    <td class="coverage">278<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_theta</pre>
+                      <pre class="language-r">      return(ret)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">469</td>
+                    <td class="num">376</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">470</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">377</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_obs &lt;- length(component(object, "y_vector"))</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">471</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">378</td>
+                    <td class="coverage">241<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_beta &lt;- length(component(object, "beta_est"))</pre>
+                      <pre class="language-r">  if (identical(tmb_data$n_groups, 1L)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">472</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">379</td>
+                    <td class="coverage">204<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    m &lt;- max(n_theta + 2, n_obs - n_beta)</pre>
+                      <pre class="language-r">    cov &lt;- cov[[1]]</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">473</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">380</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_theta * (m / (m - n_theta - 1))</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">474</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">381</td>
+                    <td class="coverage">37<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    names(cov) &lt;- levels(tmb_data$subject_groups)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">475</td>
+                    <td class="num">382</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">476</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="num">383</td>
+                    <td class="coverage">241<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  2 * component(object, "neg_log_lik") + k * df</pre>
+                      <pre class="language-r">  return(cov)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">477</td>
+                    <td class="num">384</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">478</td>
+                    <td class="num">385</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">479</td>
+                    <td class="num">386</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the Bayesian Information Criterion,</pre>
+                      <pre class="language-r">#' Build `TMB` Fit Result List</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">480</td>
+                    <td class="num">387</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   which is using the natural logarithm of the number of subjects for the</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">481</td>
+                    <td class="num">388</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   penalty parameter `k`.</pre>
+                      <pre class="language-r">#' This helper does some simple post-processing of the `TMB` object and</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">482</td>
+                    <td class="num">389</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats BIC</pre>
+                      <pre class="language-r">#' optimization results, including setting names, inverting matrices etc.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">483</td>
+                    <td class="num">390</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">484</td>
+                    <td class="num">391</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @param tmb_object (`list`)\cr created with [TMB::MakeADFun()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">485</td>
+                    <td class="num">392</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # BIC:</pre>
+                      <pre class="language-r">#' @param tmb_opt (`list`)\cr optimization result.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">486</td>
+                    <td class="num">393</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' BIC(object)</pre>
+                      <pre class="language-r">#' @param formula_parts (`mmrm_tmb_formula_parts`)\cr produced by</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">487</td>
+                    <td class="num">394</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">BIC.mmrm_tmb &lt;- function(object, ...) {</pre>
+                      <pre class="language-r">#'  [h_mmrm_tmb_formula_parts()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">488</td>
+                    <td class="num">395</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # nolint</pre>
+                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">489</td>
-                    <td class="coverage">21<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">396</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  k &lt;- log(component(object, "n_subjects"))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">490</td>
-                    <td class="coverage">21<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">397</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  AIC(object, corrected = FALSE, k = k)</pre>
+                      <pre class="language-r">#' @return List of class `mmrm_tmb` with:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">491</td>
+                    <td class="num">398</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'   - `cov`: estimated covariance matrix, or named list of estimated group specific covariance matrices.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">492</td>
+                    <td class="num">399</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   - `beta_est`: vector of coefficient estimates.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">493</td>
+                    <td class="num">400</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   - `beta_vcov`: Variance-covariance matrix for coefficient estimates.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">494</td>
+                    <td class="num">401</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods prints the object.</pre>
+                      <pre class="language-r">#'   - `beta_vcov_inv_L`: Lower triangular matrix `L` of the inverse variance-covariance matrix decomposition.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">495</td>
+                    <td class="num">402</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">#'   - `beta_vcov_inv_D`: vector of diagonal matrix `D` of the inverse variance-covariance matrix decomposition.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">496</td>
+                    <td class="num">403</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">print.mmrm_tmb &lt;- function(x,</pre>
+                      <pre class="language-r">#'   - `theta_est`: vector of variance parameter estimates.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">497</td>
+                    <td class="num">404</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                           ...) {</pre>
+                      <pre class="language-r">#'   - `theta_vcov`: variance-covariance matrix for variance parameter estimates.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">498</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">405</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("mmrm fit\n\n")</pre>
+                      <pre class="language-r">#'   - `neg_log_lik`: obtained negative log-likelihood.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">499</td>
+                    <td class="num">406</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   - `formula_parts`: input.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">500</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">407</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_print_call(</pre>
+                      <pre class="language-r">#'   - `data`: input.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">501</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">408</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    component(x, "call"), component(x, "n_obs"),</pre>
+                      <pre class="language-r">#'   - `weights`: input.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">502</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">409</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    component(x, "n_subjects"), component(x, "n_timepoints")</pre>
+                      <pre class="language-r">#'   - `reml`: input as a flag.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">503</td>
+                    <td class="num">410</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'   - `opt_details`: list with optimization details including convergence code.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">504</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">411</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_print_cov(component(x, "cov_type"), component(x, "n_theta"), component(x, "n_groups"))</pre>
+                      <pre class="language-r">#'   - `tmb_object`: original `TMB` object created with [TMB::MakeADFun()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">505</td>
+                    <td class="num">412</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   - `tmb_data`: input.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">506</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">413</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("Inference:   ")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">507</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">414</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat(ifelse(component(x, "reml"), "REML", "ML"))</pre>
+                      <pre class="language-r">#' @details Instead of inverting or decomposing `beta_vcov`, it can be more efficient to use its robust</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">508</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">415</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n")</pre>
+                      <pre class="language-r">#'   Cholesky decomposition `LDL^T`, therefore we return the corresponding two components `L` and `D`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">509</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">416</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("Deviance:    ")</pre>
+                      <pre class="language-r">#'   as well since they have been available on the `C++` side already.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">510</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">417</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat(deviance(x))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">511</td>
+                    <td class="num">418</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">512</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">419</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n\nCoefficients: ")</pre>
+                      <pre class="language-r">h_mmrm_tmb_fit &lt;- function(tmb_object,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">513</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">420</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_singular_coefs &lt;- sum(component(x, "beta_aliased"))</pre>
+                      <pre class="language-r">                           tmb_opt,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">514</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">421</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (n_singular_coefs &gt; 0) {</pre>
+                      <pre class="language-r">                           formula_parts,</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">422</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">                           tmb_data) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">515</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">423</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cat("(", n_singular_coefs, " not defined because of singularities)", sep = "")</pre>
+                      <pre class="language-r">  assert_list(tmb_object)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">516</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">424</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  assert_subset(c("fn", "gr", "par", "he"), names(tmb_object))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">517</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">425</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n")</pre>
+                      <pre class="language-r">  assert_list(tmb_opt)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">518</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">426</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  print(coef(x, complete = TRUE))</pre>
+                      <pre class="language-r">  assert_subset(c("par", "objective", "convergence", "message"), names(tmb_opt))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">519</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">427</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">520</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">428</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("\nModel Inference Optimization:")</pre>
+                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">521</td>
+                    <td class="num">429</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">522</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">430</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat(ifelse(component(x, "convergence") == 0, "\nConverged", "\nFailed to converge"))</pre>
+                      <pre class="language-r">  tmb_report &lt;- tmb_object$report(par = tmb_opt$par)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">523</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">431</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat(</pre>
+                      <pre class="language-r">  x_matrix_cols &lt;- colnames(tmb_data$x_matrix)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">524</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">432</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    " with code", component(x, "convergence"),</pre>
+                      <pre class="language-r">  cov &lt;- h_mmrm_tmb_extract_cov(tmb_report, tmb_data, formula_parts$visit_var, formula_parts$is_spatial)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">525</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">433</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "and message:",</pre>
+                      <pre class="language-r">  beta_est &lt;- tmb_report$beta</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">526</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">434</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (is.null(component(x, "conv_message"))) "No message provided." else tolower(component(x, "conv_message"))</pre>
+                      <pre class="language-r">  names(beta_est) &lt;- x_matrix_cols</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">527</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">435</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  beta_vcov &lt;- tmb_report$beta_vcov</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">528</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">436</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n")</pre>
+                      <pre class="language-r">  dimnames(beta_vcov) &lt;- list(x_matrix_cols, x_matrix_cols)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">529</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">437</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  invisible(x)</pre>
+                      <pre class="language-r">  beta_vcov_inv_L &lt;- tmb_report$XtWX_L # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">530</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">438</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  beta_vcov_inv_D &lt;- tmb_report$XtWX_D # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">531</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">439</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  theta_est &lt;- tmb_opt$par</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">532</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">440</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  names(theta_est) &lt;- NULL</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">533</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">441</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods to obtain residuals - either unscaled ('response'), 'pearson' or 'normalized'.</pre>
+                      <pre class="language-r">  theta_vcov &lt;- try(solve(tmb_object$he(tmb_opt$par)), silent = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">534</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">442</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param type (`string`)\cr unscaled (`response`), `pearson` or `normalized`. Default is `response`,</pre>
+                      <pre class="language-r">  opt_details_names &lt;- setdiff(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">535</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">443</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' and this is the only type available for use with models with a spatial covariance structure.</pre>
+                      <pre class="language-r">    names(tmb_opt),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">536</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">444</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats residuals</pre>
+                      <pre class="language-r">    c("par", "objective")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">537</td>
+                    <td class="num">445</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">538</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">446</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">  structure(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">539</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">447</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # residuals:</pre>
+                      <pre class="language-r">    list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">540</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">448</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' residuals(object, type = "response")</pre>
+                      <pre class="language-r">      cov = cov,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">541</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">449</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' residuals(object, type = "pearson")</pre>
+                      <pre class="language-r">      beta_est = beta_est,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">542</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">450</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' residuals(object, type = "normalized")</pre>
+                      <pre class="language-r">      beta_vcov = beta_vcov,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">543</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">451</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @references</pre>
+                      <pre class="language-r">      beta_vcov_inv_L = beta_vcov_inv_L,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">544</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">452</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - \insertRef{galecki2013linear}{mmrm}</pre>
+                      <pre class="language-r">      beta_vcov_inv_D = beta_vcov_inv_D,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">545</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">453</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">residuals.mmrm_tmb &lt;- function(object, type = c("response", "pearson", "normalized"), ...) {</pre>
+                      <pre class="language-r">      theta_est = theta_est,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">546</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">454</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
+                      <pre class="language-r">      theta_vcov = theta_vcov,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">547</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">455</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  switch(type,</pre>
+                      <pre class="language-r">      neg_log_lik = tmb_opt$objective,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">548</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="num">456</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "response" = h_residuals_response(object),</pre>
+                      <pre class="language-r">      formula_parts = formula_parts,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">549</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">457</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "pearson" = h_residuals_pearson(object),</pre>
+                      <pre class="language-r">      data = tmb_data$data,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">550</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">458</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "normalized" = h_residuals_normalized(object)</pre>
+                      <pre class="language-r">      weights = tmb_data$weights_vector,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">551</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">459</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">      reml = as.logical(tmb_data$reml),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">552</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">460</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      opt_details = tmb_opt[opt_details_names],</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">553</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">461</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculate Pearson Residuals</pre>
+                      <pre class="language-r">      tmb_object = tmb_object,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">554</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">462</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      tmb_data = tmb_data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">555</td>
+                    <td class="num">463</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used by [residuals.mmrm_tmb()] to calculate Pearson residuals.</pre>
+                      <pre class="language-r">    ),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">556</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">464</td>
+                    <td class="coverage">239<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    class = "mmrm_tmb"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">557</td>
+                    <td class="num">465</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">558</td>
+                    <td class="num">466</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">559</td>
+                    <td class="num">467</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Vector of residuals.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">560</td>
+                    <td class="num">468</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Low-Level Fitting Function for MMRM</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">561</td>
+                    <td class="num">469</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">562</td>
+                    <td class="num">470</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_residuals_pearson &lt;- function(object) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">563</td>
-                    <td class="coverage">6<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">564</td>
-                    <td class="coverage">6<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">471</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_residuals_response(object) * object$tmb_object$report()$diag_cov_inv_sqrt</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">565</td>
+                    <td class="num">472</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' This is the low-level function to fit an MMRM. Note that this does not</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">566</td>
+                    <td class="num">473</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' try different optimizers or adds Jacobian information etc. in contrast to</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">567</td>
+                    <td class="num">474</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculate normalized residuals</pre>
+                      <pre class="language-r">#' [mmrm()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">568</td>
+                    <td class="num">475</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">569</td>
+                    <td class="num">476</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used by [residuals.mmrm_tmb()] to calculate normalized / scaled residuals.</pre>
+                      <pre class="language-r">#' @param formula (`formula`)\cr model formula with exactly one special term</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">570</td>
+                    <td class="num">477</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   specifying the visits within subjects, see details.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">571</td>
+                    <td class="num">478</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r">#' @param data (`data.frame`)\cr input data containing the variables used in</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">572</td>
+                    <td class="num">479</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   `formula`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">573</td>
+                    <td class="num">480</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Vector of residuals</pre>
+                      <pre class="language-r">#' @param weights (`vector`)\cr input vector containing the weights.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">574</td>
+                    <td class="num">481</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @inheritParams h_mmrm_tmb_data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">575</td>
+                    <td class="num">482</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @param covariance (`cov_struct`)\cr A covariance structure type definition,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">576</td>
+                    <td class="num">483</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_residuals_normalized &lt;- function(object) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">577</td>
-                    <td class="coverage">8<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">578</td>
-                    <td class="coverage">8<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  object$tmb_object$report()$epsilonTilde</pre>
+                      <pre class="language-r">#'   or value that can be coerced to a covariance structure using</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">579</td>
+                    <td class="num">484</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'   [as.cov_struct()]. If no value is provided, a structure is derived from</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">580</td>
+                    <td class="num">485</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculate response residuals.</pre>
+                      <pre class="language-r">#'   the provided formula.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">581</td>
+                    <td class="num">486</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param control (`mmrm_control`)\cr list of control options produced by</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">582</td>
+                    <td class="num">487</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used by [residuals.mmrm_tmb()] to calculate response residuals.</pre>
+                      <pre class="language-r">#'   [mmrm_control()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">583</td>
+                    <td class="num">488</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @inheritParams fit_single_optimizer</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">584</td>
+                    <td class="num">489</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">585</td>
+                    <td class="num">490</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @return List of class `mmrm_tmb`, see [h_mmrm_tmb_fit()] for details.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">586</td>
+                    <td class="num">491</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Vector of residuals</pre>
+                      <pre class="language-r">#'   In addition, it contains elements `call` and `optimizer`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">587</td>
+                    <td class="num">492</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">588</td>
+                    <td class="num">493</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">589</td>
+                    <td class="num">494</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_residuals_response &lt;- function(object) {</pre>
+                      <pre class="language-r">#' The `formula` typically looks like:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">590</td>
-                    <td class="coverage">15<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">495</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">591</td>
-                    <td class="coverage">15<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">496</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  component(object, "y_vector") - unname(fitted(object))</pre>
+                      <pre class="language-r">#' `FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">592</td>
+                    <td class="num">497</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">593</td>
+                    <td class="num">498</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' which specifies response and covariates as usual, and exactly one special term</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">594</td>
+                    <td class="num">499</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tmb_methods simulate responses from a fitted model according</pre>
+                      <pre class="language-r">#' defines which covariance structure is used and what are the visit and</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">595</td>
+                    <td class="num">500</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   to the simulation `method`, returning a `data.frame` of dimension `[n, m]`</pre>
+                      <pre class="language-r">#' subject variables.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">596</td>
+                    <td class="num">501</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   where n is the number of rows in `newdata`,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">597</td>
+                    <td class="num">502</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   and m is the number `nsim` of simulated responses.</pre>
+                      <pre class="language-r">#' Always use only the first optimizer if multiple optimizers are provided.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">598</td>
+                    <td class="num">503</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">599</td>
+                    <td class="num">504</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param seed unused argument from [stats::simulate()].</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">600</td>
+                    <td class="num">505</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param method (`string`)\cr simulation method to use. If "conditional",</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">601</td>
+                    <td class="num">506</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   simulated values are sampled given the estimated covariance matrix of `object`.</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">602</td>
+                    <td class="num">507</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   If "marginal", the variance of the estimated covariance matrix is taken into account.</pre>
+                      <pre class="language-r">#' formula &lt;- FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">603</td>
+                    <td class="num">508</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' data &lt;- fev_data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">604</td>
+                    <td class="num">509</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @importFrom stats simulate</pre>
+                      <pre class="language-r">#' system.time(result &lt;- fit_mmrm(formula, data, rep(1, nrow(fev_data))))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">605</td>
+                    <td class="num">510</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">fit_mmrm &lt;- function(formula,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">606</td>
+                    <td class="num">511</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">simulate.mmrm_tmb &lt;- function(object,</pre>
+                      <pre class="language-r">                     data,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">607</td>
+                    <td class="num">512</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                              nsim = 1,</pre>
+                      <pre class="language-r">                     weights,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">608</td>
+                    <td class="num">513</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                              seed = NULL,</pre>
+                      <pre class="language-r">                     reml = TRUE,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">609</td>
+                    <td class="num">514</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                              newdata,</pre>
+                      <pre class="language-r">                     covariance = NULL,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">610</td>
+                    <td class="num">515</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                              ...,</pre>
+                      <pre class="language-r">                     tmb_data,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">611</td>
+                    <td class="num">516</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                              method = c("conditional", "marginal")) {</pre>
+                      <pre class="language-r">                     formula_parts,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">612</td>
-                    <td class="coverage">15<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">517</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
+                      <pre class="language-r">                     control = mmrm_control()) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">613</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">518</td>
+                    <td class="coverage">252<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_null(seed)</pre>
+                      <pre class="language-r">  if (missing(formula_parts) || missing(tmb_data)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">614</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">519</td>
+                    <td class="coverage">67<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(newdata)) {</pre>
+                      <pre class="language-r">    covariance &lt;- h_reconcile_cov_struct(formula, covariance)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">615</td>
-                    <td class="coverage">12<em>x</em></td>
+                    <td class="num">520</td>
+                    <td class="coverage">65<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    newdata &lt;- object$data</pre>
+                      <pre class="language-r">    formula_parts &lt;- h_mmrm_tmb_formula_parts(formula, covariance)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">616</td>
+                    <td class="num">521</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">617</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">522</td>
+                    <td class="coverage">65<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_data_frame(newdata)</pre>
+                      <pre class="language-r">    if (!formula_parts$is_spatial &amp;&amp; !is.factor(data[[formula_parts$visit_var]])) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">618</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">523</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  method &lt;- match.arg(method)</pre>
+                      <pre class="language-r">      stop("Time variable must be a factor for non-spatial covariance structures")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">619</td>
+                    <td class="num">524</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">620</td>
+                    <td class="num">525</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">621</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">526</td>
+                    <td class="coverage">64<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tmb_data &lt;- h_mmrm_tmb_data(</pre>
+                      <pre class="language-r">    assert_class(control, "mmrm_control")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">622</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">527</td>
+                    <td class="coverage">64<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    object$formula_parts, newdata,</pre>
+                      <pre class="language-r">    assert_list(control$optimizers, min.len = 1)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">623</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">528</td>
+                    <td class="coverage">64<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    weights = rep(1, nrow(newdata)),</pre>
+                      <pre class="language-r">    assert_numeric(weights, any.missing = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">624</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    reml = TRUE,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">625</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    singular = "keep",</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">626</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    drop_visit_levels = FALSE,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">627</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">529</td>
+                    <td class="coverage">64<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    allow_na_response = TRUE,</pre>
+                      <pre class="language-r">    assert_true(all(weights &gt; 0))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">628</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">530</td>
+                    <td class="coverage">64<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    drop_levels = FALSE,</pre>
+                      <pre class="language-r">    tmb_data &lt;- h_mmrm_tmb_data(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">629</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">531</td>
+                    <td class="coverage">64<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    xlev = component(object, "xlev"),</pre>
+                      <pre class="language-r">      formula_parts, data, weights, reml,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">630</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">532</td>
+                    <td class="coverage">64<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    contrasts = component(object, "contrasts")</pre>
+                      <pre class="language-r">      singular = if (control$accept_singular) "drop" else "error", drop_visit_levels = control$drop_visit_levels</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">631</td>
+                    <td class="num">533</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">632</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  ret &lt;- if (method == "conditional") {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">633</td>
-                    <td class="coverage">8<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    predict_res &lt;- h_get_prediction(tmb_data, object$theta_est, object$beta_est, object$beta_vcov)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">634</td>
-                    <td class="coverage">8<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    as.data.frame(h_get_sim_per_subj(predict_res, tmb_data$n_subjects, nsim))</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">635</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  } else if (method == "marginal") {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">636</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    theta_chol &lt;- t(chol(object$theta_vcov))</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">637</td>
-                    <td class="coverage">7<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    n_theta &lt;- length(object$theta_est)</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">638</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">534</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    as.data.frame(</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">639</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">535</td>
+                    <td class="coverage">185<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      sapply(seq_len(nsim), function(x) {</pre>
+                      <pre class="language-r">    assert_class(tmb_data, "mmrm_tmb_data")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">640</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">536</td>
+                    <td class="coverage">185<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        newtheta &lt;- object$theta_est + theta_chol %*% matrix(stats::rnorm(n_theta), ncol = 1)</pre>
+                      <pre class="language-r">    assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">641</td>
+                    <td class="num">537</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        # Recalculate betas with sampled thetas.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">642</td>
-                    <td class="coverage">503<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">        hold &lt;- object$tmb_object$report(newtheta)</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">643</td>
-                    <td class="coverage"></td>
+                    <td class="num">538</td>
+                    <td class="coverage">249<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        # Resample betas given new beta distribution.</pre>
+                      <pre class="language-r">  tmb_parameters &lt;- h_mmrm_tmb_parameters(formula_parts, tmb_data, start = control$start, n_groups = tmb_data$n_groups)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">644</td>
+                    <td class="num">539</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        # We first solve L^\top w = D^{-1/2}z_{sample}:</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">645</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">540</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        w_sample &lt;- backsolve(</pre>
+                      <pre class="language-r">  tmb_object &lt;- TMB::MakeADFun(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">646</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">541</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          r = hold$XtWX_L,</pre>
+                      <pre class="language-r">    data = tmb_data,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">647</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">542</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          x = stats::rnorm(length(hold$beta)) / sqrt(hold$XtWX_D),</pre>
+                      <pre class="language-r">    parameters = tmb_parameters,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">648</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">543</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          upper.tri = FALSE,</pre>
+                      <pre class="language-r">    hessian = TRUE,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">649</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">544</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          transpose = TRUE</pre>
+                      <pre class="language-r">    DLL = "mmrm",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">650</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">545</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        )</pre>
+                      <pre class="language-r">    silent = TRUE</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">651</td>
+                    <td class="num">546</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        # Then we add the mean vector, the beta estimate.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">652</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">547</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        beta_sample &lt;- hold$beta + w_sample</pre>
+                      <pre class="language-r">  h_mmrm_tmb_assert_start(tmb_object)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">653</td>
-                    <td class="coverage">503<em>x</em></td>
+                    <td class="num">548</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        predict_res &lt;- h_get_prediction(tmb_data, newtheta, beta_sample, hold$beta_vcov)</pre>
+                      <pre class="language-r">  used_optimizer &lt;- control$optimizers[[1L]]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">654</td>
-                    <td class="coverage">503<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">        h_get_sim_per_subj(predict_res, tmb_data$n_subjects, 1L)</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">655</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      })</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">656</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">657</td>
-                    <td class="coverage"></td>
+                    <td class="num">549</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  used_optimizer_name &lt;- names(control$optimizers)[1L]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">658</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">550</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  orig_row_names &lt;- row.names(newdata)</pre>
+                      <pre class="language-r">  args &lt;- with(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">659</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">551</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  new_order &lt;- match(orig_row_names, row.names(tmb_data$full_frame))</pre>
+                      <pre class="language-r">    tmb_object,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">660</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  ret[new_order, , drop = FALSE]</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">661</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">662</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r"></pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">663</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' Get simulated values by patient.</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">664</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">665</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @param predict_res (`list`)\cr from [h_get_prediction()].</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">666</td>
-                    <td class="coverage"></td>
+                    <td class="num">552</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param nsub (`count`)\cr number of subjects.</pre>
+                      <pre class="language-r">    c(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">667</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">553</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param nsim (`count`)\cr number of values to simulate.</pre>
+                      <pre class="language-r">      list(par, fn, gr),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">668</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">554</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      attr(used_optimizer, "args")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">669</td>
+                    <td class="num">555</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">670</td>
+                    <td class="num">556</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_sim_per_subj &lt;- function(predict_res, nsub, nsim) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">671</td>
-                    <td class="coverage">517<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(predict_res)</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">672</td>
-                    <td class="coverage">517<em>x</em></td>
+                    <td class="num">557</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(nsub, positive = TRUE)</pre>
+                      <pre class="language-r">  if (identical(attr(used_optimizer, "use_hessian"), TRUE)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">673</td>
-                    <td class="coverage">516<em>x</em></td>
+                    <td class="num">558</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
+                      <pre class="language-r">    args$hessian &lt;- tmb_object$he</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">674</td>
+                    <td class="num">559</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">675</td>
-                    <td class="coverage">515<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  ret &lt;- matrix(</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">676</td>
-                    <td class="coverage">515<em>x</em></td>
+                    <td class="num">560</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    predict_res$prediction[, "fit"],</pre>
+                      <pre class="language-r">  tmb_opt &lt;- do.call(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">677</td>
-                    <td class="coverage">515<em>x</em></td>
+                    <td class="num">561</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ncol = nsim,</pre>
+                      <pre class="language-r">    what = used_optimizer,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">678</td>
-                    <td class="coverage">515<em>x</em></td>
+                    <td class="num">562</td>
+                    <td class="coverage">246<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    nrow = nrow(predict_res$prediction)</pre>
+                      <pre class="language-r">    args = args</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">679</td>
+                    <td class="num">563</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">680</td>
-                    <td class="coverage">515<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  for (i in seq_len(nsub)) {</pre>
-                    </td>
-                  </tr>
                   <tr class="never">
-                    <td class="num">681</td>
+                    <td class="num">564</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Skip subjects which are not included in predict_res.</pre>
+                      <pre class="language-r">  # Ensure negative log likelihood is stored in `objective` element of list.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">682</td>
-                    <td class="coverage">82699<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    if (length(predict_res$index[[i]]) &gt; 0) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">683</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      # Obtain indices of data.frame belonging to subject i</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">684</td>
-                    <td class="coverage"></td>
+                    <td class="num">565</td>
+                    <td class="coverage">237<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # (increment by 1, since indices from cpp are 0-order).</pre>
+                      <pre class="language-r">  if ("value" %in% names(tmb_opt)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">685</td>
-                    <td class="coverage">66631<em>x</em></td>
+                    <td class="num">566</td>
+                    <td class="coverage">227<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      inds &lt;- predict_res$index[[i]] + 1</pre>
+                      <pre class="language-r">    tmb_opt$objective &lt;- tmb_opt$value</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">686</td>
-                    <td class="coverage">66631<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      obs &lt;- length(inds)</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">687</td>
-                    <td class="coverage"></td>
+                    <td class="num">567</td>
+                    <td class="coverage">227<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    tmb_opt$value &lt;- NULL</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">688</td>
+                    <td class="num">568</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # Get relevant covariance matrix for subject i.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">689</td>
-                    <td class="coverage">66631<em>x</em></td>
+                    <td class="num">569</td>
+                    <td class="coverage">237<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      covmat_i &lt;- predict_res$covariance[[i]]</pre>
+                      <pre class="language-r">  fit &lt;- h_mmrm_tmb_fit(tmb_object, tmb_opt, formula_parts, tmb_data)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">690</td>
-                    <td class="coverage">66631<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      theta_chol &lt;- t(chol(covmat_i))</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">691</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r"></pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">692</td>
-                    <td class="coverage"></td>
+                    <td class="num">570</td>
+                    <td class="coverage">237<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # Simulate epsilon from covariance matrix.</pre>
+                      <pre class="language-r">  h_mmrm_tmb_check_conv(tmb_opt, fit)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">693</td>
-                    <td class="coverage">66631<em>x</em></td>
+                    <td class="num">571</td>
+                    <td class="coverage">237<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      mus &lt;- ret[inds, , drop = FALSE]</pre>
+                      <pre class="language-r">  fit$call &lt;- match.call()</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">694</td>
-                    <td class="coverage">66631<em>x</em></td>
+                    <td class="num">572</td>
+                    <td class="coverage">237<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      epsilons &lt;- theta_chol %*% matrix(stats::rnorm(nsim * obs), ncol = nsim)</pre>
+                      <pre class="language-r">  fit$call$formula &lt;- formula_parts$formula</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">695</td>
-                    <td class="coverage">66631<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      ret[inds, ] &lt;- mus + epsilons</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">696</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">697</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">698</td>
-                    <td class="coverage"></td>
+                    <td class="num">573</td>
+                    <td class="coverage">237<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  fit$optimizer &lt;- used_optimizer_name</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">699</td>
-                    <td class="coverage">515<em>x</em></td>
+                    <td class="num">574</td>
+                    <td class="coverage">237<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret</pre>
+                      <pre class="language-r">  fit</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">700</td>
+                    <td class="num">575</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
@@ -13163,14 +12820,14 @@ <h2>mmrm coverage - 97.05%</h2>
                 </tbody>
               </table>
             </div>
-            <div id="R/tmb.R" class="hidden">
+            <div id="R/tidiers.R" class="hidden">
               <table class="table-condensed">
                 <tbody>
                   <tr class="never">
                     <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Processing the Formula for `TMB` Fit</pre>
+                      <pre class="language-r">#' Tidying Methods for `mmrm` Objects</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -13184,133 +12841,133 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula (`formula`)\cr Original formula.</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param covariance (`cov_struct`)\cr A covariance structure from which</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   additional formula parts should be added.</pre>
+                      <pre class="language-r">#' These methods tidy the estimates from an `mmrm` object into a</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' summary.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List of class `mmrm_tmb_formula_parts` with elements:</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param x (`mmrm`)\cr fitted model.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `formula`: the original input.</pre>
+                      <pre class="language-r">#' @param conf.int (`flag`)\cr if `TRUE` columns for the lower (`conf.low`) and upper bounds</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `model_formula`: `formula` with the covariance term is removed.</pre>
+                      <pre class="language-r">#'   (`conf.high`) of coefficient estimates are included.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `model_formula`: `formula` with the covariance term removed.</pre>
+                      <pre class="language-r">#' @param conf.level (`number`)\cr defines the range of the optional confidence internal.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">12</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `full_formula`: same as `model_formula` but includes the covariance</pre>
+                      <pre class="language-r">#' @param newdata (`data.frame` or `NULL`)\cr optional new data frame.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   structure's subject, visit and (optionally) group variables.</pre>
+                      <pre class="language-r">#' @param se_fit (`flag`)\cr whether to return standard errors of fit.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `cov_type`: `string` with covariance term type (e.g. `"us"`).</pre>
+                      <pre class="language-r">#' @param interval (`string`)\cr type of interval calculation.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `is_spatial`: `flag` indicator of whether the covariance structure is</pre>
+                      <pre class="language-r">#' @param type.residuals (`string`)\cr passed on to [residuals.mmrm_tmb()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   spatial</pre>
+                      <pre class="language-r">#' @param ... only used by `augment()` to pass arguments to the [predict.mmrm_tmb()] method.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">17</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `visit_var`: `character` with the visit variable name.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">18</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `subject_var`: `string` with the subject variable name.</pre>
+                      <pre class="language-r">#' @name mmrm_tidiers</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `group_var`: `string` with the group variable name. If no group specified,</pre>
+                      <pre class="language-r">#' @aliases mmrm_tidiers</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   this element is `NULL`.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `model_var`: `character` with the variables names of the formula, except `subject_var`.</pre>
+                      <pre class="language-r">#' @seealso [`mmrm_methods`], [`mmrm_tmb_methods`] for additional methods.</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -13324,42 +12981,42 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_mmrm_tmb_formula_parts &lt;- function(</pre>
+                      <pre class="language-r">#' fit &lt;- mmrm(</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">25</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    formula,</pre>
+                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    covariance = as.cov_struct(formula, warn_partial = FALSE)) {</pre>
+                      <pre class="language-r">#'   data = fev_data</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">27</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_formula(formula)</pre>
+                      <pre class="language-r">#' )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">28</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_true(identical(length(formula), 3L))</pre>
+                      <pre class="language-r">NULL</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -13369,4837 +13026,4688 @@ <h2>mmrm coverage - 97.05%</h2>
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">30</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  model_formula &lt;- h_drop_covariance_terms(formula)</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tidiers derives tidy `tibble` from an `mmrm` object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">31</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">32</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  structure(</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">33</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    list(</pre>
+                      <pre class="language-r">#' # Applying tidy method to return summary table of covariate estimates.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">34</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      formula = formula,</pre>
+                      <pre class="language-r">#' fit |&gt; tidy()</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">35</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      model_formula = model_formula,</pre>
+                      <pre class="language-r">#' fit |&gt; tidy(conf.int = TRUE, conf.level = 0.9)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">36</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      full_formula = h_add_covariance_terms(model_formula, covariance),</pre>
+                      <pre class="language-r">tidy.mmrm &lt;- function(x, # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">37</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      cov_type = tmb_cov_type(covariance),</pre>
+                      <pre class="language-r">                      conf.int = FALSE, # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">38</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      is_spatial = covariance$type == "sp_exp",</pre>
+                      <pre class="language-r">                      conf.level = 0.95, # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">39</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      visit_var = covariance$visits,</pre>
+                      <pre class="language-r">                      ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">40</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      subject_var = covariance$subject,</pre>
+                      <pre class="language-r">  assert_flag(conf.int)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">41</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      group_var = if (length(covariance$group) &lt; 1) NULL else covariance$group,</pre>
+                      <pre class="language-r">  assert_number(conf.level, lower = 0, upper = 1)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">42</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      model_var = setdiff(all.vars(formula[[3]]), covariance$subject)</pre>
+                      <pre class="language-r">  tbl &lt;- tibble::as_tibble(summary(x)$coefficients, rownames = "term")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">43</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ),</pre>
+                      <pre class="language-r">  colnames(tbl) &lt;- c("term", "estimate", "std.error", "df", "statistic", "p.value")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">44</td>
-                    <td class="coverage">270<em>x</em></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    class = "mmrm_tmb_formula_parts"</pre>
+                      <pre class="language-r">  coefs &lt;- coef(x)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">45</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  if (length(coefs) != nrow(tbl)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="missed">
                     <td class="num">46</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    coefs &lt;- tibble::enframe(coefs, name = "term", value = "estimate")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="missed">
                     <td class="num">47</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    tbl &lt;- merge(coefs, tbl, by = c("term", "estimate"))</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">48</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Data for `TMB` Fit</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">49</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  if (conf.int) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">50</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula_parts (`mmrm_tmb_formula_parts`)\cr list with formula parts</pre>
+                      <pre class="language-r">    ci &lt;- h_tbl_confint_terms(x, level = conf.level)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">51</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   from [h_mmrm_tmb_formula_parts()].</pre>
+                      <pre class="language-r">    tbl &lt;- tibble::as_tibble(merge(tbl, ci, by = "term"))</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param data (`data.frame`)\cr which contains variables used in `formula_parts`.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">53</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param weights (`vector`)\cr weights to be used in the fitting process.</pre>
+                      <pre class="language-r">  tbl</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">54</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param reml (`flag`)\cr whether restricted maximum likelihood (REML) estimation is used,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">55</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   otherwise maximum likelihood (ML) is used.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">56</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param singular (`string`)\cr choices of method deal with rank-deficient matrices. "error" to</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tidiers derives `glance` `tibble` from an `mmrm` object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">57</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   stop the function return the error, "drop" to drop these columns, and "keep" to keep all the columns.</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">58</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param drop_visit_levels (`flag`)\cr whether to drop levels for visit variable, if visit variable is a factor.</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">59</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param allow_na_response (`flag`)\cr whether NA in response is allowed.</pre>
+                      <pre class="language-r">#' # Applying glance method to return summary table of goodness of fit statistics.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">60</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param drop_levels (`flag`)\cr whether drop levels for covariates. If not dropped could lead to singular matrix.</pre>
+                      <pre class="language-r">#' fit |&gt; glance()</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">61</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">glance.mmrm &lt;- function(x, ...) { # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">62</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List of class `mmrm_tmb_data` with elements:</pre>
+                      <pre class="language-r">  tibble::as_tibble(summary(x)$aic_list)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">63</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `full_frame`: `data.frame` with `n` rows containing all variables needed in the model.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">64</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `data`: `data.frame` of input dataset.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">65</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `x_matrix`: `matrix` with `n` rows and `p` columns specifying the overall design matrix.</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tidiers derives `augment` `tibble` from an `mmrm` object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">66</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `x_cols_aliased`: `logical` with potentially more than `p` elements indicating which</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">67</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      columns in the original design matrix have been left out to obtain a full rank</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">68</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      `x_matrix`.</pre>
+                      <pre class="language-r">#' # Applying augment method to return merged `tibble` of model data, fitted and residuals.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">69</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `y_vector`: length `n` `numeric` specifying the overall response vector.</pre>
+                      <pre class="language-r">#' fit |&gt; augment()</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `weights_vector`: length `n` `numeric` specifying the weights vector.</pre>
+                      <pre class="language-r">#' fit |&gt; augment(interval = "confidence")</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">71</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `n_visits`: `int` with the number of visits, which is the dimension of the</pre>
+                      <pre class="language-r">#' fit |&gt; augment(type.residuals = "pearson")</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">72</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      covariance matrix.</pre>
+                      <pre class="language-r">augment.mmrm &lt;- function(x, # nolint</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `n_subjects`: `int` with the number of subjects.</pre>
+                      <pre class="language-r">                         newdata = NULL,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">74</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `subject_zero_inds`: length `n_subjects` `integer` containing the zero-based start</pre>
+                      <pre class="language-r">                         interval = c("none", "confidence", "prediction"),</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">75</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     indices for each subject.</pre>
+                      <pre class="language-r">                         se_fit = (interval != "none"),</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">76</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `subject_n_visits`: length `n_subjects` `integer` containing the number of</pre>
+                      <pre class="language-r">                         type.residuals = c("response", "pearson", "normalized"), # nolint</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">77</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     observed visits for each subjects. So the sum of this vector equals `n`.</pre>
+                      <pre class="language-r">                         ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">78</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `cov_type`: `string` value specifying the covariance type.</pre>
+                      <pre class="language-r">  type.residuals &lt;- match.arg(type.residuals) # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">79</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `is_spatial_int`: `int` specifying whether the covariance structure is spatial(1) or not(0).</pre>
+                      <pre class="language-r">  resid_df &lt;- NULL</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">80</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `reml`: `int` specifying whether REML estimation is used (1), otherwise ML (0).</pre>
+                      <pre class="language-r">  if (is.null(newdata)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">81</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `subject_groups`: `factor` specifying the grouping for each subject.</pre>
+                      <pre class="language-r">    newdata &lt;- stats::get_all_vars(x, data = stats::na.omit(x$data))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">82</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `n_groups`: `int` with the number of total groups</pre>
+                      <pre class="language-r">    resid_df &lt;- data.frame(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">83</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      .rownames = rownames(newdata),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">84</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details Note that the `subject_var` must not be factor but can also be character.</pre>
+                      <pre class="language-r">      .resid = unname(residuals(x, type = type.residuals))</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   If it is character, then it will be converted to factor internally. Here</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">86</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   the levels will be the unique values, sorted alphabetically and numerically if there</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">87</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   is a common string prefix of numbers in the character elements. For full control</pre>
+                      <pre class="language-r">  interval &lt;- match.arg(interval)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   on the order please use a factor.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">89</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  tbl &lt;- h_newdata_add_pred(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">90</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    x,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">91</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_mmrm_tmb_data &lt;- function(formula_parts,</pre>
+                      <pre class="language-r">    newdata = newdata,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">92</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            data,</pre>
+                      <pre class="language-r">    se_fit = se_fit,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">93</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            weights,</pre>
+                      <pre class="language-r">    interval = interval,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">94</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            reml,</pre>
+                      <pre class="language-r">    ...</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">95</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            singular = c("drop", "error", "keep"),</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">96</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            drop_visit_levels,</pre>
+                      <pre class="language-r">  if (!is.null(resid_df)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">97</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            allow_na_response = FALSE,</pre>
+                      <pre class="language-r">    tbl &lt;- merge(tbl, resid_df, by = ".rownames")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">98</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            drop_levels = TRUE,</pre>
+                      <pre class="language-r">    tbl$.rownames &lt;- as.numeric(tbl$.rownames)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">99</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            xlev = NULL,</pre>
+                      <pre class="language-r">    tbl &lt;- tbl[order(tbl$.rownames), , drop = FALSE]</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">100</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                            contrasts = NULL) {</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">101</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
+                      <pre class="language-r">  tibble::as_tibble(tbl)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">102</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_data_frame(data)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">103</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  varname &lt;- formula_parts[grepl("_var", names(formula_parts))]</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">104</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_names(</pre>
+                      <pre class="language-r">#' Extract `tibble` with Confidence Intervals and Term Names</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">105</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    names(data),</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">106</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    must.include = unlist(varname, use.names = FALSE)</pre>
+                      <pre class="language-r">#' This is used in [tidy.mmrm()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">107</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">108</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_true(is.factor(data[[formula_parts$subject_var]]) || is.character(data[[formula_parts$subject_var]]))</pre>
+                      <pre class="language-r">#' @param x (`mmrm`)\cr fit object.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">109</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(weights, len = nrow(data))</pre>
+                      <pre class="language-r">#' @param ... passed to [stats::confint()], hence not used at the moment.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">110</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(reml)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">111</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  singular &lt;- match.arg(singular)</pre>
+                      <pre class="language-r">#' @return A `tibble` with `term`, `conf.low`, `conf.high` columns.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">112</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(drop_visit_levels)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">113</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">114</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.character(data[[formula_parts$subject_var]])) {</pre>
+                      <pre class="language-r">h_tbl_confint_terms &lt;- function(x, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">115</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    data[[formula_parts$subject_var]] &lt;- factor(</pre>
+                      <pre class="language-r">  df &lt;- stats::confint(x, ...)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">116</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      data[[formula_parts$subject_var]],</pre>
+                      <pre class="language-r">  tbl &lt;- tibble::as_tibble(df, rownames = "term", .name_repair = "minimal")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">117</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      levels = stringr::str_sort(unique(data[[formula_parts$subject_var]]), numeric = TRUE)</pre>
+                      <pre class="language-r">  names(tbl) &lt;- c("term", "conf.low", "conf.high")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">118</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">  tbl</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">119</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">120</td>
-                    <td class="coverage">312<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  data_order &lt;- if (formula_parts$is_spatial) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">121</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    order(data[[formula_parts$subject_var]])</pre>
+                      <pre class="language-r">#' Add Prediction Results to New Data</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">122</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">123</td>
-                    <td class="coverage">296<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    subject_visit_data &lt;- data[, c(formula_parts$subject_var, formula_parts$visit_var)]</pre>
+                      <pre class="language-r">#' This is used in [augment.mmrm()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">124</td>
-                    <td class="coverage">296<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    is_duplicated &lt;- duplicated(subject_visit_data)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">125</td>
-                    <td class="coverage">296<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (any(is_duplicated)) {</pre>
+                      <pre class="language-r">#' @param x (`mmrm`)\cr fit.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">126</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      stop(</pre>
+                      <pre class="language-r">#' @param newdata (`data.frame`)\cr data to predict.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">127</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "time points have to be unique for each subject, detected following duplicates in data:\n",</pre>
+                      <pre class="language-r">#' @param se_fit (`flag`)\cr whether to return standard error of prediction,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">128</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        paste(utils::capture.output(print(subject_visit_data[is_duplicated, ])), collapse = "\n")</pre>
+                      <pre class="language-r">#'   can only be used when `interval` is not "none".</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">129</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">#' @param interval (`string`)\cr type of interval.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">130</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#' @param ... passed to [predict.mmrm_tmb()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">131</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    order(data[[formula_parts$subject_var]], data[[formula_parts$visit_var]])</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">132</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @return The `newdata` as a `tibble` with additional columns `.fitted`,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">133</td>
-                    <td class="coverage">311<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (identical(formula_parts$is_spatial, FALSE)) {</pre>
+                      <pre class="language-r">#'   `.lower`, `.upper` (if interval is not `none`) and `.se.fit` (if `se_fit`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">134</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_confirm_large_levels(length(levels(data[[formula_parts$visit_var]])))</pre>
+                      <pre class="language-r">#'   requested).</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">136</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  data &lt;- data[data_order, ]</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">137</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  weights &lt;- weights[data_order]</pre>
+                      <pre class="language-r">h_newdata_add_pred &lt;- function(x,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">138</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  data &lt;- data.frame(data, weights)</pre>
+                      <pre class="language-r">                               newdata,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">139</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Weights is always the last column.</pre>
+                      <pre class="language-r">                               se_fit,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">140</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  weights_name &lt;- colnames(data)[ncol(data)]</pre>
+                      <pre class="language-r">                               interval,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">141</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # If `y` is allowed to be NA, then first replace y with 1:n, then replace it with original y.</pre>
+                      <pre class="language-r">                               ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">142</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!allow_na_response) {</pre>
+                      <pre class="language-r">  assert_class(x, "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">143</td>
-                    <td class="coverage">260<em>x</em></td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_warn_na_action()</pre>
+                      <pre class="language-r">  assert_data_frame(newdata)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">144</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  assert_flag(se_fit)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">145</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  full_frame &lt;- eval(</pre>
+                      <pre class="language-r">  assert_string(interval)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">146</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    bquote(stats::model.frame(</pre>
+                      <pre class="language-r">  if (interval == "none") {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">147</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      formula_parts$full_formula,</pre>
+                      <pre class="language-r">    assert_false(se_fit)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">148</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      data = data,</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">149</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      weights = .(as.symbol(weights_name)),</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">150</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      na.action = "na.pass",</pre>
+                      <pre class="language-r">  tbl &lt;- h_df_to_tibble(newdata)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">151</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      xlev = xlev</pre>
+                      <pre class="language-r">  pred_results &lt;- predict(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">152</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ))</pre>
+                      <pre class="language-r">    x,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">153</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    newdata = newdata,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">154</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (drop_levels) {</pre>
+                      <pre class="language-r">    na.action = stats::na.pass,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">155</td>
-                    <td class="coverage">262<em>x</em></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    full_frame &lt;- h_drop_levels(full_frame, formula_parts$subject_var, formula_parts$visit_var, names(xlev))</pre>
+                      <pre class="language-r">    se.fit = se_fit,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">156</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    interval = interval,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">157</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  has_response &lt;- !identical(attr(attr(full_frame, "terms"), "response"), 0L)</pre>
+                      <pre class="language-r">    ...</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">158</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  keep_ind &lt;- if (allow_na_response &amp;&amp; has_response) {</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">159</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Note that response is always the first column if there is response.</pre>
+                      <pre class="language-r">  if (interval == "none") {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">160</td>
-                    <td class="coverage">50<em>x</em></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stats::complete.cases(full_frame[, -1L, drop = FALSE])</pre>
+                      <pre class="language-r">    assert_numeric(pred_results)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">161</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">    tbl$.fitted &lt;- unname(pred_results)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">162</td>
-                    <td class="coverage">260<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stats::complete.cases(full_frame)</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">163</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    assert_matrix(pred_results)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">164</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  full_frame &lt;- full_frame[keep_ind, ]</pre>
+                      <pre class="language-r">    tbl$.fitted &lt;- unname(pred_results[, "fit"])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">165</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (drop_visit_levels &amp;&amp; !formula_parts$is_spatial &amp;&amp; h_extra_levels(full_frame[[formula_parts$visit_var]])) {</pre>
+                      <pre class="language-r">    tbl$.lower &lt;- unname(pred_results[, "lwr"])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">166</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    visit_vec &lt;- full_frame[[formula_parts$visit_var]]</pre>
+                      <pre class="language-r">    tbl$.upper &lt;- unname(pred_results[, "upr"])</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">167</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    old_levels &lt;- levels(visit_vec)</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">168</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    full_frame[[formula_parts$visit_var]] &lt;- droplevels(visit_vec)</pre>
+                      <pre class="language-r">  if (se_fit) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">169</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    new_levels &lt;- levels(full_frame[[formula_parts$visit_var]])</pre>
+                      <pre class="language-r">    tbl$.se.fit &lt;- unname(pred_results[, "se"])</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">170</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    dropped &lt;- setdiff(old_levels, new_levels)</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">171</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    message(</pre>
+                      <pre class="language-r">  tbl</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">172</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "In ", formula_parts$visit_var, " there are dropped visits: ", toString(dropped),</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">173</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ".\n Additional attributes including contrasts are lost.\n",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">174</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "To avoid this behavior, make sure use `drop_visit_levels = FALSE`."</pre>
+                      <pre class="language-r">#' Coerce a Data Frame to a `tibble`</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">175</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">176</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' This is used in [h_newdata_add_pred()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">177</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  is_factor_col &lt;- vapply(full_frame, is.factor, FUN.VALUE = TRUE)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">178</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  is_factor_col &lt;- intersect(names(is_factor_col)[is_factor_col], all.vars(formula_parts$model_formula))</pre>
+                      <pre class="language-r">#' @details This is only a thin wrapper around [tibble::as_tibble()], except</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">179</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x_matrix &lt;- stats::model.matrix(</pre>
+                      <pre class="language-r">#' giving a useful error message and it checks for `rownames` and adds them</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">180</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    formula_parts$model_formula,</pre>
+                      <pre class="language-r">#' as a new column `.rownames` if they are not just a numeric sequence as</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">181</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    data = full_frame,</pre>
+                      <pre class="language-r">#' per the [tibble::has_rownames()] decision.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">182</td>
-                    <td class="coverage">310<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    contrasts.arg = h_default_value(contrasts, lapply(full_frame[is_factor_col], contrasts))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">183</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' @param data (`data.frame`)\cr what to coerce.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">184</td>
-                    <td class="coverage">309<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x_cols_aliased &lt;- stats::setNames(rep(FALSE, ncol(x_matrix)), nm = colnames(x_matrix))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">185</td>
-                    <td class="coverage">309<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  qr_x_mat &lt;- qr(x_matrix)</pre>
+                      <pre class="language-r">#' @return The `data` as a `tibble`, potentially with a `.rownames` column.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">186</td>
-                    <td class="coverage">309<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (qr_x_mat$rank &lt; ncol(x_matrix)) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">187</td>
-                    <td class="coverage">23<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cols_to_drop &lt;- utils::tail(qr_x_mat$pivot, ncol(x_matrix) - qr_x_mat$rank)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">188</td>
-                    <td class="coverage">23<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (identical(singular, "error")) {</pre>
+                      <pre class="language-r">h_df_to_tibble &lt;- function(data) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">189</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      stop(</pre>
+                      <pre class="language-r">  tryCatch(tbl &lt;- tibble::as_tibble(data), error = function(cnd) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">190</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "design matrix only has rank ", qr_x_mat$rank, " and ", length(cols_to_drop),</pre>
+                      <pre class="language-r">    stop("Could not coerce data to `tibble`. Try explicitly passing a",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">191</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        " columns (", toString(colnames(x_matrix)[cols_to_drop]), ") could be dropped",</pre>
+                      <pre class="language-r">      "dataset to either the `data` or `newdata` argument.",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">192</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        " to achieve full rank ", ncol(x_matrix), " by using `accept_singular = TRUE`"</pre>
+                      <pre class="language-r">      call. = FALSE</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">193</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">194</td>
-                    <td class="coverage">22<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    } else if (identical(singular, "drop")) {</pre>
+                      <pre class="language-r">  })</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">195</td>
-                    <td class="coverage">11<em>x</em></td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      assign_attr &lt;- attr(x_matrix, "assign")</pre>
+                      <pre class="language-r">  if (tibble::has_rownames(data)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">196</td>
-                    <td class="coverage">11<em>x</em></td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      contrasts_attr &lt;- attr(x_matrix, "contrasts")</pre>
+                      <pre class="language-r">    tbl &lt;- tibble::add_column(tbl, .rownames = rownames(data), .before = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">197</td>
-                    <td class="coverage">11<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      x_matrix &lt;- x_matrix[, -cols_to_drop, drop = FALSE]</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">198</td>
-                    <td class="coverage">11<em>x</em></td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      x_cols_aliased[cols_to_drop] &lt;- TRUE</pre>
+                      <pre class="language-r">  tbl</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">199</td>
-                    <td class="coverage">11<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      attr(x_matrix, "assign") &lt;- assign_attr[-cols_to_drop]</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">200</td>
-                    <td class="coverage">11<em>x</em></td>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/interop-emmeans.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
+                  <tr class="never">
+                    <td class="num">1</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      attr(x_matrix, "contrasts") &lt;- contrasts_attr</pre>
+                      <pre class="language-r">#' Support for `emmeans`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">201</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">202</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">203</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">4</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  y_vector &lt;- if (has_response) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">204</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">5</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    as.numeric(stats::model.response(full_frame))</pre>
+                      <pre class="language-r">#' This package includes methods that allow `mmrm` objects to be used</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">205</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">#' with the `emmeans` package. `emmeans` computes estimated marginal means</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">206</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">7</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    rep(NA_real_, nrow(full_frame))</pre>
+                      <pre class="language-r">#' (also called least-square means) for the coefficients of the MMRM.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">207</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' We can also e.g. obtain differences between groups by applying</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">208</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">9</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  weights_vector &lt;- as.numeric(stats::model.weights(full_frame))</pre>
+                      <pre class="language-r">#' [`pairs()`][emmeans::pairs.emmGrid()] on the object returned</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">209</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">10</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_subjects &lt;- length(unique(full_frame[[formula_parts$subject_var]]))</pre>
+                      <pre class="language-r">#' by [emmeans::emmeans()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">210</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">11</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  subject_zero_inds &lt;- which(!duplicated(full_frame[[formula_parts$subject_var]])) - 1L</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">211</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">12</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  subject_n_visits &lt;- c(utils::tail(subject_zero_inds, -1L), nrow(full_frame)) - subject_zero_inds</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">212</td>
+                    <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # It is possible that `subject_var` is factor with more levels (and this does not affect fit)</pre>
+                      <pre class="language-r">#' fit &lt;- mmrm(</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">213</td>
+                    <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # so no check is needed for `subject_visits`.</pre>
+                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">214</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">15</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_true(all(subject_n_visits &gt; 0))</pre>
+                      <pre class="language-r">#'   data = fev_data</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">215</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">16</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(formula_parts$group_var)) {</pre>
+                      <pre class="language-r">#' )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">216</td>
-                    <td class="coverage">41<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">17</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_factor(data[[formula_parts$group_var]])</pre>
+                      <pre class="language-r">#' if (require(emmeans)) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">217</td>
-                    <td class="coverage">41<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">18</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    subject_groups &lt;- full_frame[[formula_parts$group_var]][subject_zero_inds + 1L]</pre>
+                      <pre class="language-r">#'   emmeans(fit, ~ ARMCD | AVISIT)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">218</td>
-                    <td class="coverage">41<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">19</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_groups &lt;- nlevels(subject_groups)</pre>
+                      <pre class="language-r">#'   pairs(emmeans(fit, ~ ARMCD | AVISIT), reverse = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">219</td>
+                    <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">#' }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">220</td>
-                    <td class="coverage">267<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">21</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    subject_groups &lt;- factor(rep(0L, n_subjects))</pre>
+                      <pre class="language-r">#' @name emmeans_support</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">221</td>
-                    <td class="coverage">267<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">22</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_groups &lt;- 1L</pre>
+                      <pre class="language-r">NULL</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">222</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">223</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">24</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coordinates &lt;- full_frame[, formula_parts$visit_var, drop = FALSE]</pre>
+                      <pre class="language-r">#' Returns a `data.frame` for `emmeans` Purposes</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">224</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">25</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (formula_parts$is_spatial) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">225</td>
-                    <td class="coverage">16<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">26</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    lapply(coordinates, assert_numeric)</pre>
+                      <pre class="language-r">#' @seealso See [emmeans::recover_data()] for background.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">226</td>
-                    <td class="coverage">16<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">27</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    coordinates_matrix &lt;- as.matrix(coordinates)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">227</td>
-                    <td class="coverage">16<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">28</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_visits &lt;- max(subject_n_visits)</pre>
+                      <pre class="language-r">#' @noRd</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">228</td>
+                    <td class="num">29</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">recover_data.mmrm &lt;- function(object, ...) { # nolint</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">229</td>
-                    <td class="coverage">292<em>x</em></td>
+                    <td class="num">30</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert(identical(ncol(coordinates), 1L))</pre>
+                      <pre class="language-r">  fun_call &lt;- stats::getCall(object)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">230</td>
-                    <td class="coverage">292<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">31</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_factor(coordinates[[1L]])</pre>
+                      <pre class="language-r">  # subject_var is excluded because it should not contain fixed effect.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">231</td>
-                    <td class="coverage">292<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">32</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    coordinates_matrix &lt;- as.matrix(as.integer(coordinates[[1L]]) - 1, ncol = 1)</pre>
+                      <pre class="language-r">  # visit_var is not excluded because emmeans can provide marginal mean</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">232</td>
-                    <td class="coverage">292<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">33</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_visits &lt;- nlevels(coordinates[[1L]])</pre>
+                      <pre class="language-r">  # by each visit if visit_var is not spatial.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">233</td>
-                    <td class="coverage">292<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    assert_true(all(subject_n_visits &lt;= n_visits))</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">234</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">235</td>
-                    <td class="coverage">308<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  structure(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">236</td>
-                    <td class="coverage">308<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    list(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">237</td>
-                    <td class="coverage">308<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      full_frame = full_frame,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">238</td>
-                    <td class="coverage">308<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      data = data,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">239</td>
-                    <td class="coverage">308<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      x_matrix = x_matrix,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">240</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">34</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      x_cols_aliased = x_cols_aliased,</pre>
+                      <pre class="language-r">  model_frame &lt;- stats::model.frame(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">241</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">35</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      coordinates = coordinates_matrix,</pre>
+                      <pre class="language-r">    object,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">242</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">36</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      y_vector = y_vector,</pre>
+                      <pre class="language-r">    include = c(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">243</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">37</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      weights_vector = weights_vector,</pre>
+                      <pre class="language-r">      if (!object$formula_parts$is_spatial) "visit_var" else NULL,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">244</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">38</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      n_visits = n_visits,</pre>
+                      <pre class="language-r">      "response_var", "group_var"</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">245</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">39</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      n_subjects = n_subjects,</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">246</td>
-                    <td class="coverage">308<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">40</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      subject_zero_inds = subject_zero_inds,</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">247</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">41</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      subject_n_visits = subject_n_visits,</pre>
+                      <pre class="language-r">  model_terms &lt;- stats::delete.response(stats::terms(model_frame))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">248</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">42</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      cov_type = formula_parts$cov_type,</pre>
+                      <pre class="language-r">  emmeans::recover_data(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">249</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">43</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      is_spatial_int = as.integer(formula_parts$is_spatial),</pre>
+                      <pre class="language-r">    fun_call,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">250</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">44</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      reml = as.integer(reml),</pre>
+                      <pre class="language-r">    trms = model_terms,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">251</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">45</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      subject_groups = subject_groups,</pre>
+                      <pre class="language-r">    na.action = "na.omit",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">252</td>
-                    <td class="coverage">308<em>x</em></td>
+                    <td class="num">46</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      n_groups = n_groups</pre>
+                      <pre class="language-r">    frame = model_frame,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">253</td>
+                    <td class="num">47</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ),</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">254</td>
-                    <td class="coverage">308<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    class = "mmrm_tmb_data"</pre>
+                      <pre class="language-r">    ...</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">255</td>
+                    <td class="num">48</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">256</td>
+                    <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">257</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">258</td>
+                    <td class="num">51</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Start Parameters for `TMB` Fit</pre>
+                      <pre class="language-r">#' Returns a List of Model Details for `emmeans` Purposes</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">259</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">260</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula_parts (`mmrm_tmb_formula_parts`)\cr produced by</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">261</td>
+                    <td class="num">53</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'  [h_mmrm_tmb_formula_parts()].</pre>
+                      <pre class="language-r">#' @seealso See [emmeans::emm_basis()] for background.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">262</td>
+                    <td class="num">54</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">263</td>
+                    <td class="num">55</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param start (`numeric` or `NULL`)\cr optional start values for variance</pre>
+                      <pre class="language-r">#' @noRd</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">264</td>
+                    <td class="num">56</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   parameters.</pre>
+                      <pre class="language-r">emm_basis.mmrm &lt;- function(object, # nolint</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">265</td>
+                    <td class="num">57</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_groups (`int`)\cr number of groups.</pre>
+                      <pre class="language-r">                           trms,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">266</td>
+                    <td class="num">58</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with element `theta` containing the start values for the variance</pre>
+                      <pre class="language-r">                           xlev,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">267</td>
+                    <td class="num">59</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   parameters.</pre>
+                      <pre class="language-r">                           grid,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">268</td>
+                    <td class="num">60</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                           ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">269</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">61</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  model_frame &lt;- stats::model.frame(trms, grid, na.action = stats::na.pass, xlev = xlev)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">270</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">62</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_mmrm_tmb_parameters &lt;- function(formula_parts,</pre>
+                      <pre class="language-r">  contrasts &lt;- component(object, "contrasts")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">271</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">63</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                  tmb_data,</pre>
+                      <pre class="language-r">  model_mat &lt;- stats::model.matrix(trms, model_frame, contrasts.arg = contrasts)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">272</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">64</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                  start,</pre>
+                      <pre class="language-r">  beta_hat &lt;- component(object, "beta_est")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">273</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">65</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                                  n_groups = 1L) {</pre>
+                      <pre class="language-r">  nbasis &lt;- if (length(beta_hat) &lt; ncol(model_mat)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">274</td>
-                    <td class="coverage">265<em>x</em></td>
+                    <td class="num">66</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
+                      <pre class="language-r">    kept &lt;- match(names(beta_hat), colnames(model_mat))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">275</td>
-                    <td class="coverage">265<em>x</em></td>
+                    <td class="num">67</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
+                      <pre class="language-r">    beta_hat &lt;- NA * model_mat[1L, ]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">276</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">68</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    beta_hat[kept] &lt;- component(object, "beta_est")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">277</td>
-                    <td class="coverage">265<em>x</em></td>
+                    <td class="num">69</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  m &lt;- tmb_data$n_visits</pre>
+                      <pre class="language-r">    orig_model_mat &lt;- stats::model.matrix(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">278</td>
-                    <td class="coverage">265<em>x</em></td>
+                    <td class="num">70</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  start_value0 &lt;- std_start(formula_parts$cov_type, m, n_groups)</pre>
+                      <pre class="language-r">      trms,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">279</td>
-                    <td class="coverage">265<em>x</em></td>
+                    <td class="num">71</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  theta_dim &lt;- length(start_value0)</pre>
+                      <pre class="language-r">      stats::model.frame(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">280</td>
-                    <td class="coverage">265<em>x</em></td>
+                    <td class="num">72</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  start_values &lt;- if (is.null(start)) {</pre>
+                      <pre class="language-r">        object,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">281</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">73</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    start_value0</pre>
+                      <pre class="language-r">        include = c(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">282</td>
-                    <td class="coverage">265<em>x</em></td>
+                    <td class="num">74</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else if (test_function(start)) {</pre>
+                      <pre class="language-r">          if (!object$formula_parts$is_spatial) "visit_var" else NULL,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">283</td>
-                    <td class="coverage">233<em>x</em></td>
+                    <td class="num">75</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    do.call(start, utils::modifyList(formula_parts, tmb_data))</pre>
+                      <pre class="language-r">          "response_var", "group_var"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">284</td>
+                    <td class="num">76</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">285</td>
-                    <td class="coverage">17<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    start</pre>
+                      <pre class="language-r">        )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">286</td>
+                    <td class="num">77</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">287</td>
-                    <td class="coverage">264<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(start_values, len = theta_dim, any.missing = FALSE, finite = TRUE)</pre>
+                      <pre class="language-r">      ),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">288</td>
-                    <td class="coverage">262<em>x</em></td>
+                    <td class="num">78</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(theta = start_values)</pre>
+                      <pre class="language-r">      contrasts.arg = contrasts</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">289</td>
+                    <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">290</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">80</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    estimability::nonest.basis(orig_model_mat)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">291</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Asserting Sane Start Values for `TMB` Fit</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">292</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">82</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    estimability::all.estble</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">293</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_object (`list`)\cr created with [TMB::MakeADFun()].</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">294</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">84</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  dfargs &lt;- list(object = object)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">295</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">85</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Nothing, only used for assertions.</pre>
+                      <pre class="language-r">  dffun &lt;- function(k, dfargs) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">296</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">86</td>
+                    <td class="coverage">113<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    mmrm::df_md(dfargs$object, contrast = k)$denom_df</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">297</td>
+                    <td class="num">87</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">298</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">88</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_mmrm_tmb_assert_start &lt;- function(tmb_object) {</pre>
+                      <pre class="language-r">  list(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">299</td>
-                    <td class="coverage">249<em>x</em></td>
+                    <td class="num">89</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(tmb_object)</pre>
+                      <pre class="language-r">    X = model_mat,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">300</td>
-                    <td class="coverage">249<em>x</em></td>
+                    <td class="num">90</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(c("fn", "gr", "par"), names(tmb_object))</pre>
+                      <pre class="language-r">    bhat = beta_hat,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">301</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">91</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    nbasis = nbasis,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">302</td>
-                    <td class="coverage">249<em>x</em></td>
+                    <td class="num">92</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.na(tmb_object$fn(tmb_object$par))) {</pre>
+                      <pre class="language-r">    V = component(object, "beta_vcov"),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">303</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">93</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("negative log-likelihood is NaN at starting parameter values")</pre>
+                      <pre class="language-r">    dffun = dffun,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">304</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">94</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    dfargs = dfargs</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">305</td>
-                    <td class="coverage">248<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">95</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (any(is.na(tmb_object$gr(tmb_object$par)))) {</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">306</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">96</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("some elements of gradient are NaN at starting parameter values")</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/kenwardroger.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">307</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' Obtain Kenward-Roger Adjustment Components</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">308</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">309</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @description Obtains the components needed downstream for the computation of Kenward-Roger degrees of freedom.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">310</td>
+                    <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Checking the `TMB` Optimization Result</pre>
+                      <pre class="language-r">#' Used in [mmrm()] fitting if method is "Kenward-Roger".</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">311</td>
+                    <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">312</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_opt (`list`)\cr optimization result.</pre>
+                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">313</td>
+                    <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param mmrm_tmb (`mmrm_tmb`)\cr result from [h_mmrm_tmb_fit()].</pre>
+                      <pre class="language-r">#' @param theta (`numeric`)\cr theta estimate.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">314</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">315</td>
+                    <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Nothing, only used to generate warnings in case that the model</pre>
+                      <pre class="language-r">#' @details the function returns a named list, \eqn{P}, \eqn{Q} and \eqn{R}, which corresponds to the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">316</td>
+                    <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' did not converge.</pre>
+                      <pre class="language-r">#' paper in 1997. The matrices are stacked in columns so that \eqn{P}, \eqn{Q} and \eqn{R} has the same</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">317</td>
+                    <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' column number(number of beta parameters). The number of rows, is dependent on</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">318</td>
+                    <td class="num">12</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' the total number of theta and number of groups, if the fit is a grouped mmrm.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">319</td>
+                    <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_mmrm_tmb_check_conv &lt;- function(tmb_opt, mmrm_tmb) {</pre>
+                      <pre class="language-r">#' For \eqn{P} matrix, it is stacked sequentially. For \eqn{Q} and \eqn{R} matrix, it is stacked so</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">320</td>
-                    <td class="coverage">245<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">14</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(tmb_opt)</pre>
+                      <pre class="language-r">#' that the \eqn{Q_{ij}} and \eqn{R_{ij}} is stacked from \eqn{j} then to \eqn{i}, i.e. \eqn{R_{i1}}, \eqn{R_{i2}}, etc.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">321</td>
-                    <td class="coverage">245<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">15</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(c("par", "objective", "convergence", "message"), names(tmb_opt))</pre>
+                      <pre class="language-r">#' \eqn{Q} and \eqn{R} only contains intra-group results and inter-group results should be all zero matrices</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">322</td>
-                    <td class="coverage">245<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">16</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(mmrm_tmb, "mmrm_tmb")</pre>
+                      <pre class="language-r">#' so they are not stacked in the result.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">323</td>
+                    <td class="num">17</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">324</td>
-                    <td class="coverage">245<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">18</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(tmb_opt$convergence) &amp;&amp; tmb_opt$convergence != 0) {</pre>
+                      <pre class="language-r">#' @return Named list with elements:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">325</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">19</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning("Model convergence problem: ", tmb_opt$message, ".")</pre>
+                      <pre class="language-r">#' - `P`: `matrix` of \eqn{P} component.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">326</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">20</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return()</pre>
+                      <pre class="language-r">#' - `Q`: `matrix` of \eqn{Q} component.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">327</td>
+                    <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' - `R`: `matrix` of \eqn{R} component.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">328</td>
-                    <td class="coverage">242<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">22</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  theta_vcov &lt;- mmrm_tmb$theta_vcov</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">329</td>
-                    <td class="coverage">242<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">23</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is(theta_vcov, "try-error")) {</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">330</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">24</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning("Model convergence problem: hessian is singular, theta_vcov not available.")</pre>
+                      <pre class="language-r">h_get_kr_comp &lt;- function(tmb_data, theta) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">331</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">25</td>
+                    <td class="coverage">47<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return()</pre>
+                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">332</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">26</td>
+                    <td class="coverage">47<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  assert_class(theta, "numeric")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">333</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">27</td>
+                    <td class="coverage">47<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!all(is.finite(theta_vcov))) {</pre>
+                      <pre class="language-r">  .Call(`_mmrm_get_pqr`, PACKAGE = "mmrm", tmb_data, theta)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">334</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">28</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning("Model convergence problem: theta_vcov contains non-finite values.")</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">335</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">29</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return()</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">336</td>
+                    <td class="num">30</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' Calculation of Kenward-Roger Degrees of Freedom for Multi-Dimensional Contrast</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">337</td>
-                    <td class="coverage">236<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">31</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  eigen_vals &lt;- eigen(theta_vcov, only.values = TRUE)$values</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">338</td>
-                    <td class="coverage">236<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">32</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (mode(eigen_vals) == "complex" || any(eigen_vals &lt;= 0)) {</pre>
+                      <pre class="language-r">#' @description Used in [df_md()] if method is "Kenward-Roger" or "Kenward-Roger-Linear".</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">339</td>
+                    <td class="num">33</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Note: complex eigen values signal that the matrix is not symmetric, therefore not positive definite.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">340</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">34</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning("Model convergence problem: theta_vcov is not positive definite.")</pre>
+                      <pre class="language-r">#' @inheritParams h_df_md_sat</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">341</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">35</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return()</pre>
+                      <pre class="language-r">#' @inherit h_df_md_sat return</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">342</td>
+                    <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">37</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">h_df_md_kr &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">343</td>
-                    <td class="coverage">233<em>x</em></td>
+                    <td class="num">38</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  qr_rank &lt;- qr(theta_vcov)$rank</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">344</td>
-                    <td class="coverage">233<em>x</em></td>
+                    <td class="num">39</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (qr_rank &lt; ncol(theta_vcov)) {</pre>
+                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">345</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">40</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning("Model convergence problem: theta_vcov is numerically singular.")</pre>
+                      <pre class="language-r">  if (component(object, "reml") != 1) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="missed">
+                    <td class="num">41</td>
+                    <td class="coverage">!</td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    stop("Kenward-Roger is only for REML")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">346</td>
+                    <td class="num">42</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">347</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">43</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  kr_comp &lt;- object$kr_comp</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">348</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">44</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  w &lt;- component(object, "theta_vcov")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">349</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">45</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Extract covariance matrix from `TMB` report and input data</pre>
+                      <pre class="language-r">  v_adj &lt;- object$beta_vcov_adj</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">350</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">46</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  df &lt;- h_kr_df(v0 = object$beta_vcov, l = contrast, w = w, p = kr_comp$P)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">351</td>
+                    <td class="num">47</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This helper does some simple post-processing to extract covariance matrix or named</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">352</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">48</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' list of covariance matrices if the fitting is using grouped covariance matrices.</pre>
+                      <pre class="language-r">  h_test_md(object, contrast, df = df$m, f_stat_factor = df$lambda)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">353</td>
+                    <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">354</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_report (`list`)\cr report created with [TMB::MakeADFun()] report function.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">355</td>
+                    <td class="num">51</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
+                      <pre class="language-r">#' Calculation of Kenward-Roger Degrees of Freedom for One-Dimensional Contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">356</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param visit_var (`character`)\cr character vector of the visit variable</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">357</td>
+                    <td class="num">53</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param is_spatial (`flag`)\cr indicator whether the covariance structure is spatial.</pre>
+                      <pre class="language-r">#' @description Used in [df_1d()] if method is</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">358</td>
+                    <td class="num">54</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Return a simple covariance matrix if there is no grouping, or a named</pre>
+                      <pre class="language-r">#' "Kenward-Roger" or "Kenward-Roger-Linear".</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">359</td>
+                    <td class="num">55</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' list of estimated grouped covariance matrices,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">360</td>
+                    <td class="num">56</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' with its name equal to the group levels.</pre>
+                      <pre class="language-r">#' @inheritParams h_df_1d_sat</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">361</td>
+                    <td class="num">57</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @inherit h_df_1d_sat return</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">362</td>
+                    <td class="num">58</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">363</td>
+                    <td class="num">59</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_mmrm_tmb_extract_cov &lt;- function(tmb_report, tmb_data, visit_var, is_spatial) {</pre>
+                      <pre class="language-r">h_df_1d_kr &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">364</td>
-                    <td class="coverage">241<em>x</em></td>
+                    <td class="num">60</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  d &lt;- dim(tmb_report$covariance_lower_chol)</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">365</td>
-                    <td class="coverage">241<em>x</em></td>
+                    <td class="num">61</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  visit_names &lt;- if (!is_spatial) {</pre>
+                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">366</td>
-                    <td class="coverage">228<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    levels(tmb_data$full_frame[[visit_var]])</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">367</td>
-                    <td class="coverage"></td>
+                    <td class="num">62</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">  if (component(object, "reml") != 1) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">368</td>
-                    <td class="coverage">13<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">63</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    c(0, 1)</pre>
+                      <pre class="language-r">    stop("Kenward-Roger is only for REML!")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">369</td>
+                    <td class="num">64</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">370</td>
-                    <td class="coverage">241<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">65</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cov &lt;- lapply(</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">371</td>
-                    <td class="coverage">241<em>x</em></td>
+                    <td class="num">66</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    seq_len(d[1] / d[2]),</pre>
+                      <pre class="language-r">  df &lt;- h_kr_df(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">372</td>
-                    <td class="coverage">241<em>x</em></td>
+                    <td class="num">67</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    function(i) {</pre>
+                      <pre class="language-r">    v0 = object$beta_vcov,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">373</td>
-                    <td class="coverage">278<em>x</em></td>
+                    <td class="num">68</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ret &lt;- tcrossprod(tmb_report$covariance_lower_chol[seq(1 + (i - 1) * d[2], i * d[2]), ])</pre>
+                      <pre class="language-r">    l = matrix(contrast, nrow = 1),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">374</td>
-                    <td class="coverage">278<em>x</em></td>
+                    <td class="num">69</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      dimnames(ret) &lt;- list(visit_names, visit_names)</pre>
+                      <pre class="language-r">    w = component(object, "theta_vcov"),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">375</td>
-                    <td class="coverage">278<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">      return(ret)</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">376</td>
-                    <td class="coverage"></td>
+                    <td class="num">70</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">    p = object$kr_comp$P</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">377</td>
+                    <td class="num">71</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">378</td>
-                    <td class="coverage">241<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  if (identical(tmb_data$n_groups, 1L)) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">379</td>
-                    <td class="coverage">204<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    cov &lt;- cov[[1]]</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">380</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">381</td>
-                    <td class="coverage">37<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    names(cov) &lt;- levels(tmb_data$subject_groups)</pre>
-                    </td>
-                  </tr>
                   <tr class="never">
-                    <td class="num">382</td>
+                    <td class="num">72</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">383</td>
-                    <td class="coverage">241<em>x</em></td>
+                    <td class="num">73</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  return(cov)</pre>
+                      <pre class="language-r">  h_test_1d(object, contrast, df$m)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">384</td>
+                    <td class="num">74</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">385</td>
+                    <td class="num">75</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">386</td>
+                    <td class="num">76</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Build `TMB` Fit Result List</pre>
+                      <pre class="language-r">#' Obtain the Adjusted Kenward-Roger degrees of freedom</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">387</td>
+                    <td class="num">77</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">388</td>
+                    <td class="num">78</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This helper does some simple post-processing of the `TMB` object and</pre>
+                      <pre class="language-r">#' @description Obtains the adjusted Kenward-Roger degrees of freedom and F statistic scale parameter.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">389</td>
+                    <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' optimization results, including setting names, inverting matrices etc.</pre>
+                      <pre class="language-r">#' Used in [h_df_md_kr()] or [h_df_1d_kr].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">390</td>
+                    <td class="num">80</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">391</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_object (`list`)\cr created with [TMB::MakeADFun()].</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">392</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_opt (`list`)\cr optimization result.</pre>
+                      <pre class="language-r">#' @param v0 (`matrix`)\cr unadjusted covariance matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">393</td>
+                    <td class="num">82</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula_parts (`mmrm_tmb_formula_parts`)\cr produced by</pre>
+                      <pre class="language-r">#' @param l (`matrix`)\cr linear combination matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">394</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'  [h_mmrm_tmb_formula_parts()].</pre>
+                      <pre class="language-r">#' @param w (`matrix`)\cr hessian matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">395</td>
+                    <td class="num">84</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
+                      <pre class="language-r">#' @param p (`matrix`)\cr P matrix from [h_get_kr_comp()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">396</td>
+                    <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">397</td>
+                    <td class="num">86</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List of class `mmrm_tmb` with:</pre>
+                      <pre class="language-r">#' @return Named list with elements:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">398</td>
+                    <td class="num">87</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `cov`: estimated covariance matrix, or named list of estimated group specific covariance matrices.</pre>
+                      <pre class="language-r">#' - `m`: `numeric` degrees of freedom.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">399</td>
+                    <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `beta_est`: vector of coefficient estimates.</pre>
+                      <pre class="language-r">#' - `lambda`: `numeric` F statistic scale parameter.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">400</td>
+                    <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `beta_vcov`: Variance-covariance matrix for coefficient estimates.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">401</td>
+                    <td class="num">90</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `beta_vcov_inv_L`: Lower triangular matrix `L` of the inverse variance-covariance matrix decomposition.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">402</td>
+                    <td class="num">91</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `beta_vcov_inv_D`: vector of diagonal matrix `D` of the inverse variance-covariance matrix decomposition.</pre>
+                      <pre class="language-r">h_kr_df &lt;- function(v0, l, w, p) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">403</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">92</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `theta_est`: vector of variance parameter estimates.</pre>
+                      <pre class="language-r">  n_beta &lt;- ncol(v0)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">404</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">93</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `theta_vcov`: variance-covariance matrix for variance parameter estimates.</pre>
+                      <pre class="language-r">  assert_matrix(v0, ncols = n_beta, nrows = n_beta)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">405</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">94</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `neg_log_lik`: obtained negative log-likelihood.</pre>
+                      <pre class="language-r">  assert_matrix(l, ncols = n_beta)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">406</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">95</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `formula_parts`: input.</pre>
+                      <pre class="language-r">  n_theta &lt;- ncol(w)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">407</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">96</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `data`: input.</pre>
+                      <pre class="language-r">  assert_matrix(w, ncols = n_theta, nrows = n_theta)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">408</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">97</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `weights`: input.</pre>
+                      <pre class="language-r">  n_visits &lt;- ncol(p)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">409</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">98</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `reml`: input as a flag.</pre>
+                      <pre class="language-r">  assert_matrix(p, nrows = n_visits * n_theta)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">410</td>
+                    <td class="num">99</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `opt_details`: list with optimization details including convergence code.</pre>
+                      <pre class="language-r">  # see vignettes/kenward.Rmd#279</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">411</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#'   - `tmb_object`: original `TMB` object created with [TMB::MakeADFun()].</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">412</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#'   - `tmb_data`: input.</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">413</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">414</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @details Instead of inverting or decomposing `beta_vcov`, it can be more efficient to use its robust</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">415</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#'   Cholesky decomposition `LDL^T`, therefore we return the corresponding two components `L` and `D`</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">416</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#'   as well since they have been available on the `C++` side already.</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">417</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">418</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">419</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">h_mmrm_tmb_fit &lt;- function(tmb_object,</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">420</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">                           tmb_opt,</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">421</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">100</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                           formula_parts,</pre>
+                      <pre class="language-r">  slvol &lt;- solve(h_quad_form_mat(l, v0))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">422</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">101</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                           tmb_data) {</pre>
+                      <pre class="language-r">  m &lt;- h_quad_form_mat(t(l), slvol)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">423</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">102</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(tmb_object)</pre>
+                      <pre class="language-r">  nl &lt;- nrow(l)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">424</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">103</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(c("fn", "gr", "par", "he"), names(tmb_object))</pre>
+                      <pre class="language-r">  mv0 &lt;- m %*% v0</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">425</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">104</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(tmb_opt)</pre>
+                      <pre class="language-r">  pl &lt;- lapply(seq_len(nrow(p) / ncol(p)), function(x) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">426</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">105</td>
+                    <td class="coverage">108<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(c("par", "objective", "convergence", "message"), names(tmb_opt))</pre>
+                      <pre class="language-r">    ii &lt;- (x - 1) * ncol(p) + 1</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">427</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">106</td>
+                    <td class="coverage">108<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
+                      <pre class="language-r">    jj &lt;- x * ncol(p)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">428</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">107</td>
+                    <td class="coverage">108<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
+                      <pre class="language-r">    p[ii:jj, ]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">429</td>
+                    <td class="num">108</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  })</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">430</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">109</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tmb_report &lt;- tmb_object$report(par = tmb_opt$par)</pre>
+                      <pre class="language-r">  mv0pv0 &lt;- lapply(pl, function(x) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">431</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">110</td>
+                    <td class="coverage">108<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x_matrix_cols &lt;- colnames(tmb_data$x_matrix)</pre>
+                      <pre class="language-r">    mv0 %*% x %*% v0</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">432</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">111</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cov &lt;- h_mmrm_tmb_extract_cov(tmb_report, tmb_data, formula_parts$visit_var, formula_parts$is_spatial)</pre>
+                      <pre class="language-r">  })</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">433</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">112</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  beta_est &lt;- tmb_report$beta</pre>
+                      <pre class="language-r">  a1 &lt;- 0</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">434</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">113</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  names(beta_est) &lt;- x_matrix_cols</pre>
+                      <pre class="language-r">  a2 &lt;- 0</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">435</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">114</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  beta_vcov &lt;- tmb_report$beta_vcov</pre>
+                      <pre class="language-r">  # see vignettes/kenward.Rmd#283</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">436</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">115</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  dimnames(beta_vcov) &lt;- list(x_matrix_cols, x_matrix_cols)</pre>
+                      <pre class="language-r">  for (i in seq_len(length(pl))) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">437</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">116</td>
+                    <td class="coverage">108<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  beta_vcov_inv_L &lt;- tmb_report$XtWX_L # nolint</pre>
+                      <pre class="language-r">    for (j in seq_len(length(pl))) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">438</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">117</td>
+                    <td class="coverage">592<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  beta_vcov_inv_D &lt;- tmb_report$XtWX_D # nolint</pre>
+                      <pre class="language-r">      a1 &lt;- a1 + w[i, j] * h_tr(mv0pv0[[i]]) * h_tr(mv0pv0[[j]])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">439</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">118</td>
+                    <td class="coverage">592<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  theta_est &lt;- tmb_opt$par</pre>
+                      <pre class="language-r">      a2 &lt;- a2 + w[i, j] * h_tr(mv0pv0[[i]] %*% mv0pv0[[j]])</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">440</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">119</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  names(theta_est) &lt;- NULL</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">441</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">120</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  theta_vcov &lt;- try(solve(tmb_object$he(tmb_opt$par)), silent = TRUE)</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">442</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">121</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  opt_details_names &lt;- setdiff(</pre>
+                      <pre class="language-r">  b &lt;- 1 / (2 * nl) * (a1 + 6 * a2)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">443</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">122</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    names(tmb_opt),</pre>
+                      <pre class="language-r">  e &lt;- 1 + a2 / nl</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">444</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">123</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    c("par", "objective")</pre>
+                      <pre class="language-r">  e_star &lt;- 1 / (1 - a2 / nl)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">445</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">124</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  g &lt;- ((nl + 1) * a1 - (nl + 4) * a2) / ((nl + 2) * a2)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">446</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">125</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  structure(</pre>
+                      <pre class="language-r">  denom &lt;- (3 * nl + 2 - 2 * g)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">447</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">126</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    list(</pre>
+                      <pre class="language-r">  c1 &lt;- g / denom</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">448</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">127</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      cov = cov,</pre>
+                      <pre class="language-r">  c2 &lt;- (nl - g) / denom</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">449</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">128</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      beta_est = beta_est,</pre>
+                      <pre class="language-r">  c3 &lt;- (nl + 2 - g) / denom</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">450</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">129</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      beta_vcov = beta_vcov,</pre>
+                      <pre class="language-r">  v_star &lt;- 2 / nl * (1 + c1 * b) / (1 - c2 * b)^2 / (1 - c3 * b)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">451</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">130</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      beta_vcov_inv_L = beta_vcov_inv_L,</pre>
+                      <pre class="language-r">  rho &lt;- v_star / (2 * e_star^2)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">452</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">131</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      beta_vcov_inv_D = beta_vcov_inv_D,</pre>
+                      <pre class="language-r">  m &lt;- 4 + (nl + 2) / (nl * rho - 1)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">453</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">132</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      theta_est = theta_est,</pre>
+                      <pre class="language-r">  lambda &lt;- m / (e_star * (m - 2))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">454</td>
-                    <td class="coverage">239<em>x</em></td>
+                    <td class="num">133</td>
+                    <td class="coverage">28<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      theta_vcov = theta_vcov,</pre>
+                      <pre class="language-r">  list(m = m, lambda = lambda)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">455</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">134</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      neg_log_lik = tmb_opt$objective,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">456</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">135</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      formula_parts = formula_parts,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">457</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">136</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      data = tmb_data$data,</pre>
+                      <pre class="language-r">#' Obtain the Adjusted Covariance Matrix</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">458</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">137</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      weights = tmb_data$weights_vector,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">459</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">138</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      reml = as.logical(tmb_data$reml),</pre>
+                      <pre class="language-r">#' @description Obtains the Kenward-Roger adjusted covariance matrix for the</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">460</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">139</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      opt_details = tmb_opt[opt_details_names],</pre>
+                      <pre class="language-r">#'   coefficient estimates.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">461</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">140</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      tmb_object = tmb_object,</pre>
+                      <pre class="language-r">#' Used in [mmrm()] fitting if method is "Kenward-Roger" or "Kenward-Roger-Linear".</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">462</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">141</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      tmb_data = tmb_data</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">463</td>
+                    <td class="num">142</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ),</pre>
+                      <pre class="language-r">#' @param v (`matrix`)\cr unadjusted covariance matrix.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">464</td>
-                    <td class="coverage">239<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">143</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    class = "mmrm_tmb"</pre>
+                      <pre class="language-r">#' @param w (`matrix`)\cr hessian matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">465</td>
+                    <td class="num">144</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' @param p (`matrix`)\cr P matrix from [h_get_kr_comp()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">466</td>
+                    <td class="num">145</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @param q (`matrix`)\cr Q matrix from [h_get_kr_comp()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">467</td>
+                    <td class="num">146</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @param r (`matrix`)\cr R matrix from [h_get_kr_comp()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">468</td>
+                    <td class="num">147</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Low-Level Fitting Function for MMRM</pre>
+                      <pre class="language-r">#' @param linear (`flag`)\cr whether to use linear Kenward-Roger approximation.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">469</td>
+                    <td class="num">148</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">470</td>
+                    <td class="num">149</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">#' @return The matrix of adjusted covariance matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">471</td>
+                    <td class="num">150</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">472</td>
+                    <td class="num">151</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is the low-level function to fit an MMRM. Note that this does not</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">473</td>
+                    <td class="num">152</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' try different optimizers or adds Jacobian information etc. in contrast to</pre>
+                      <pre class="language-r">h_var_adj &lt;- function(v, w, p, q, r, linear = FALSE) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">474</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">153</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' [mmrm()].</pre>
+                      <pre class="language-r">  assert_flag(linear)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">475</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">154</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  n_beta &lt;- ncol(v)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">476</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">155</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param formula (`formula`)\cr model formula with exactly one special term</pre>
+                      <pre class="language-r">  assert_matrix(v, nrows = n_beta)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">477</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">156</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   specifying the visits within subjects, see details.</pre>
+                      <pre class="language-r">  n_theta &lt;- ncol(w)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">478</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">157</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param data (`data.frame`)\cr input data containing the variables used in</pre>
+                      <pre class="language-r">  assert_matrix(w, nrows = n_theta)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">479</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">158</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `formula`.</pre>
+                      <pre class="language-r">  n_visits &lt;- ncol(p)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">480</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">159</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param weights (`vector`)\cr input vector containing the weights.</pre>
+                      <pre class="language-r">  theta_per_group &lt;- nrow(q) / nrow(p)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">481</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">160</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_mmrm_tmb_data</pre>
+                      <pre class="language-r">  n_groups &lt;- n_theta / theta_per_group</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">482</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">161</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param covariance (`cov_struct`)\cr A covariance structure type definition,</pre>
+                      <pre class="language-r">  assert_matrix(p, nrows = n_theta * n_visits)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">483</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">162</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   or value that can be coerced to a covariance structure using</pre>
+                      <pre class="language-r">  assert_matrix(q, nrows = theta_per_group^2 * n_groups * n_visits, ncols = n_visits)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">484</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">163</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   [as.cov_struct()]. If no value is provided, a structure is derived from</pre>
+                      <pre class="language-r">  assert_matrix(r, nrows = theta_per_group^2 * n_groups * n_visits, ncols = n_visits)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">485</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">164</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   the provided formula.</pre>
+                      <pre class="language-r">  if (linear) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">486</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">165</td>
+                    <td class="coverage">13<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param control (`mmrm_control`)\cr list of control options produced by</pre>
+                      <pre class="language-r">    r &lt;- matrix(0, nrow = nrow(r), ncol = ncol(r))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">487</td>
+                    <td class="num">166</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   [mmrm_control()].</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">488</td>
+                    <td class="num">167</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams fit_single_optimizer</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">489</td>
+                    <td class="num">168</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  # see vignettes/kenward.Rmd#131</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">490</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">169</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List of class `mmrm_tmb`, see [h_mmrm_tmb_fit()] for details.</pre>
+                      <pre class="language-r">  ret &lt;- v</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">491</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">170</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   In addition, it contains elements `call` and `optimizer`.</pre>
+                      <pre class="language-r">  for (i in seq_len(n_theta)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">492</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">171</td>
+                    <td class="coverage">264<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    for (j in seq_len(n_theta)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">493</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">172</td>
+                    <td class="coverage">2164<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">      gi &lt;- ceiling(i / theta_per_group)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">494</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">173</td>
+                    <td class="coverage">2164<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' The `formula` typically looks like:</pre>
+                      <pre class="language-r">      gj &lt;- ceiling(j / theta_per_group)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">174</td>
+                    <td class="coverage">2164<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      iid &lt;- (i - 1) * n_beta + 1</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">175</td>
+                    <td class="coverage">2164<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      jid &lt;- (j - 1) * n_beta + 1</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">495</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">176</td>
+                    <td class="coverage">2164<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      ii &lt;- i - (gi - 1) * theta_per_group</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">496</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">177</td>
+                    <td class="coverage">2164<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)`</pre>
+                      <pre class="language-r">      jj &lt;- j - (gi - 1) * theta_per_group</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">497</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">178</td>
+                    <td class="coverage">2164<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      ijid &lt;- ((ii - 1) * theta_per_group + jj - 1) * n_beta + (gi - 1) * n_beta * theta_per_group^2 + 1</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">498</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">179</td>
+                    <td class="coverage">2164<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' which specifies response and covariates as usual, and exactly one special term</pre>
+                      <pre class="language-r">      if (gi != gj) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">499</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">180</td>
+                    <td class="coverage">592<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' defines which covariance structure is used and what are the visit and</pre>
+                      <pre class="language-r">        ret &lt;- ret + 2 * w[i, j] * v %*% (-p[iid:(iid + n_beta - 1), ] %*% v %*% p[jid:(jid + n_beta - 1), ]) %*% v</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">500</td>
+                    <td class="num">181</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' subject variables.</pre>
+                      <pre class="language-r">      } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">501</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">182</td>
+                    <td class="coverage">1572<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        ret &lt;- ret + 2 * w[i, j] * v %*% (</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">502</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">183</td>
+                    <td class="coverage">1572<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Always use only the first optimizer if multiple optimizers are provided.</pre>
+                      <pre class="language-r">          q[ijid:(ijid + n_beta - 1), ] -</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">503</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">184</td>
+                    <td class="coverage">1572<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">            p[iid:(iid + n_beta - 1), ] %*% v %*% p[jid:(jid + n_beta - 1), ] -</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">504</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">185</td>
+                    <td class="coverage">1572<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">            1 / 4 * r[ijid:(ijid + n_beta - 1), ]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">505</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">186</td>
+                    <td class="coverage">1572<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        ) %*% v</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">506</td>
+                    <td class="num">187</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">      }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">507</td>
+                    <td class="num">188</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' formula &lt;- FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">508</td>
+                    <td class="num">189</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' data &lt;- fev_data</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">509</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">190</td>
+                    <td class="coverage">49<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' system.time(result &lt;- fit_mmrm(formula, data, rep(1, nrow(fev_data))))</pre>
+                      <pre class="language-r">  ret</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">510</td>
+                    <td class="num">191</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">fit_mmrm &lt;- function(formula,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/interop-car.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">511</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                     data,</pre>
+                      <pre class="language-r">#' Register `mmrm` For Use With `car::Anova`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">512</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                     weights,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">513</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                     reml = TRUE,</pre>
+                      <pre class="language-r">#' @inheritParams base::requireNamespace</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">514</td>
+                    <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                     covariance = NULL,</pre>
+                      <pre class="language-r">#' @return A logical value indicating whether registration was successful.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">515</td>
+                    <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                     tmb_data,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">516</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                     formula_parts,</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">517</td>
+                    <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                     control = mmrm_control()) {</pre>
+                      <pre class="language-r">car_add_mmrm &lt;- function(quietly = FALSE) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">518</td>
-                    <td class="coverage">252<em>x</em></td>
+                    <td class="num">8</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(formula_parts) || missing(tmb_data)) {</pre>
+                      <pre class="language-r">  if (!requireNamespace("car", quietly = quietly)) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">519</td>
-                    <td class="coverage">67<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">9</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    covariance &lt;- h_reconcile_cov_struct(formula, covariance)</pre>
+                      <pre class="language-r">    return(FALSE)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">520</td>
-                    <td class="coverage">65<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">10</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    formula_parts &lt;- h_mmrm_tmb_formula_parts(formula, covariance)</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">521</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">11</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  envir &lt;- asNamespace("mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">522</td>
-                    <td class="coverage">65<em>x</em></td>
+                    <td class="num">12</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (!formula_parts$is_spatial &amp;&amp; !is.factor(data[[formula_parts$visit_var]])) {</pre>
+                      <pre class="language-r">  h_register_s3("car", "Anova", "mmrm", envir)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">523</td>
+                    <td class="num">13</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      stop("Time variable must be a factor for non-spatial covariance structures")</pre>
+                      <pre class="language-r">  TRUE</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">524</td>
+                    <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">525</td>
+                    <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">526</td>
-                    <td class="coverage">64<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">16</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_class(control, "mmrm_control")</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">527</td>
-                    <td class="coverage">64<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">17</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_list(control$optimizers, min.len = 1)</pre>
+                      <pre class="language-r">#' Obtain Contrast for Specified Effect</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">528</td>
-                    <td class="coverage">64<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">18</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_numeric(weights, any.missing = FALSE)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">529</td>
-                    <td class="coverage">64<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">19</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_true(all(weights &gt; 0))</pre>
+                      <pre class="language-r">#' This is support function to obtain contrast matrix for type II/III testing.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">530</td>
-                    <td class="coverage">64<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">20</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tmb_data &lt;- h_mmrm_tmb_data(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">531</td>
-                    <td class="coverage">64<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">21</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      formula_parts, data, weights, reml,</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">532</td>
-                    <td class="coverage">64<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">22</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      singular = if (control$accept_singular) "drop" else "error", drop_visit_levels = control$drop_visit_levels</pre>
+                      <pre class="language-r">#' @param effect (`string`) the name of the effect.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">533</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' @param type (`string`) type of test, "II", "III", '2', or '3'.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">534</td>
+                    <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">#' @param tol (`numeric`) threshold blow which values are treated as 0.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">535</td>
-                    <td class="coverage">185<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">25</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_class(tmb_data, "mmrm_tmb_data")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">536</td>
-                    <td class="coverage">185<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">26</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_class(formula_parts, "mmrm_tmb_formula_parts")</pre>
+                      <pre class="language-r">#' @return A `matrix` of the contrast.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">537</td>
+                    <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">538</td>
-                    <td class="coverage">249<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">28</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tmb_parameters &lt;- h_mmrm_tmb_parameters(formula_parts, tmb_data, start = control$start, n_groups = tmb_data$n_groups)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">539</td>
+                    <td class="num">29</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">h_get_contrast &lt;- function(object, effect, type = c("II", "III", "2", "3"), tol = sqrt(.Machine$double.eps)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">540</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">30</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tmb_object &lt;- TMB::MakeADFun(</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">541</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">31</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    data = tmb_data,</pre>
+                      <pre class="language-r">  assert_string(effect)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">542</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">32</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    parameters = tmb_parameters,</pre>
+                      <pre class="language-r">  assert_double(tol, finite = TRUE, len = 1L)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">543</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">33</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    hessian = TRUE,</pre>
+                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">544</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">34</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    DLL = "mmrm",</pre>
+                      <pre class="language-r">  mx &lt;- component(object, "x_matrix")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">545</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">35</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    silent = TRUE</pre>
+                      <pre class="language-r">  asg &lt;- attr(mx, "assign")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">546</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">36</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  formula &lt;- object$formula_parts$model_formula</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">547</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">37</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_mmrm_tmb_assert_start(tmb_object)</pre>
+                      <pre class="language-r">  tms &lt;- terms(formula)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">548</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">38</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  used_optimizer &lt;- control$optimizers[[1L]]</pre>
+                      <pre class="language-r">  fcts &lt;- attr(tms, "factors")[-1L, , drop = FALSE] # Discard the response.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">549</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">39</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  used_optimizer_name &lt;- names(control$optimizers)[1L]</pre>
+                      <pre class="language-r">  ods &lt;- attr(tms, "order")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">550</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">40</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  args &lt;- with(</pre>
+                      <pre class="language-r">  assert_subset(effect, colnames(fcts))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">551</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">41</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tmb_object,</pre>
+                      <pre class="language-r">  idx &lt;- which(effect == colnames(fcts))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">552</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">42</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    c(</pre>
+                      <pre class="language-r">  cols &lt;- which(asg == idx)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">553</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">43</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      list(par, fn, gr),</pre>
+                      <pre class="language-r">  xlev &lt;- component(object, "xlev")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">554</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">44</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      attr(used_optimizer, "args")</pre>
+                      <pre class="language-r">  contains_intercept &lt;- (!0 %in% asg) &amp;&amp; h_first_contain_categorical(effect, fcts, names(xlev))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">555</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">45</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">  coef_rows &lt;- length(cols) - as.integer(contains_intercept)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">556</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">46</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  l_mx &lt;- matrix(0, nrow = coef_rows, ncol = length(asg))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">557</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">47</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (identical(attr(used_optimizer, "use_hessian"), TRUE)) {</pre>
+                      <pre class="language-r">  if (coef_rows == 0L) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">558</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="num">48</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    args$hessian &lt;- tmb_object$he</pre>
+                      <pre class="language-r">    return(l_mx)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">559</td>
+                    <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">560</td>
-                    <td class="coverage">246<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  tmb_opt &lt;- do.call(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">561</td>
-                    <td class="coverage">246<em>x</em></td>
+                    <td class="num">50</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    what = used_optimizer,</pre>
+                      <pre class="language-r">  if (contains_intercept) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">562</td>
-                    <td class="coverage">246<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    args = args</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">563</td>
-                    <td class="coverage"></td>
+                    <td class="num">51</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    l_mx[, cols] &lt;- cbind(-1, diag(rep(1, coef_rows)))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">564</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Ensure negative log likelihood is stored in `objective` element of list.</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">565</td>
-                    <td class="coverage">237<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  if ("value" %in% names(tmb_opt)) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">566</td>
-                    <td class="coverage">227<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    tmb_opt$objective &lt;- tmb_opt$value</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">567</td>
-                    <td class="coverage">227<em>x</em></td>
+                    <td class="num">53</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tmb_opt$value &lt;- NULL</pre>
+                      <pre class="language-r">    l_mx[, cols] &lt;- diag(rep(1, coef_rows))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">568</td>
+                    <td class="num">54</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">569</td>
-                    <td class="coverage">237<em>x</em></td>
+                    <td class="num">55</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fit &lt;- h_mmrm_tmb_fit(tmb_object, tmb_opt, formula_parts, tmb_data)</pre>
+                      <pre class="language-r">  for (i in setdiff(seq_len(ncol(fcts)), idx)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">570</td>
-                    <td class="coverage">237<em>x</em></td>
+                    <td class="num">56</td>
+                    <td class="coverage">120<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_mmrm_tmb_check_conv(tmb_opt, fit)</pre>
+                      <pre class="language-r">    additional_vars &lt;- names(which(fcts[, i] &gt; fcts[, idx]))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">571</td>
-                    <td class="coverage">237<em>x</em></td>
+                    <td class="num">57</td>
+                    <td class="coverage">120<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fit$call &lt;- match.call()</pre>
+                      <pre class="language-r">    additional_numeric &lt;- any(!additional_vars %in% names(xlev))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">572</td>
-                    <td class="coverage">237<em>x</em></td>
+                    <td class="num">58</td>
+                    <td class="coverage">120<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fit$call$formula &lt;- formula_parts$formula</pre>
+                      <pre class="language-r">    current_col &lt;- which(asg == i)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">573</td>
-                    <td class="coverage">237<em>x</em></td>
+                    <td class="num">59</td>
+                    <td class="coverage">120<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fit$optimizer &lt;- used_optimizer_name</pre>
+                      <pre class="language-r">    if (ods[i] &gt;= ods[idx] &amp;&amp; all(fcts[, i] &gt;= fcts[, idx]) &amp;&amp; !additional_numeric) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">574</td>
-                    <td class="coverage">237<em>x</em></td>
+                    <td class="num">60</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fit</pre>
+                      <pre class="language-r">      sub_mat &lt;- switch(type,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">575</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">61</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">        "2" = ,</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/zzz.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
-                  <tr class="never">
-                    <td class="num">1</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">62</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Dynamic Registration for Package Interoperability</pre>
+                      <pre class="language-r">        "II" = {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">2</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">63</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">          x1 &lt;- mx[, cols, drop = FALSE]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">3</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">64</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @seealso See `vignette("xtending", package = "emmeans")` for background.</pre>
+                      <pre class="language-r">          x0 &lt;- mx[, -c(cols, current_col), drop = FALSE]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">4</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">65</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">          x2 &lt;- mx[, current_col, drop = FALSE]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">5</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">66</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @noRd</pre>
+                      <pre class="language-r">          m &lt;- diag(rep(1, nrow(x0))) - x0 %*% solve(t(x0) %*% x0) %*% t(x0)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">6</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">67</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">.onLoad &lt;- function(libname, pkgname) { # nolint</pre>
+                      <pre class="language-r">          ret &lt;- solve(t(x1) %*% m %*% x1) %*% t(x1) %*% m %*% x2</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">7</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">68</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (utils::packageVersion("TMB") &lt; "1.9.15") {</pre>
+                      <pre class="language-r">          if (contains_intercept) {</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">8</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">69</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning("TMB version 1.9.15 or higher is required for reproducible model fits", call. = FALSE)</pre>
+                      <pre class="language-r">            ret[-1, ] - ret[1, ]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">          } else {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">71</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">            ret</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">72</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">          }</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">11</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">73</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  register_on_load(</pre>
+                      <pre class="language-r">        },</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">12</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">74</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "emmeans", c("1.6", NA),</pre>
+                      <pre class="language-r">        "3" = ,</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">13</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">75</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    callback = function() emmeans::.emm_register("mmrm", pkgname),</pre>
+                      <pre class="language-r">        "III" = {</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">14</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">76</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    message = "mmrm() registered as emmeans extension"</pre>
+                      <pre class="language-r">          lvls &lt;- h_obtain_lvls(effect, additional_vars, xlev)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">15</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">77</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">          t_levels &lt;- lvls$total</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">16</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">78</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">          nms_base &lt;- colnames(mx)[cols]</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">17</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">79</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  register_on_load(</pre>
+                      <pre class="language-r">          nms &lt;- colnames(mx)[current_col]</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">18</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">80</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "parsnip", c("1.1.0", NA),</pre>
+                      <pre class="language-r">          nms_base_split &lt;- strsplit(nms_base, ":")</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">19</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">81</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    callback = parsnip_add_mmrm,</pre>
+                      <pre class="language-r">          nms_split &lt;- strsplit(nms, ":")</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">20</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">82</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    message = emit_tidymodels_register_msg</pre>
+                      <pre class="language-r">          base_idx &lt;- h_get_index(nms_split, nms_base_split)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">21</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">83</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">          mt &lt;- l_mx[, cols, drop = FALSE] / t_levels</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">22</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">84</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  register_on_load(</pre>
+                      <pre class="language-r">          ret &lt;- mt[, base_idx, drop = FALSE]</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">23</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">85</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "car", c("3.1.2", NA),</pre>
+                      <pre class="language-r">          # if there is extra levels, replace it with -1/t_levels</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">24</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">86</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    callback = car_add_mmrm,</pre>
+                      <pre class="language-r">          ret[is.na(ret)] &lt;- -1 / t_levels</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">25</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">87</td>
+                    <td class="coverage">16<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    message = "mmrm() registered as car::Anova extension"</pre>
+                      <pre class="language-r">          ret</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">26</td>
+                    <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">        }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">27</td>
+                    <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">28</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">90</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      l_mx[, current_col] &lt;- sub_mat</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">29</td>
+                    <td class="num">91</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Helper Function for Registering Functionality With Suggests Packages</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">30</td>
+                    <td class="num">92</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">31</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">93</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams check_package_version</pre>
+                      <pre class="language-r">  l_mx[abs(l_mx) &lt; tol] &lt;- 0</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">94</td>
+                    <td class="coverage">44<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  l_mx</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">32</td>
+                    <td class="num">95</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">33</td>
+                    <td class="num">96</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param callback (`function(...) ANY`)\cr a callback to execute upon package</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">34</td>
+                    <td class="num">97</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   load. Note that no arguments are passed to this function. Any necessary</pre>
+                      <pre class="language-r">#' Conduct type II/III hypothesis testing on the MMRM fit results.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">35</td>
+                    <td class="num">98</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   data must be provided upon construction.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">36</td>
+                    <td class="num">99</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param mod (`mmrm`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">37</td>
+                    <td class="num">100</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param message (`NULL` or `string`)\cr an optional message to print after</pre>
+                      <pre class="language-r">#' @param ... not used.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">38</td>
+                    <td class="num">101</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   the callback is executed upon successful registration.</pre>
+                      <pre class="language-r">#' @inheritParams h_get_contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">39</td>
+                    <td class="num">102</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">40</td>
+                    <td class="num">103</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A logical (invisibly) indicating whether registration was successful.</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">41</td>
+                    <td class="num">104</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'  If not, a onLoad hook was set for the next time the package is loaded.</pre>
+                      <pre class="language-r">#' `Anova` will return `anova` object with one row per variable and columns</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">42</td>
+                    <td class="num">105</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' `Num Df`(numerator degrees of freedom), `Denom Df`(denominator degrees of freedom),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">43</td>
+                    <td class="num">106</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' `F Statistic` and `Pr(&gt;=F)`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">44</td>
+                    <td class="num">107</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">register_on_load &lt;- function(pkg,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">45</td>
+                    <td class="num">108</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             ver = c(NA_character_, NA_character_),</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">46</td>
+                    <td class="num">109</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             callback,</pre>
+                      <pre class="language-r"># Please do not load `car` and then create the documentation. The Rd file will be different.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">47</td>
+                    <td class="num">110</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             message = NULL) {</pre>
+                      <pre class="language-r">Anova.mmrm &lt;- function(mod, type = c("II", "III", "2", "3"), tol = sqrt(.Machine$double.eps), ...) { # nolint</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">48</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">111</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (isNamespaceLoaded(pkg) &amp;&amp; check_package_version(pkg, ver)) {</pre>
+                      <pre class="language-r">  assert_double(tol, finite = TRUE, len = 1L)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">49</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">112</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    callback()</pre>
+                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">50</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">113</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (is.character(message)) packageStartupMessage(message)</pre>
+                      <pre class="language-r">  vars &lt;- colnames(attr(terms(mod$formula_parts$model_formula), "factors"))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">51</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">114</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (is.function(message)) packageStartupMessage(message())</pre>
+                      <pre class="language-r">  ret &lt;- lapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">52</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">115</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(invisible(TRUE))</pre>
+                      <pre class="language-r">    vars,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">53</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">116</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    function(x) df_md(mod, h_get_contrast(mod, x, type, tol))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">117</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">55</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">118</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  setHook(</pre>
+                      <pre class="language-r">  ret_df &lt;- do.call(rbind.data.frame, ret)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">56</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">119</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    packageEvent(pkg, event = "onLoad"),</pre>
+                      <pre class="language-r">  row.names(ret_df) &lt;- vars</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">57</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">120</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    action = "append",</pre>
+                      <pre class="language-r">  colnames(ret_df) &lt;- c("Num Df", "Denom Df", "F Statistic", "Pr(&gt;=F)")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">58</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">121</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    function(...) {</pre>
+                      <pre class="language-r">  class(ret_df) &lt;- c("anova", "data.frame")</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">59</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">122</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      register_on_load(</pre>
+                      <pre class="language-r">  attr(ret_df, "heading") &lt;- sprintf(</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">60</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">123</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        pkg = pkg,</pre>
+                      <pre class="language-r">    "Analysis of Fixed Effect Table (Type %s F tests)",</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">61</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">124</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        ver = ver,</pre>
+                      <pre class="language-r">    switch(type,</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">62</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">125</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        callback = callback,</pre>
+                      <pre class="language-r">      "2" = ,</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">63</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">126</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        message = message</pre>
+                      <pre class="language-r">      "II" = "II",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">64</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">127</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">      "3" = ,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">65</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">128</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">      "III" = "III"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">66</td>
+                    <td class="num">129</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">67</td>
+                    <td class="num">130</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">68</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">131</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  invisible(FALSE)</pre>
+                      <pre class="language-r">  ret_df</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">69</td>
+                    <td class="num">132</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">70</td>
+                    <td class="num">133</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">71</td>
+                    <td class="num">134</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Check Suggested Dependency Against Version Requirements</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">72</td>
+                    <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Obtain Levels Prior and Posterior</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">73</td>
+                    <td class="num">136</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param pkg (`string`)\cr package name.</pre>
+                      <pre class="language-r">#' @param var (`string`) name of the effect.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">74</td>
+                    <td class="num">137</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ver (`character`)\cr of length 2 whose elements can be provided to</pre>
+                      <pre class="language-r">#' @param additional_vars (`character`) names of additional variables.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">75</td>
+                    <td class="num">138</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   [numeric_version()], representing a minimum and maximum (inclusive) version</pre>
+                      <pre class="language-r">#' @param xlev (`list`) named list of character levels.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">76</td>
+                    <td class="num">139</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   requirement for interoperability. When `NA`, no version requirement is</pre>
+                      <pre class="language-r">#' @param factors (`matrix`) the factor matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">77</td>
+                    <td class="num">140</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   imposed. Defaults to no version requirement.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">78</td>
+                    <td class="num">141</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">h_obtain_lvls &lt;- function(var, additional_vars, xlev, factors) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">79</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">142</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A logical (invisibly) indicating whether the loaded package meets</pre>
+                      <pre class="language-r">  assert_string(var)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">80</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">143</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   the version requirements. A warning is emitted otherwise.</pre>
+                      <pre class="language-r">  assert_character(additional_vars)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">81</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">144</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_list(xlev, types = "character")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">82</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">145</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  nms &lt;- names(xlev)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">83</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">146</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">check_package_version &lt;- function(pkg, ver = c(NA_character_, NA_character_)) {</pre>
+                      <pre class="language-r">  assert_subset(additional_vars, nms)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">84</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">147</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_character(ver, len = 2L)</pre>
+                      <pre class="language-r">  if (var %in% nms) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">85</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">148</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  pkg_ver &lt;- utils::packageVersion(pkg)</pre>
+                      <pre class="language-r">    prior_vars &lt;- intersect(nms[seq_len(match(var, nms) - 1)], additional_vars)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">86</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">149</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ver &lt;- numeric_version(ver, strict = FALSE)</pre>
+                      <pre class="language-r">    prior_lvls &lt;- vapply(xlev[prior_vars], length, FUN.VALUE = 1L)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">87</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">150</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    post_vars &lt;- intersect(nms[seq(match(var, nms) + 1, length(nms))], additional_vars)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">88</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">151</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  warn_version &lt;- function(pkg, pkg_ver, ver) {</pre>
+                      <pre class="language-r">    post_lvls &lt;- vapply(xlev[post_vars], length, FUN.VALUE = 1L)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">89</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">152</td>
+                    <td class="coverage">14<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ver_na &lt;- is.na(ver)</pre>
+                      <pre class="language-r">    total_lvls &lt;- prod(prior_lvls) * prod(post_lvls)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">153</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warning(sprintf(</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">154</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Cannot register mmrm for use with %s (v%s). Version %s required.",</pre>
+                      <pre class="language-r">    prior_lvls &lt;- vapply(xlev[additional_vars], length, FUN.VALUE = 1L)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">155</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      pkg, pkg_ver,</pre>
+                      <pre class="language-r">    post_lvls &lt;- 2L</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">156</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (!any(ver_na)) {</pre>
+                      <pre class="language-r">    total_lvls &lt;- prod(prior_lvls)</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">94</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">157</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        sprintf("%s to %s", ver[1], ver[2])</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">95</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">158</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else if (ver_na[2]) {</pre>
+                      <pre class="language-r">  list(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">96</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">159</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        paste0("&gt;= ", ver[1])</pre>
+                      <pre class="language-r">    prior = prior_lvls,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">97</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">160</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else if (ver_na[1]) {</pre>
+                      <pre class="language-r">    post = post_lvls,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">98</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">161</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        paste0("&lt;= ", ver[2])</pre>
+                      <pre class="language-r">    total = total_lvls</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">99</td>
+                    <td class="num">162</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      }</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">100</td>
+                    <td class="num">163</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ))</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">101</td>
+                    <td class="num">164</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">102</td>
+                    <td class="num">165</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' Check if the Effect is the First Categorical Effect</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">103</td>
-                    <td class="coverage">6<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">166</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (identical(pkg_ver &lt; ver[1], TRUE) || identical(pkg_ver &gt; ver[2], TRUE)) {</pre>
+                      <pre class="language-r">#' @param effect (`string`) name of the effect.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">104</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">167</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    warn_version(pkg, pkg_ver, ver)</pre>
+                      <pre class="language-r">#' @param categorical (`character`) names of the categorical values.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">105</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">168</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(invisible(FALSE))</pre>
+                      <pre class="language-r">#' @param factors (`matrix`) the factor matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">106</td>
+                    <td class="num">169</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">107</td>
+                    <td class="num">170</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">h_first_contain_categorical &lt;- function(effect, factors, categorical) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">108</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">171</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  invisible(TRUE)</pre>
+                      <pre class="language-r">  assert_string(effect)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">109</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">172</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  assert_matrix(factors)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">110</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">173</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  assert_character(categorical)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">111</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">174</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Format a Message to Emit When Tidymodels is Loaded</pre>
+                      <pre class="language-r">  mt &lt;- match(effect, colnames(factors))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">112</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">175</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  varnms &lt;- row.names(factors)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">113</td>
+                    <td class="num">176</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A character message to emit. Either a ansi-formatted cli output if</pre>
+                      <pre class="language-r">  # if the effect is not categorical in any value, return FALSE</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">114</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">177</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   package 'cli' is available or a plain-text message otherwise.</pre>
+                      <pre class="language-r">  if (!any(varnms[factors[, mt] &gt; 0] %in% categorical)) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">178</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    return(FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">115</td>
+                    <td class="num">179</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">116</td>
+                    <td class="num">180</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  # keep only categorical rows that is in front of the current factor</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">181</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  factors &lt;- factors[row.names(factors) %in% categorical, seq_len(mt - 1L), drop = FALSE]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">117</td>
+                    <td class="num">182</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">emit_tidymodels_register_msg &lt;- function() {</pre>
+                      <pre class="language-r">  # if previous cols are all numerical, return TRUE</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">118</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">183</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  pkg &lt;- utils::packageName()</pre>
+                      <pre class="language-r">  if (ncol(factors) &lt; 1L) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">119</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">184</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ver &lt;- utils::packageVersion(pkg)</pre>
+                      <pre class="language-r">    return(TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">120</td>
+                    <td class="num">185</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">121</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">186</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (isTRUE(getOption("tidymodels.quiet"))) {</pre>
+                      <pre class="language-r">  col_ind &lt;- apply(factors, 2, prod)</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">122</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">187</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return()</pre>
+                      <pre class="language-r">  # if any of the previous cols are categorical, return FALSE</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">123</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">188</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  return(!any(col_ind &gt; 0))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">124</td>
+                    <td class="num">189</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">125</td>
+                    <td class="num">190</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # if tidymodels is attached, cli packages come as a dependency</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">126</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">191</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  has_cli &lt;- requireNamespace("cli", quietly = TRUE)</pre>
+                      <pre class="language-r">#' Test if the First Vector is Subset of the Second Vector</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">127</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">192</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (has_cli) {</pre>
+                      <pre class="language-r">#' @param x (`vector`) the first list.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">128</td>
+                    <td class="num">193</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # unfortunately, cli does not expose many formatting tools for emitting</pre>
+                      <pre class="language-r">#' @param y (`vector`) the second list.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">129</td>
+                    <td class="num">194</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # messages (only via conditions to stderr) which can't be suppressed using</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">130</td>
+                    <td class="num">195</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # suppressPackageStartupMessages() so formatting must be done adhoc,</pre>
+                      <pre class="language-r">h_get_index &lt;- function(x, y) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">131</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">196</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # similar to how it's done in {tidymodels} R/attach.R</pre>
+                      <pre class="language-r">  assert_list(x)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">132</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">197</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    paste0(</pre>
+                      <pre class="language-r">  assert_list(y)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">133</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">198</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      cli::rule(</pre>
+                      <pre class="language-r">  vapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">134</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">199</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        left = cli::style_bold("Model Registration"),</pre>
+                      <pre class="language-r">    x,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">135</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">200</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        right = paste(pkg, ver)</pre>
+                      <pre class="language-r">    \(i) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">136</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">201</td>
+                    <td class="coverage">68<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ),</pre>
+                      <pre class="language-r">      r &lt;- vapply(y, \(j) test_subset(j, i), FUN.VALUE = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">137</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">202</td>
+                    <td class="coverage">68<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "\n",</pre>
+                      <pre class="language-r">      if (sum(r) == 1L) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">138</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">203</td>
+                    <td class="coverage">65<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      cli::col_green(cli::symbol$tick), " ",</pre>
+                      <pre class="language-r">        which(r)</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">204</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">139</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">205</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      cli::col_blue("mmrm"), "::", cli::col_green("mmrm()")</pre>
+                      <pre class="language-r">        NA_integer_</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">140</td>
+                    <td class="num">206</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">      }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">141</td>
+                    <td class="num">207</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">    },</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">142</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">208</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    paste0(pkg, "::mmrm() registered for use with tidymodels")</pre>
+                      <pre class="language-r">    FUN.VALUE = 1L</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">143</td>
+                    <td class="num">209</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">144</td>
+                    <td class="num">210</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
@@ -18208,14 +17716,14 @@ <h2>mmrm coverage - 97.05%</h2>
                 </tbody>
               </table>
             </div>
-            <div id="R/interop-car.R" class="hidden">
+            <div id="R/cov_struct.R" class="hidden">
               <table class="table-condensed">
                 <tbody>
                   <tr class="never">
                     <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Register `mmrm` For Use With `car::Anova`</pre>
+                      <pre class="language-r">#' Covariance Type Database</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -18229,168 +17737,168 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams base::requireNamespace</pre>
+                      <pre class="language-r">#' An internal constant for covariance type information.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A logical value indicating whether registration was successful.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @format A data frame with 5 variables and one record per covariance type:</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">car_add_mmrm &lt;- function(quietly = FALSE) {</pre>
+                      <pre class="language-r">#' \describe{</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">8</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!requireNamespace("car", quietly = quietly)) {</pre>
+                      <pre class="language-r">#'   \item{name}{</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
+                  <tr class="never">
                     <td class="num">9</td>
-                    <td class="coverage">!</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(FALSE)</pre>
+                      <pre class="language-r">#'     The long-form name of the covariance structure type</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'   }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">11</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  envir &lt;- asNamespace("mmrm")</pre>
+                      <pre class="language-r">#'   \item{abbr}{</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">12</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_register_s3("car", "Anova", "mmrm", envir)</pre>
+                      <pre class="language-r">#'     The abbreviated name of the covariance structure type</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">13</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  TRUE</pre>
+                      <pre class="language-r">#'   }</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'   \item{habbr}{</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'     The abbreviated name of the heterogeneous version of a covariance</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'     structure type (The abbreviated name (`abbr`) with a trailing `"h"` if</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">17</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain Contrast for Specified Effect</pre>
+                      <pre class="language-r">#'     the structure has a heterogeneous implementation or `NA` otherwise).</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">18</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   }</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is support function to obtain contrast matrix for type II/III testing.</pre>
+                      <pre class="language-r">#'   \item{heterogeneous}{</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'     A logical value indicating whether the covariance structure has a</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r">#'     heterogeneous counterpart.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">22</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param effect (`string`) the name of the effect.</pre>
+                      <pre class="language-r">#'   }</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param type (`string`) type of test, "II", "III", '2', or '3'.</pre>
+                      <pre class="language-r">#'   \item{spatial}{</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tol (`numeric`) threshold blow which values are treated as 0.</pre>
+                      <pre class="language-r">#'     A logical value indicating whether the covariance structure is spatial.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">25</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   }</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A `matrix` of the contrast.</pre>
+                      <pre class="language-r">#' }</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -18411,511 +17919,511 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">29</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_contrast &lt;- function(object, effect, type = c("II", "III", "2", "3"), tol = sqrt(.Machine$double.eps)) {</pre>
+                      <pre class="language-r">COV_TYPES &lt;- local({ # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">30</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">  type &lt;- function(name, abbr, habbr, heterogeneous, spatial) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">31</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(effect)</pre>
+                      <pre class="language-r">    args &lt;- as.list(match.call()[-1])</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">32</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_double(tol, finite = TRUE, len = 1L)</pre>
+                      <pre class="language-r">    do.call(data.frame, args)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">33</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">34</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  mx &lt;- component(object, "x_matrix")</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">35</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  asg &lt;- attr(mx, "assign")</pre>
+                      <pre class="language-r">  as.data.frame(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">36</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  formula &lt;- object$formula_parts$model_formula</pre>
+                      <pre class="language-r">    col.names = names(formals(type)),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">37</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tms &lt;- terms(formula)</pre>
+                      <pre class="language-r">    rbind(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">38</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fcts &lt;- attr(tms, "factors")[-1L, , drop = FALSE] # Discard the response.</pre>
+                      <pre class="language-r">      type("unstructured", "us", NA, FALSE, FALSE),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">39</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ods &lt;- attr(tms, "order")</pre>
+                      <pre class="language-r">      type("Toeplitz", "toep", "toeph", TRUE, FALSE),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">40</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(effect, colnames(fcts))</pre>
+                      <pre class="language-r">      type("auto-regressive order one", "ar1", "ar1h", TRUE, FALSE),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">41</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  idx &lt;- which(effect == colnames(fcts))</pre>
+                      <pre class="language-r">      type("ante-dependence", "ad", "adh", TRUE, FALSE),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">42</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cols &lt;- which(asg == idx)</pre>
+                      <pre class="language-r">      type("compound symmetry", "cs", "csh", TRUE, FALSE),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">43</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  xlev &lt;- component(object, "xlev")</pre>
+                      <pre class="language-r">      type("spatial exponential", "sp_exp", NA, FALSE, TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">44</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  contains_intercept &lt;- (!0 %in% asg) &amp;&amp; h_first_contain_categorical(effect, fcts, names(xlev))</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">45</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_rows &lt;- length(cols) - as.integer(contains_intercept)</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">46</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  l_mx &lt;- matrix(0, nrow = coef_rows, ncol = length(asg))</pre>
+                      <pre class="language-r">})</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">47</td>
-                    <td class="coverage">45<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (coef_rows == 0L) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">48</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(l_mx)</pre>
+                      <pre class="language-r">#' Covariance Types</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">50</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (contains_intercept) {</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">51</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    l_mx[, cols] &lt;- cbind(-1, diag(rep(1, coef_rows)))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">#' @param form (`character`)\cr covariance structure type name form. One or</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">53</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    l_mx[, cols] &lt;- diag(rep(1, coef_rows))</pre>
+                      <pre class="language-r">#'   more of `"name"`, `"abbr"` (abbreviation), or `"habbr"` (heterogeneous</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">54</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'   abbreviation).</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">55</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  for (i in setdiff(seq_len(ncol(fcts)), idx)) {</pre>
+                      <pre class="language-r">#' @param filter (`character`)\cr covariance structure type filter. One or</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">56</td>
-                    <td class="coverage">120<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    additional_vars &lt;- names(which(fcts[, i] &gt; fcts[, idx]))</pre>
+                      <pre class="language-r">#'   more of `"heterogeneous"` or `"spatial"`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">57</td>
-                    <td class="coverage">120<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    additional_numeric &lt;- any(!additional_vars %in% names(xlev))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">58</td>
-                    <td class="coverage">120<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    current_col &lt;- which(asg == i)</pre>
+                      <pre class="language-r">#' @return A character vector of accepted covariance structure type names and</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">59</td>
-                    <td class="coverage">120<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (ods[i] &gt;= ods[idx] &amp;&amp; all(fcts[, i] &gt;= fcts[, idx]) &amp;&amp; !additional_numeric) {</pre>
+                      <pre class="language-r">#'   abbreviations.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">60</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      sub_mat &lt;- switch(type,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">61</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "2" = ,</pre>
+                      <pre class="language-r">#' @section Abbreviations for Covariance Structures:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">62</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "II" = {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">63</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          x1 &lt;- mx[, cols, drop = FALSE]</pre>
+                      <pre class="language-r">#' ## Common Covariance Structures:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">64</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          x0 &lt;- mx[, -c(cols, current_col), drop = FALSE]</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">65</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          x2 &lt;- mx[, current_col, drop = FALSE]</pre>
+                      <pre class="language-r">#' \tabular{clll}{</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">66</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          m &lt;- diag(rep(1, nrow(x0))) - x0 %*% solve(t(x0) %*% x0) %*% t(x0)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">67</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          ret &lt;- solve(t(x1) %*% m %*% x1) %*% t(x1) %*% m %*% x2</pre>
+                      <pre class="language-r">#' \strong{Structure}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">68</td>
-                    <td class="coverage">8<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          if (contains_intercept) {</pre>
+                      <pre class="language-r">#' \tab \strong{Description}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">69</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">            ret[-1, ] - ret[1, ]</pre>
+                      <pre class="language-r">#' \tab \strong{Parameters}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          } else {</pre>
+                      <pre class="language-r">#' \tab \strong{\eqn{(i, j)} element}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">71</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">            ret</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">72</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        },</pre>
+                      <pre class="language-r">#' ad</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">74</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "3" = ,</pre>
+                      <pre class="language-r">#' \tab Ante-dependence</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">75</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "III" = {</pre>
+                      <pre class="language-r">#' \tab \eqn{m}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">76</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          lvls &lt;- h_obtain_lvls(effect, additional_vars, xlev)</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\prod_{k=i}^{j-1}\rho_{k}}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">77</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          t_levels &lt;- lvls$total</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">78</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          nms_base &lt;- colnames(mx)[cols]</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">79</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          nms &lt;- colnames(mx)[current_col]</pre>
+                      <pre class="language-r">#' adh</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">80</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          nms_base_split &lt;- strsplit(nms_base, ":")</pre>
+                      <pre class="language-r">#' \tab Heterogeneous ante-dependence</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">81</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          nms_split &lt;- strsplit(nms, ":")</pre>
+                      <pre class="language-r">#' \tab \eqn{2m-1}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">82</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          base_idx &lt;- h_get_index(nms_split, nms_base_split)</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\prod_{k=i}^{j-1}\rho_{k}}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">83</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          mt &lt;- l_mx[, cols, drop = FALSE] / t_levels</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">84</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          ret &lt;- mt[, base_idx, drop = FALSE]</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          # if there is extra levels, replace it with -1/t_levels</pre>
+                      <pre class="language-r">#' ar1</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">86</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          ret[is.na(ret)] &lt;- -1 / t_levels</pre>
+                      <pre class="language-r">#' \tab First-order auto-regressive</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">87</td>
-                    <td class="coverage">16<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          ret</pre>
+                      <pre class="language-r">#' \tab \eqn{2}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        }</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\rho^{\left \vert {i-j} \right \vert}}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">90</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      l_mx[, current_col] &lt;- sub_mat</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">91</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#' ar1h</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">92</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' \tab Heterogeneous first-order auto-regressive</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">93</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  l_mx[abs(l_mx) &lt; tol] &lt;- 0</pre>
+                      <pre class="language-r">#' \tab \eqn{m+1}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">94</td>
-                    <td class="coverage">44<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  l_mx</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\rho^{\left \vert {i-j} \right \vert}}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">95</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">96</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">97</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Conduct type II/III hypothesis testing on the MMRM fit results.</pre>
+                      <pre class="language-r">#' cs</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">98</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' \tab Compound symmetry</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">99</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param mod (`mmrm`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r">#' \tab \eqn{2}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">100</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... not used.</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\left[ \rho I(i \neq j)+I(i=j) \right]}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">101</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_get_contrast</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -18929,14393 +18437,14697 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">103</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">#' csh</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">104</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `Anova` will return `anova` object with one row per variable and columns</pre>
+                      <pre class="language-r">#' \tab Heterogeneous compound symmetry</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">105</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `Num Df`(numerator degrees of freedom), `Denom Df`(denominator degrees of freedom),</pre>
+                      <pre class="language-r">#' \tab \eqn{m+1}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">106</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `F Statistic` and `Pr(&gt;=F)`.</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\left[ \rho I(i \neq j)+I(i=j) \right]}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">107</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">108</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">109</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"># Please do not load `car` and then create the documentation. The Rd file will be different.</pre>
+                      <pre class="language-r">#' toep</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">110</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">Anova.mmrm &lt;- function(mod, type = c("II", "III", "2", "3"), tol = sqrt(.Machine$double.eps), ...) { # nolint</pre>
+                      <pre class="language-r">#' \tab Toeplitz</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">111</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_double(tol, finite = TRUE, len = 1L)</pre>
+                      <pre class="language-r">#' \tab \eqn{m}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">112</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma_{\left \vert {i-j} \right \vert +1}}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">113</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  vars &lt;- colnames(attr(terms(mod$formula_parts$model_formula), "factors"))</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">114</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret &lt;- lapply(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">115</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    vars,</pre>
+                      <pre class="language-r">#' toeph</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">116</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    function(x) df_md(mod, h_get_contrast(mod, x, type, tol))</pre>
+                      <pre class="language-r">#' \tab Heterogeneous Toeplitz</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">117</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' \tab \eqn{2m-1}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">118</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret_df &lt;- do.call(rbind.data.frame, ret)</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\rho_{\left \vert {i-j} \right \vert}}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">119</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  row.names(ret_df) &lt;- vars</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">120</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  colnames(ret_df) &lt;- c("Num Df", "Denom Df", "F Statistic", "Pr(&gt;=F)")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">121</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  class(ret_df) &lt;- c("anova", "data.frame")</pre>
+                      <pre class="language-r">#' us</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">122</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  attr(ret_df, "heading") &lt;- sprintf(</pre>
+                      <pre class="language-r">#' \tab Unstructured</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">123</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Analysis of Fixed Effect Table (Type %s F tests)",</pre>
+                      <pre class="language-r">#' \tab \eqn{m(m+1)/2}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">124</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    switch(type,</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma_{ij}}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">125</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "2" = ,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">126</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "II" = "II",</pre>
+                      <pre class="language-r">#' }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">127</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "3" = ,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">128</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "III" = "III"</pre>
+                      <pre class="language-r">#' where \eqn{i} and \eqn{j} denote \eqn{i}-th and \eqn{j}-th time points,</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">129</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' respectively, out of total \eqn{m} time points, \eqn{1 \leq i, j \leq m}.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">130</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">131</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret_df</pre>
+                      <pre class="language-r">#' @note The **ante-dependence** covariance structure in this package refers to</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">132</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' homogeneous ante-dependence, while the ante-dependence covariance structure</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">133</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' from SAS `PROC MIXED` refers to heterogeneous ante-dependence and the</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">134</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' homogeneous version is not available in SAS.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain Levels Prior and Posterior</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">136</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param var (`string`) name of the effect.</pre>
+                      <pre class="language-r">#' @note For all non-spatial covariance structures, the time variable must</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">137</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param additional_vars (`character`) names of additional variables.</pre>
+                      <pre class="language-r">#' be coded as a factor.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">138</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param xlev (`list`) named list of character levels.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">139</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param factors (`matrix`) the factor matrix.</pre>
+                      <pre class="language-r">#' ## Spatial Covariance structures:</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">140</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">141</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_obtain_lvls &lt;- function(var, additional_vars, xlev, factors) {</pre>
+                      <pre class="language-r">#' \tabular{clll}{</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">142</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(var)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">143</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_character(additional_vars)</pre>
+                      <pre class="language-r">#' \strong{Structure}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">144</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(xlev, types = "character")</pre>
+                      <pre class="language-r">#' \tab \strong{Description}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">145</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  nms &lt;- names(xlev)</pre>
+                      <pre class="language-r">#' \tab \strong{Parameters}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">146</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_subset(additional_vars, nms)</pre>
+                      <pre class="language-r">#' \tab \strong{\eqn{(i, j)} element}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">147</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (var %in% nms) {</pre>
+                      <pre class="language-r">#' \cr</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">148</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    prior_vars &lt;- intersect(nms[seq_len(match(var, nms) - 1)], additional_vars)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">149</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    prior_lvls &lt;- vapply(xlev[prior_vars], length, FUN.VALUE = 1L)</pre>
+                      <pre class="language-r">#' sp_exp</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">150</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    post_vars &lt;- intersect(nms[seq(match(var, nms) + 1, length(nms))], additional_vars)</pre>
+                      <pre class="language-r">#' \tab spatial exponential</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">151</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    post_lvls &lt;- vapply(xlev[post_vars], length, FUN.VALUE = 1L)</pre>
+                      <pre class="language-r">#' \tab \eqn{2}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">152</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    total_lvls &lt;- prod(prior_lvls) * prod(post_lvls)</pre>
+                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\rho^{-d_{ij}}}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">153</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">154</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    prior_lvls &lt;- vapply(xlev[additional_vars], length, FUN.VALUE = 1L)</pre>
+                      <pre class="language-r">#' }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">155</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    post_lvls &lt;- 2L</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">156</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    total_lvls &lt;- prod(prior_lvls)</pre>
+                      <pre class="language-r">#' where \eqn{d_{ij}} denotes the Euclidean distance between time points</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">157</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' \eqn{i} and \eqn{j}.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">158</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">159</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    prior = prior_lvls,</pre>
+                      <pre class="language-r">#' @family covariance types</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">160</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    post = post_lvls,</pre>
+                      <pre class="language-r">#' @name covariance_types</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">161</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    total = total_lvls</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">162</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">cov_types &lt;- function(</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">163</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    form = c("name", "abbr", "habbr"),</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">164</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    filter = c("heterogeneous", "spatial")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">165</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1666<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Check if the Effect is the First Categorical Effect</pre>
+                      <pre class="language-r">  form &lt;- match.arg(form, several.ok = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">166</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1666<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param effect (`string`) name of the effect.</pre>
+                      <pre class="language-r">  filter &lt;- if (missing(filter)) c() else match.arg(filter, several.ok = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">167</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1666<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param categorical (`character`) names of the categorical values.</pre>
+                      <pre class="language-r">  df &lt;- COV_TYPES[form][rowSums(!COV_TYPES[filter]) == 0, ]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">168</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">1666<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param factors (`matrix`) the factor matrix.</pre>
+                      <pre class="language-r">  Filter(Negate(is.na), unlist(t(df), use.names = FALSE))</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">169</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">170</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_first_contain_categorical &lt;- function(effect, factors, categorical) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">171</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(effect)</pre>
+                      <pre class="language-r">#' Retrieve Associated Abbreviated Covariance Structure Type Name</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">172</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(factors)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">173</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_character(categorical)</pre>
+                      <pre class="language-r">#' @param type (`string`)\cr either a full name or abbreviate covariance</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">174</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  mt &lt;- match(effect, colnames(factors))</pre>
+                      <pre class="language-r">#'   structure type name to collapse into an abbreviated type.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">175</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  varnms &lt;- row.names(factors)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">176</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # if the effect is not categorical in any value, return FALSE</pre>
+                      <pre class="language-r">#' @return The corresponding abbreviated covariance type name.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">177</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!any(varnms[factors[, mt] &gt; 0] %in% categorical)) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">178</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(FALSE)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">179</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">cov_type_abbr &lt;- function(type) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">180</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">299<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # keep only categorical rows that is in front of the current factor</pre>
+                      <pre class="language-r">  row &lt;- which(COV_TYPES == type, arr.ind = TRUE)[, 1]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">181</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="coverage">299<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  factors &lt;- factors[row.names(factors) %in% categorical, seq_len(mt - 1L), drop = FALSE]</pre>
+                      <pre class="language-r">  COV_TYPES$abbr[row]</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">182</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # if previous cols are all numerical, return TRUE</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">183</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (ncol(factors) &lt; 1L) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">184</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(TRUE)</pre>
+                      <pre class="language-r">#' Retrieve Associated Full Covariance Structure Type Name</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">185</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">186</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  col_ind &lt;- apply(factors, 2, prod)</pre>
+                      <pre class="language-r">#' @param type (`string`)\cr either a full name or abbreviate covariance</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">187</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # if any of the previous cols are categorical, return FALSE</pre>
+                      <pre class="language-r">#'   structure type name to convert to a long-form type.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">188</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  return(!any(col_ind &gt; 0))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">189</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @return The corresponding abbreviated covariance type name.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">190</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">191</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Test if the First Vector is Subset of the Second Vector</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">192</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`vector`) the first list.</pre>
+                      <pre class="language-r">cov_type_name &lt;- function(type) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">193</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param y (`vector`) the second list.</pre>
+                      <pre class="language-r">  row &lt;- which(COV_TYPES == type, arr.ind = TRUE)[, 1]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">194</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  COV_TYPES$name[row]</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">195</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_index &lt;- function(x, y) {</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">196</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(x)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">197</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(y)</pre>
+                      <pre class="language-r">#' Produce A Covariance Identifier Passing to TMB</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">198</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  vapply(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">199</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x,</pre>
+                      <pre class="language-r">#' @param cov (`cov_struct`)\cr a covariance structure object.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">200</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    \(i) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">201</td>
-                    <td class="coverage">68<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      r &lt;- vapply(y, \(j) test_subset(j, i), FUN.VALUE = TRUE)</pre>
+                      <pre class="language-r">#' @return A string used for method dispatch when passed to TMB.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">202</td>
-                    <td class="coverage">68<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (sum(r) == 1L) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">203</td>
-                    <td class="coverage">65<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        which(r)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">204</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else {</pre>
+                      <pre class="language-r">tmb_cov_type &lt;- function(cov) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">205</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage">266<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        NA_integer_</pre>
+                      <pre class="language-r">  paste0(cov$type, if (cov$heterogeneous) "h")</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">206</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      }</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">207</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    },</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">208</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    FUN.VALUE = 1L</pre>
+                      <pre class="language-r">#' Define a Covariance Structure</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">209</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">210</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/kenwardroger.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">211</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain Kenward-Roger Adjustment Components</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">212</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param type (`string`)\cr the name of the covariance structure type to use.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">213</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtains the components needed downstream for the computation of Kenward-Roger degrees of freedom.</pre>
+                      <pre class="language-r">#'   For available options, see `cov_types()`. If a type abbreviation is used</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">214</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Used in [mmrm()] fitting if method is "Kenward-Roger".</pre>
+                      <pre class="language-r">#'   that implies heterogeneity (e.g. `cph`) and no value is provided to</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">215</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   `heterogeneous`, then the heterogeneity is derived from the type name.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">216</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
+                      <pre class="language-r">#' @param visits (`character`)\cr a vector of variable names to use for the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">217</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param theta (`numeric`)\cr theta estimate.</pre>
+                      <pre class="language-r">#'   longitudinal terms of the covariance structure. Multiple terms are only</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">218</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   permitted for the `"spatial"` covariance type.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">219</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details the function returns a named list, \eqn{P}, \eqn{Q} and \eqn{R}, which corresponds to the</pre>
+                      <pre class="language-r">#' @param subject (`string`)\cr the name of the variable that encodes a subject</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">220</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' paper in 1997. The matrices are stacked in columns so that \eqn{P}, \eqn{Q} and \eqn{R} has the same</pre>
+                      <pre class="language-r">#'   identifier.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">221</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' column number(number of beta parameters). The number of rows, is dependent on</pre>
+                      <pre class="language-r">#' @param group (`string`)\cr optionally, the name of the variable that encodes</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">222</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' the total number of theta and number of groups, if the fit is a grouped mmrm.</pre>
+                      <pre class="language-r">#'   a grouping variable for subjects.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">13</td>
+                    <td class="num">223</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' For \eqn{P} matrix, it is stacked sequentially. For \eqn{Q} and \eqn{R} matrix, it is stacked so</pre>
+                      <pre class="language-r">#' @param heterogeneous (`flag`)\cr</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">14</td>
+                    <td class="num">224</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' that the \eqn{Q_{ij}} and \eqn{R_{ij}} is stacked from \eqn{j} then to \eqn{i}, i.e. \eqn{R_{i1}}, \eqn{R_{i2}}, etc.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">15</td>
+                    <td class="num">225</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \eqn{Q} and \eqn{R} only contains intra-group results and inter-group results should be all zero matrices</pre>
+                      <pre class="language-r">#' @return A `cov_struct` object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">226</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' so they are not stacked in the result.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">17</td>
+                    <td class="num">227</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">18</td>
+                    <td class="num">228</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Named list with elements:</pre>
+                      <pre class="language-r">#' cov_struct("csh", "AVISITN", "USUBJID")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">19</td>
+                    <td class="num">229</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `P`: `matrix` of \eqn{P} component.</pre>
+                      <pre class="language-r">#' cov_struct("spatial", c("VISITA", "VISITB"), group = "GRP", subject = "SBJ")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">20</td>
+                    <td class="num">230</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `Q`: `matrix` of \eqn{Q} component.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">21</td>
+                    <td class="num">231</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `R`: `matrix` of \eqn{R} component.</pre>
+                      <pre class="language-r">#' @family covariance types</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">22</td>
+                    <td class="num">232</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">23</td>
+                    <td class="num">233</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">cov_struct &lt;- function(</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">24</td>
+                    <td class="num">234</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_get_kr_comp &lt;- function(tmb_data, theta) {</pre>
+                      <pre class="language-r">    type = cov_types(), visits, subject, group = character(),</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">235</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    heterogeneous = FALSE) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">236</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  # if heterogeneous isn't provided, derive from provided type</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">25</td>
-                    <td class="coverage">47<em>x</em></td>
+                    <td class="num">237</td>
+                    <td class="coverage">296<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
+                      <pre class="language-r">  if (missing(heterogeneous)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">26</td>
-                    <td class="coverage">47<em>x</em></td>
+                    <td class="num">238</td>
+                    <td class="coverage">294<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(theta, "numeric")</pre>
+                      <pre class="language-r">    heterogeneous &lt;- switch(type,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">27</td>
-                    <td class="coverage">47<em>x</em></td>
+                    <td class="num">239</td>
+                    <td class="coverage">294<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  .Call(`_mmrm_get_pqr`, PACKAGE = "mmrm", tmb_data, theta)</pre>
+                      <pre class="language-r">      toeph = ,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">28</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">240</td>
+                    <td class="coverage">294<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      ar1h = ,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">29</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">241</td>
+                    <td class="coverage">294<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      adh = ,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">30</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">242</td>
+                    <td class="coverage">294<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Kenward-Roger Degrees of Freedom for Multi-Dimensional Contrast</pre>
+                      <pre class="language-r">      csh = TRUE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">31</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">243</td>
+                    <td class="coverage">294<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      heterogeneous</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">32</td>
+                    <td class="num">244</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_md()] if method is "Kenward-Roger" or "Kenward-Roger-Linear".</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">33</td>
+                    <td class="num">245</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">34</td>
+                    <td class="num">246</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_df_md_sat</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">35</td>
+                    <td class="num">247</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inherit h_df_md_sat return</pre>
+                      <pre class="language-r">  # coerce all type options into abbreviated form</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">36</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">248</td>
+                    <td class="coverage">296<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">249</td>
+                    <td class="coverage">295<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  type &lt;- cov_type_abbr(type)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">37</td>
+                    <td class="num">250</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_md_kr &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">38</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">251</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">  x &lt;- structure(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">39</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">252</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">    list(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">40</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">253</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (component(object, "reml") != 1) {</pre>
+                      <pre class="language-r">      type = type,</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">41</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">254</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("Kenward-Roger is only for REML")</pre>
+                      <pre class="language-r">      heterogeneous = heterogeneous,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">42</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">255</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">      visits = visits,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">43</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">256</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  kr_comp &lt;- object$kr_comp</pre>
+                      <pre class="language-r">      subject = subject,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">44</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">257</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  w &lt;- component(object, "theta_vcov")</pre>
+                      <pre class="language-r">      group = group</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">258</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    ),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">45</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">259</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  v_adj &lt;- object$beta_vcov_adj</pre>
+                      <pre class="language-r">    class = c("cov_struct", "mmrm_cov_struct", "list")</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">46</td>
-                    <td class="coverage">6<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">260</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- h_kr_df(v0 = object$beta_vcov, l = contrast, w = w, p = kr_comp$P)</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">47</td>
+                    <td class="num">261</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">48</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">262</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_test_md(object, contrast, df = df$m, f_stat_factor = df$lambda)</pre>
+                      <pre class="language-r">  validate_cov_struct(x)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">49</td>
+                    <td class="num">263</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">264</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">51</td>
+                    <td class="num">265</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Kenward-Roger Degrees of Freedom for One-Dimensional Contrast</pre>
+                      <pre class="language-r">#' Reconcile Possible Covariance Structure Inputs</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">52</td>
+                    <td class="num">266</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">53</td>
+                    <td class="num">267</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_1d()] if method is</pre>
+                      <pre class="language-r">#' @inheritParams mmrm</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">268</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' "Kenward-Roger" or "Kenward-Roger-Linear".</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">55</td>
+                    <td class="num">269</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @return The value `covariance` if it's provided or a covariance structure</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">56</td>
+                    <td class="num">270</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_df_1d_sat</pre>
+                      <pre class="language-r">#'   derived from the provided `formula` otherwise. An error is raised of both</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">57</td>
+                    <td class="num">271</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inherit h_df_1d_sat return</pre>
+                      <pre class="language-r">#'   are provided.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">58</td>
+                    <td class="num">272</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">59</td>
+                    <td class="num">273</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_1d_kr &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">60</td>
-                    <td class="coverage">21<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">274</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">h_reconcile_cov_struct &lt;- function(formula = NULL, covariance = NULL) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">61</td>
-                    <td class="coverage">21<em>x</em></td>
+                    <td class="num">275</td>
+                    <td class="coverage">238<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">  assert_multi_class(covariance, c("formula", "cov_struct"), null.ok = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">62</td>
-                    <td class="coverage">21<em>x</em></td>
+                    <td class="num">276</td>
+                    <td class="coverage">238<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (component(object, "reml") != 1) {</pre>
+                      <pre class="language-r">  assert_formula(formula, null.ok = FALSE)</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">63</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">277</td>
+                    <td class="coverage">238<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("Kenward-Roger is only for REML!")</pre>
+                      <pre class="language-r">  if (inherits(covariance, "formula")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">64</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">278</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    covariance &lt;- as.cov_struct(covariance)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">65</td>
+                    <td class="num">279</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">66</td>
-                    <td class="coverage">21<em>x</em></td>
+                    <td class="num">280</td>
+                    <td class="coverage">238<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- h_kr_df(</pre>
+                      <pre class="language-r">  if (!is.null(covariance) &amp;&amp; length(h_extract_covariance_terms(formula)) &gt; 0) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">67</td>
-                    <td class="coverage">21<em>x</em></td>
+                    <td class="num">281</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    v0 = object$beta_vcov,</pre>
+                      <pre class="language-r">    stop(paste0(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">68</td>
-                    <td class="coverage">21<em>x</em></td>
+                    <td class="num">282</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    l = matrix(contrast, nrow = 1),</pre>
+                      <pre class="language-r">      "Redundant covariance structure definition in `formula` and ",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">69</td>
-                    <td class="coverage">21<em>x</em></td>
+                    <td class="num">283</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    w = component(object, "theta_vcov"),</pre>
+                      <pre class="language-r">      "`covariance` arguments"</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">70</td>
-                    <td class="coverage">21<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">284</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    p = object$kr_comp$P</pre>
+                      <pre class="language-r">    ))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">71</td>
+                    <td class="num">285</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">72</td>
+                    <td class="num">286</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">73</td>
-                    <td class="coverage">21<em>x</em></td>
+                    <td class="num">287</td>
+                    <td class="coverage">236<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_test_1d(object, contrast, df$m)</pre>
+                      <pre class="language-r">  if (!is.null(covariance)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">74</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">288</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    return(covariance)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">75</td>
+                    <td class="num">289</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">76</td>
+                    <td class="num">290</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain the Adjusted Kenward-Roger degrees of freedom</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">77</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">291</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  as.cov_struct(formula, warn_partial = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">78</td>
+                    <td class="num">292</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtains the adjusted Kenward-Roger degrees of freedom and F statistic scale parameter.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">79</td>
+                    <td class="num">293</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Used in [h_df_md_kr()] or [h_df_1d_kr].</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">80</td>
+                    <td class="num">294</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Validate Covariance Structure Data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">81</td>
+                    <td class="num">295</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param v0 (`matrix`)\cr unadjusted covariance matrix.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">82</td>
+                    <td class="num">296</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param l (`matrix`)\cr linear combination matrix.</pre>
+                      <pre class="language-r">#' Run checks against relational integrity of covariance definition</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">83</td>
+                    <td class="num">297</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param w (`matrix`)\cr hessian matrix.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">84</td>
+                    <td class="num">298</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param p (`matrix`)\cr P matrix from [h_get_kr_comp()].</pre>
+                      <pre class="language-r">#' @param x (`cov_struct`)\cr a covariance structure object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">85</td>
+                    <td class="num">299</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">86</td>
+                    <td class="num">300</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Named list with elements:</pre>
+                      <pre class="language-r">#' @return `x` if successful, or an error is thrown otherwise.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">87</td>
+                    <td class="num">301</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `m`: `numeric` degrees of freedom.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">88</td>
+                    <td class="num">302</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `lambda`: `numeric` F statistic scale parameter.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">89</td>
+                    <td class="num">303</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">validate_cov_struct &lt;- function(x) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">90</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">304</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  checks &lt;- checkmate::makeAssertCollection()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">91</td>
+                    <td class="num">305</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_kr_df &lt;- function(v0, l, w, p) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">306</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_beta &lt;- ncol(v0)</pre>
+                      <pre class="language-r">  with(x, {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">307</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(v0, ncols = n_beta, nrows = n_beta)</pre>
+                      <pre class="language-r">    assert_character(subject, len = 1, add = checks)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">94</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">308</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(l, ncols = n_beta)</pre>
+                      <pre class="language-r">    assert_logical(heterogeneous, len = 1, add = checks)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">95</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">309</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_theta &lt;- ncol(w)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">96</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">310</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(w, ncols = n_theta, nrows = n_theta)</pre>
+                      <pre class="language-r">    if (length(group) &gt; 1 || length(visits) &lt; 1) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">97</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">311</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_visits &lt;- ncol(p)</pre>
+                      <pre class="language-r">      checks$push(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">98</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">312</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(p, nrows = n_visits * n_theta)</pre>
+                      <pre class="language-r">        "Covariance structure must be of the form `time | (group /) subject`"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">99</td>
+                    <td class="num">313</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # see vignettes/kenward.Rmd#279</pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">100</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">314</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  slvol &lt;- solve(h_quad_form_mat(l, v0))</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">101</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">315</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  m &lt;- h_quad_form_mat(t(l), slvol)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">102</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">316</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  nl &lt;- nrow(l)</pre>
+                      <pre class="language-r">    if (!type %in% cov_types(filter = "spatial") &amp;&amp; length(visits) &gt; 1) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">103</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">317</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  mv0 &lt;- m %*% v0</pre>
+                      <pre class="language-r">      checks$push(paste(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">104</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">318</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  pl &lt;- lapply(seq_len(nrow(p) / ncol(p)), function(x) {</pre>
+                      <pre class="language-r">        "Non-spatial covariance structures must have a single longitudinal",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">105</td>
-                    <td class="coverage">108<em>x</em></td>
+                    <td class="num">319</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ii &lt;- (x - 1) * ncol(p) + 1</pre>
+                      <pre class="language-r">        "variable"</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">106</td>
-                    <td class="coverage">108<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">320</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    jj &lt;- x * ncol(p)</pre>
+                      <pre class="language-r">      ))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">107</td>
-                    <td class="coverage">108<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">321</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    p[ii:jj, ]</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">108</td>
+                    <td class="num">322</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  })</pre>
                     </td>
                   </tr>
+                  <tr class="never">
+                    <td class="num">323</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
                   <tr class="covered">
-                    <td class="num">109</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">324</td>
+                    <td class="coverage">295<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  mv0pv0 &lt;- lapply(pl, function(x) {</pre>
+                      <pre class="language-r">  reportAssertions(checks)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">110</td>
-                    <td class="coverage">108<em>x</em></td>
+                    <td class="num">325</td>
+                    <td class="coverage">289<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mv0 %*% x %*% v0</pre>
+                      <pre class="language-r">  x</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">111</td>
+                    <td class="num">326</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">112</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">327</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  a1 &lt;- 0</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">113</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">328</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  a2 &lt;- 0</pre>
+                      <pre class="language-r">#' Format Covariance Structure Object</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">114</td>
+                    <td class="num">329</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # see vignettes/kenward.Rmd#283</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">115</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">330</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  for (i in seq_len(length(pl))) {</pre>
+                      <pre class="language-r">#' @param x (`cov_struct`)\cr a covariance structure object.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">116</td>
-                    <td class="coverage">108<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">331</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    for (j in seq_len(length(pl))) {</pre>
+                      <pre class="language-r">#' @param ... Additional arguments unused.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">117</td>
-                    <td class="coverage">592<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">332</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      a1 &lt;- a1 + w[i, j] * h_tr(mv0pv0[[i]]) * h_tr(mv0pv0[[j]])</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">118</td>
-                    <td class="coverage">592<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">333</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      a2 &lt;- a2 + w[i, j] * h_tr(mv0pv0[[i]] %*% mv0pv0[[j]])</pre>
+                      <pre class="language-r">#' @return A formatted string for `x`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">119</td>
+                    <td class="num">334</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">120</td>
+                    <td class="num">335</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">121</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">336</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  b &lt;- 1 / (2 * nl) * (a1 + 6 * a2)</pre>
+                      <pre class="language-r">format.cov_struct &lt;- function(x, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">122</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">337</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  e &lt;- 1 + a2 / nl</pre>
+                      <pre class="language-r">  sprintf(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">123</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">338</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  e_star &lt;- 1 / (1 - a2 / nl)</pre>
+                      <pre class="language-r">    "&lt;covariance structure&gt;\n%s%s:\n\n  %s | %s%s\n",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">124</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">339</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  g &lt;- ((nl + 1) * a1 - (nl + 4) * a2) / ((nl + 2) * a2)</pre>
+                      <pre class="language-r">    if (x$heterogeneous) "heterogeneous " else "",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">125</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">340</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  denom &lt;- (3 * nl + 2 - 2 * g)</pre>
+                      <pre class="language-r">    cov_type_name(x$type),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">126</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">341</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  c1 &lt;- g / denom</pre>
+                      <pre class="language-r">    format_symbols(x$visits),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">127</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">342</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  c2 &lt;- (nl - g) / denom</pre>
+                      <pre class="language-r">    if (length(x$group) &gt; 0) paste0(format_symbols(x$group), " / ") else "",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">128</td>
-                    <td class="coverage">28<em>x</em></td>
+                    <td class="num">343</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  c3 &lt;- (nl + 2 - g) / denom</pre>
+                      <pre class="language-r">    format_symbols(x$subject)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">129</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">344</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  v_star &lt;- 2 / nl * (1 + c1 * b) / (1 - c2 * b)^2 / (1 - c3 * b)</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">130</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">345</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  rho &lt;- v_star / (2 * e_star^2)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">131</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">346</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  m &lt;- 4 + (nl + 2) / (nl * rho - 1)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">132</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">347</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  lambda &lt;- m / (e_star * (m - 2))</pre>
+                      <pre class="language-r">#' Print a Covariance Structure Object</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">133</td>
-                    <td class="coverage">28<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">348</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(m = m, lambda = lambda)</pre>
+                      <pre class="language-r">#'</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">349</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @param x (`cov_struct`)\cr a covariance structure object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">134</td>
+                    <td class="num">350</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @param ... Additional arguments unused.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">135</td>
+                    <td class="num">351</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">136</td>
+                    <td class="num">352</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain the Adjusted Covariance Matrix</pre>
+                      <pre class="language-r">#' @return `x` invisibly.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">137</td>
+                    <td class="num">353</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">138</td>
+                    <td class="num">354</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtains the Kenward-Roger adjusted covariance matrix for the</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">139</td>
+                    <td class="num">355</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   coefficient estimates.</pre>
+                      <pre class="language-r">print.cov_struct &lt;- function(x, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">140</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">356</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Used in [mmrm()] fitting if method is "Kenward-Roger" or "Kenward-Roger-Linear".</pre>
+                      <pre class="language-r">  cat(format(x, ...), "\n")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">357</td>
+                    <td class="coverage">3<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  invisible(x)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">141</td>
+                    <td class="num">358</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">142</td>
+                    <td class="num">359</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param v (`matrix`)\cr unadjusted covariance matrix.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">143</td>
+                    <td class="num">360</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param w (`matrix`)\cr hessian matrix.</pre>
+                      <pre class="language-r">#' Coerce into a Covariance Structure Definition</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">144</td>
+                    <td class="num">361</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param p (`matrix`)\cr P matrix from [h_get_kr_comp()].</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">145</td>
+                    <td class="num">362</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param q (`matrix`)\cr Q matrix from [h_get_kr_comp()].</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">146</td>
+                    <td class="num">363</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param r (`matrix`)\cr R matrix from [h_get_kr_comp()].</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">147</td>
+                    <td class="num">364</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param linear (`flag`)\cr whether to use linear Kenward-Roger approximation.</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">148</td>
+                    <td class="num">365</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' A covariance structure can be parsed from a model definition formula or call.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">149</td>
+                    <td class="num">366</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The matrix of adjusted covariance matrix.</pre>
+                      <pre class="language-r">#' Generally, covariance structures defined using non-standard evaluation take</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">150</td>
+                    <td class="num">367</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' the following form:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">151</td>
+                    <td class="num">368</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">152</td>
+                    <td class="num">369</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_var_adj &lt;- function(v, w, p, q, r, linear = FALSE) {</pre>
+                      <pre class="language-r">#' ```</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">153</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">370</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(linear)</pre>
+                      <pre class="language-r">#' type( (visit, )* visit | (group /)? subject )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">154</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">371</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_beta &lt;- ncol(v)</pre>
+                      <pre class="language-r">#' ```</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">155</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">372</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(v, nrows = n_beta)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">156</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">373</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_theta &lt;- ncol(w)</pre>
+                      <pre class="language-r">#' For example, formulas may include terms such as</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">157</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">374</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(w, nrows = n_theta)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">158</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">375</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_visits &lt;- ncol(p)</pre>
+                      <pre class="language-r">#' ```r</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">159</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">376</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  theta_per_group &lt;- nrow(q) / nrow(p)</pre>
+                      <pre class="language-r">#' us(time | subject)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">160</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">377</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_groups &lt;- n_theta / theta_per_group</pre>
+                      <pre class="language-r">#' cp(time | group / subject)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">161</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">378</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(p, nrows = n_theta * n_visits)</pre>
+                      <pre class="language-r">#' sp_exp(coord1, coord2 | group / subject)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">162</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">379</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(q, nrows = theta_per_group^2 * n_groups * n_visits, ncols = n_visits)</pre>
+                      <pre class="language-r">#' ```</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">163</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">380</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(r, nrows = theta_per_group^2 * n_groups * n_visits, ncols = n_visits)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">164</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">381</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (linear) {</pre>
+                      <pre class="language-r">#' Note that only `sp_exp` (spatial) covariance structures may provide multiple</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">165</td>
-                    <td class="coverage">13<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">382</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    r &lt;- matrix(0, nrow = nrow(r), ncol = ncol(r))</pre>
+                      <pre class="language-r">#' coordinates, which identify the Euclidean distance between the time points.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">166</td>
+                    <td class="num">383</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">167</td>
+                    <td class="num">384</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @param x an object from which to derive a covariance structure. See object</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">168</td>
+                    <td class="num">385</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # see vignettes/kenward.Rmd#131</pre>
+                      <pre class="language-r">#'   specific sections for details.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">169</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">386</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret &lt;- v</pre>
+                      <pre class="language-r">#' @param warn_partial (`flag`)\cr whether to emit a warning when parts of the</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">170</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">387</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  for (i in seq_len(n_theta)) {</pre>
+                      <pre class="language-r">#'   formula are disregarded.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">171</td>
-                    <td class="coverage">264<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">388</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    for (j in seq_len(n_theta)) {</pre>
+                      <pre class="language-r">#' @param ... additional arguments unused.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">172</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">389</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      gi &lt;- ceiling(i / theta_per_group)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">173</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">390</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      gj &lt;- ceiling(j / theta_per_group)</pre>
+                      <pre class="language-r">#' @return A [cov_struct()] object.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">174</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">391</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      iid &lt;- (i - 1) * n_beta + 1</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">175</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">392</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      jid &lt;- (j - 1) * n_beta + 1</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">176</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">393</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ii &lt;- i - (gi - 1) * theta_per_group</pre>
+                      <pre class="language-r">#' # provide a covariance structure as a right-sided formula</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">177</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">394</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      jj &lt;- j - (gi - 1) * theta_per_group</pre>
+                      <pre class="language-r">#' as.cov_struct(~ csh(visit | group / subject))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">178</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">395</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ijid &lt;- ((ii - 1) * theta_per_group + jj - 1) * n_beta + (gi - 1) * n_beta * theta_per_group^2 + 1</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">179</td>
-                    <td class="coverage">2164<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">396</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (gi != gj) {</pre>
+                      <pre class="language-r">#' # when part of a full formula, suppress warnings using `warn_partial = FALSE`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">180</td>
-                    <td class="coverage">592<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">397</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        ret &lt;- ret + 2 * w[i, j] * v %*% (-p[iid:(iid + n_beta - 1), ] %*% v %*% p[jid:(jid + n_beta - 1), ]) %*% v</pre>
+                      <pre class="language-r">#' as.cov_struct(y ~ x + csh(visit | group / subject), warn_partial = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">181</td>
+                    <td class="num">398</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">182</td>
-                    <td class="coverage">1572<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">399</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        ret &lt;- ret + 2 * w[i, j] * v %*% (</pre>
+                      <pre class="language-r">#' @family covariance types</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">183</td>
-                    <td class="coverage">1572<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">400</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          q[ijid:(ijid + n_beta - 1), ] -</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">184</td>
-                    <td class="coverage">1572<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">401</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">            p[iid:(iid + n_beta - 1), ] %*% v %*% p[jid:(jid + n_beta - 1), ] -</pre>
+                      <pre class="language-r">as.cov_struct &lt;- function(x, ...) { # nolint</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">185</td>
-                    <td class="coverage">1572<em>x</em></td>
+                    <td class="num">402</td>
+                    <td class="coverage">278<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">            1 / 4 * r[ijid:(ijid + n_beta - 1), ]</pre>
+                      <pre class="language-r">  UseMethod("as.cov_struct")</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">186</td>
-                    <td class="coverage">1572<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">403</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        ) %*% v</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">187</td>
+                    <td class="num">404</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">188</td>
+                    <td class="num">405</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">189</td>
+                    <td class="num">406</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">as.cov_struct.cov_struct &lt;- function(x, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">190</td>
-                    <td class="coverage">49<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">407</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ret</pre>
+                      <pre class="language-r">  x</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">191</td>
+                    <td class="num">408</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/testing.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">409</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Degrees of Freedom for One-Dimensional Contrast</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">410</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @describeIn as.cov_struct</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">411</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">#' When provided a formula, any specialized functions are assumed to be</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">412</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculates the estimate, adjusted standard error, degrees of freedom,</pre>
+                      <pre class="language-r">#' covariance structure definitions and must follow the form:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">413</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' t statistic and p-value for one-dimensional contrast.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">414</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' ```</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">415</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
+                      <pre class="language-r">#' y ~ xs + type( (visit, )* visit | (group /)? subject )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">416</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param contrast (`numeric`)\cr contrast vector. Note that this should not include</pre>
+                      <pre class="language-r">#' ```</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">417</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">418</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   actually estimated coefficients.</pre>
+                      <pre class="language-r">#' Any component on the right hand side of a formula is considered when</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">419</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with `est`, `se`, `df`, `t_stat` and `p_val`.</pre>
+                      <pre class="language-r">#' searching for a covariance definition.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">420</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">13</td>
+                    <td class="num">421</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">14</td>
+                    <td class="num">422</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">as.cov_struct.formula &lt;- function(x, warn_partial = TRUE, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">15</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">423</td>
+                    <td class="coverage">278<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' object &lt;- mmrm(</pre>
+                      <pre class="language-r">  x_calls &lt;- h_extract_covariance_terms(x)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">424</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">17</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">425</td>
+                    <td class="coverage">278<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   data = fev_data</pre>
+                      <pre class="language-r">  if (length(x_calls) &lt; 1) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">18</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">426</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' )</pre>
+                      <pre class="language-r">    stop(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">19</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">427</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' contrast &lt;- numeric(length(object$beta_est))</pre>
+                      <pre class="language-r">      "Covariance structure must be specified in formula. ",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">20</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">428</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' contrast[3] &lt;- 1</pre>
+                      <pre class="language-r">      "Possible covariance structures include: ",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">21</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">429</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' df_1d(object, contrast)</pre>
+                      <pre class="language-r">      paste0(cov_types(c("abbr", "habbr")), collapse = ", ")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">22</td>
+                    <td class="num">430</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">df_1d &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">23</td>
-                    <td class="coverage">338<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">431</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">24</td>
-                    <td class="coverage">338<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">432</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")), any.missing = FALSE)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">25</td>
-                    <td class="coverage">338<em>x</em></td>
+                    <td class="num">433</td>
+                    <td class="coverage">274<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  contrast &lt;- as.vector(contrast)</pre>
+                      <pre class="language-r">  if (length(x_calls) &gt; 1) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">26</td>
-                    <td class="coverage">338<em>x</em></td>
+                    <td class="num">434</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  switch(object$method,</pre>
+                      <pre class="language-r">    cov_struct_types &lt;- as.character(lapply(x_calls, `[[`, 1L))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">27</td>
-                    <td class="coverage">318<em>x</em></td>
+                    <td class="num">435</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Satterthwaite" = h_df_1d_sat(object, contrast),</pre>
+                      <pre class="language-r">    stop(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">28</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">436</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Kenward-Roger" = h_df_1d_kr(object, contrast),</pre>
+                      <pre class="language-r">      "Only one covariance structure can be specified. ",</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">29</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">437</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Residual" = h_df_1d_res(object, contrast),</pre>
+                      <pre class="language-r">      "Currently specified covariance structures are: ",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">30</td>
+                    <td class="num">438</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Between-Within" = h_df_1d_bw(object, contrast),</pre>
+                      <pre class="language-r">      paste0(cov_struct_types, collapse = ", ")</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">31</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">439</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("Unrecognized degrees of freedom method: ", object$method)</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">32</td>
+                    <td class="num">440</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">33</td>
+                    <td class="num">441</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">34</td>
+                    <td class="num">442</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  # flatten into list of infix operators, calls and names/atomics</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">443</td>
+                    <td class="coverage">273<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  x &lt;- flatten_call(x_calls[[1]])</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">444</td>
+                    <td class="coverage">273<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  type &lt;- as.character(x[[1]])</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">445</td>
+                    <td class="coverage">273<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  x &lt;- drop_elements(x, 1)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">35</td>
+                    <td class="num">446</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">36</td>
+                    <td class="num">447</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Degrees of Freedom for Multi-Dimensional Contrast</pre>
+                      <pre class="language-r">  # take visits until "|"</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">37</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">448</td>
+                    <td class="coverage">273<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  n &lt;- position_symbol(x, "|", nomatch = 0)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">38</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">449</td>
+                    <td class="coverage">273<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">  visits &lt;- as.character(utils::head(x, max(n - 1, 0)))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">39</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">450</td>
+                    <td class="coverage">273<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculates the estimate, standard error, degrees of freedom,</pre>
+                      <pre class="language-r">  x &lt;- drop_elements(x, n)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">40</td>
+                    <td class="num">451</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' t statistic and p-value for one-dimensional contrast, depending on the method</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">41</td>
+                    <td class="num">452</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' used in [mmrm()].</pre>
+                      <pre class="language-r">  # take group until "/"</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">42</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">453</td>
+                    <td class="coverage">273<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  n &lt;- position_symbol(x, "/", nomatch = 0)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">43</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">454</td>
+                    <td class="coverage">273<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
+                      <pre class="language-r">  group &lt;- as.character(utils::head(x, max(n - 1, 0)))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">44</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">455</td>
+                    <td class="coverage">273<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param contrast (`matrix`)\cr numeric contrast matrix, if given a `numeric`</pre>
+                      <pre class="language-r">  x &lt;- drop_elements(x, n)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">45</td>
+                    <td class="num">456</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   then this is coerced to a row vector. Note that this should not include</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">46</td>
+                    <td class="num">457</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
+                      <pre class="language-r">  # remainder is subject</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">47</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">458</td>
+                    <td class="coverage">273<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   actually estimated coefficients.</pre>
+                      <pre class="language-r">  subject &lt;- as.character(x)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">48</td>
+                    <td class="num">459</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">460</td>
+                    <td class="coverage">273<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cov_struct(type = type, visits = visits, group = group, subject = subject)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">49</td>
+                    <td class="num">461</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with `num_df`, `denom_df`, `f_stat` and `p_val` (2-sided p-value).</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/satterthwaite.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#' Obtain List of Jacobian Matrix Entries for Covariance Matrix</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">51</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">52</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @description Obtain the Jacobian matrices given the covariance function and variance parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">53</td>
+                    <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' object &lt;- mmrm(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
+                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">55</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   data = fev_data</pre>
+                      <pre class="language-r">#' @param theta_est (`numeric`)\cr variance parameters point estimate.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">56</td>
+                    <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' )</pre>
+                      <pre class="language-r">#' @param beta_vcov (`matrix`)\cr vairance covariance matrix of coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">57</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' contrast &lt;- matrix(data = 0, nrow = 2, ncol = length(object$beta_est))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">58</td>
+                    <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' contrast[1, 2] &lt;- contrast[2, 3] &lt;- 1</pre>
+                      <pre class="language-r">#' @return List with one element per variance parameter containing a matrix</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">59</td>
+                    <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' df_md(object, contrast)</pre>
+                      <pre class="language-r">#'   of the same dimensions as the covariance matrix. The values are the derivatives</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">60</td>
+                    <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">df_md &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">#'   with regards to this variance parameter.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">61</td>
-                    <td class="coverage">150<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">12</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">62</td>
-                    <td class="coverage">150<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">13</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast, any.missing = FALSE)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">63</td>
-                    <td class="coverage">150<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">14</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.matrix(contrast)) {</pre>
+                      <pre class="language-r">h_jac_list &lt;- function(tmb_data,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">64</td>
-                    <td class="coverage">113<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">15</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    contrast &lt;- matrix(contrast, ncol = length(contrast))</pre>
+                      <pre class="language-r">                       theta_est,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">65</td>
+                    <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">                       beta_vcov) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">66</td>
-                    <td class="coverage">150<em>x</em></td>
+                    <td class="num">17</td>
+                    <td class="coverage">82<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(contrast, ncols = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">67</td>
-                    <td class="coverage">150<em>x</em></td>
+                    <td class="num">18</td>
+                    <td class="coverage">82<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (nrow(contrast) == 0) {</pre>
+                      <pre class="language-r">  assert_numeric(theta_est)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">68</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">19</td>
+                    <td class="coverage">82<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(</pre>
+                      <pre class="language-r">  assert_matrix(beta_vcov)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">69</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">20</td>
+                    <td class="coverage">82<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      list(</pre>
+                      <pre class="language-r">  .Call(`_mmrm_get_jacobian`, PACKAGE = "mmrm", tmb_data, theta_est, beta_vcov)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">70</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">21</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        num_df = 0,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">71</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">22</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        denom_df = NA_real_,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">72</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">23</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        f_stat = NA_real_,</pre>
+                      <pre class="language-r">#' Quadratic Form Calculations</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">73</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">24</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        p_val = NA_real_</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">74</td>
+                    <td class="num">25</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">#' @description These helpers are mainly for easier readability and slightly better efficiency</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">75</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' of the quadratic forms used in the Satterthwaite calculations.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">76</td>
+                    <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">77</td>
-                    <td class="coverage">149<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">28</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  switch(object$method,</pre>
+                      <pre class="language-r">#' @param center (`matrix`)\cr square numeric matrix with the same dimensions as</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">78</td>
-                    <td class="coverage">145<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">29</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Satterthwaite" = h_df_md_sat(object, contrast),</pre>
+                      <pre class="language-r">#'   `x` as the center of the quadratic form.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">79</td>
-                    <td class="coverage">3<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">30</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Kenward-Roger" = h_df_md_kr(object, contrast),</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">80</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">31</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Residual" = h_df_md_res(object, contrast),</pre>
+                      <pre class="language-r">#' @name h_quad_form</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">81</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">32</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Between-Within" = h_df_md_bw(object, contrast),</pre>
+                      <pre class="language-r">NULL</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">82</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">33</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("Unrecognized degrees of freedom method: ", object$method)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">83</td>
+                    <td class="num">34</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' @describeIn h_quad_form calculates the number `vec %*% center %*% t(vec)`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">84</td>
+                    <td class="num">35</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'   as a numeric (not a matrix).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">85</td>
+                    <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">86</td>
+                    <td class="num">37</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Creating T-Statistic Test Results For One-Dimensional Contrast</pre>
+                      <pre class="language-r">#' @param vec (`numeric`)\cr interpreted as a row vector.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">87</td>
+                    <td class="num">38</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">88</td>
+                    <td class="num">39</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Creates a list of results for one-dimensional contrasts using</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">89</td>
+                    <td class="num">40</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' a t-test statistic and the given degrees of freedom.</pre>
+                      <pre class="language-r">h_quad_form_vec &lt;- function(vec, center) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">90</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">41</td>
+                    <td class="coverage">5607<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  vec &lt;- as.vector(vec)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">91</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">42</td>
+                    <td class="coverage">5607<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams df_1d</pre>
+                      <pre class="language-r">  assert_numeric(vec, any.missing = FALSE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">92</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">43</td>
+                    <td class="coverage">5607<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param df (`number`)\cr degrees of freedom for the one-dimensional contrast.</pre>
+                      <pre class="language-r">  assert_matrix(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">93</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">44</td>
+                    <td class="coverage">5607<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    center,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">94</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">45</td>
+                    <td class="coverage">5607<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with `est`, `se`, `df`, `t_stat` and `p_val` (2-sided p-value).</pre>
+                      <pre class="language-r">    mode = "numeric",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">95</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">46</td>
+                    <td class="coverage">5607<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    any.missing = FALSE,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">47</td>
+                    <td class="coverage">5607<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    nrows = length(vec),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">48</td>
+                    <td class="coverage">5607<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    ncols = length(vec)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">96</td>
+                    <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">97</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_test_1d &lt;- function(object,</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">51</td>
+                    <td class="coverage">5607<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  sum(vec * (center %*% vec))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">98</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      contrast,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">99</td>
+                    <td class="num">53</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      df) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">100</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">54</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">#' @describeIn h_quad_form calculates the quadratic form `mat %*% center %*% t(mat)`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">101</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">55</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">#'   as a matrix, the result is square and has dimensions identical to the number</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">102</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">56</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(df, lower = .Machine$double.xmin)</pre>
+                      <pre class="language-r">#'   of rows in `mat`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">103</td>
+                    <td class="num">57</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">104</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">58</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  est &lt;- sum(contrast * component(object, "beta_est"))</pre>
+                      <pre class="language-r">#' @param mat (`matrix`)\cr numeric matrix to be multiplied left and right of</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">105</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">59</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  var &lt;- h_quad_form_vec(contrast, component(object, "beta_vcov"))</pre>
+                      <pre class="language-r">#'   `center`, therefore needs to have as many columns as there are rows and columns</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">106</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">60</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  se &lt;- sqrt(var)</pre>
+                      <pre class="language-r">#'   in `center`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">107</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">61</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  t_stat &lt;- est / se</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">108</td>
-                    <td class="coverage">486<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">62</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  p_val &lt;- 2 * stats::pt(q = abs(t_stat), df = df, lower.tail = FALSE)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">109</td>
+                    <td class="num">63</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">h_quad_form_mat &lt;- function(mat, center) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">110</td>
-                    <td class="coverage">486<em>x</em></td>
+                    <td class="num">64</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">  assert_matrix(mat, mode = "numeric", any.missing = FALSE, min.cols = 1L)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">111</td>
-                    <td class="coverage">486<em>x</em></td>
+                    <td class="num">65</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    est = est,</pre>
+                      <pre class="language-r">  assert_matrix(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">112</td>
-                    <td class="coverage">486<em>x</em></td>
+                    <td class="num">66</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    se = se,</pre>
+                      <pre class="language-r">    center,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">113</td>
-                    <td class="coverage">486<em>x</em></td>
+                    <td class="num">67</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    df = df,</pre>
+                      <pre class="language-r">    mode = "numeric",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">114</td>
-                    <td class="coverage">486<em>x</em></td>
+                    <td class="num">68</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    t_stat = t_stat,</pre>
+                      <pre class="language-r">    any.missing = FALSE,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">115</td>
-                    <td class="coverage">486<em>x</em></td>
+                    <td class="num">69</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    p_val = p_val</pre>
+                      <pre class="language-r">    nrows = ncol(center),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">116</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">70</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    ncols = ncol(center)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">117</td>
+                    <td class="num">71</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">118</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">72</td>
+                    <td class="coverage">119<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  mat %*% tcrossprod(center, mat)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">119</td>
+                    <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Creating F-Statistic Test Results For Multi-Dimensional Contrast</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">120</td>
+                    <td class="num">74</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">121</td>
+                    <td class="num">75</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Creates a list of results for multi-dimensional contrasts using</pre>
+                      <pre class="language-r">#' Computation of a Gradient Given Jacobian and Contrast Vector</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">122</td>
+                    <td class="num">76</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' an F-test statistic and the given degrees of freedom.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">123</td>
+                    <td class="num">77</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @description Computes the gradient of a linear combination of `beta` given the Jacobian matrix and</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">124</td>
+                    <td class="num">78</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams df_md</pre>
+                      <pre class="language-r">#' variance parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">125</td>
+                    <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param contrast (`matrix`)\cr numeric contrast matrix.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">126</td>
+                    <td class="num">80</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param df (`number`)\cr denominator degrees of freedom for the multi-dimensional contrast.</pre>
+                      <pre class="language-r">#' @param jac_list (`list`)\cr Jacobian list produced e.g. by [h_jac_list()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">127</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param f_stat_factor (`number`)\cr optional scaling factor on top of the standard F-statistic.</pre>
+                      <pre class="language-r">#' @param contrast (`numeric`)\cr contrast vector, which needs to have the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">128</td>
+                    <td class="num">82</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   same number of elements as there are rows and columns in each element of</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">129</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with `num_df`, `denom_df`, `f_stat` and `p_val` (2-sided p-value).</pre>
+                      <pre class="language-r">#'   `jac_list`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">130</td>
+                    <td class="num">84</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">131</td>
+                    <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @return Numeric vector which contains the quadratic forms of each element of</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">132</td>
+                    <td class="num">86</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_test_md &lt;- function(object,</pre>
+                      <pre class="language-r">#'   `jac_list` with the `contrast` vector.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">133</td>
+                    <td class="num">87</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      contrast,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">134</td>
+                    <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      df,</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">135</td>
+                    <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      f_stat_factor = 1) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">136</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">137</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(contrast, ncols = length(component(object, "beta_est")))</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">138</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  num_df &lt;- nrow(contrast)</pre>
+                      <pre class="language-r">h_gradient &lt;- function(jac_list, contrast) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">139</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">90</td>
+                    <td class="coverage">491<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(df, lower = .Machine$double.xmin)</pre>
+                      <pre class="language-r">  assert_list(jac_list)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">140</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">91</td>
+                    <td class="coverage">491<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(f_stat_factor, lower = .Machine$double.xmin)</pre>
+                      <pre class="language-r">  assert_numeric(contrast)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">141</td>
+                    <td class="num">92</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">142</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  prec_contrast &lt;- solve(h_quad_form_mat(contrast, component(object, "beta_vcov")))</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">143</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  contrast_est &lt;- component(object, "beta_est") %*% t(contrast)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">144</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  f_statistic &lt;- as.numeric(f_stat_factor / num_df * h_quad_form_mat(contrast_est, prec_contrast))</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">145</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">93</td>
+                    <td class="coverage">491<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  p_val &lt;- stats::pf(</pre>
+                      <pre class="language-r">  vapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">146</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">94</td>
+                    <td class="coverage">491<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    q = f_statistic,</pre>
+                      <pre class="language-r">    jac_list,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">147</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">95</td>
+                    <td class="coverage">491<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    df1 = num_df,</pre>
+                      <pre class="language-r">    h_quad_form_vec,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">148</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">96</td>
+                    <td class="coverage">491<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    df2 = df,</pre>
+                      <pre class="language-r">    vec = contrast,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">149</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">97</td>
+                    <td class="coverage">491<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    lower.tail = FALSE</pre>
+                      <pre class="language-r">    numeric(1L)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">150</td>
+                    <td class="num">98</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">151</td>
+                    <td class="num">99</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">152</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">153</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    num_df = num_df,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">154</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    denom_df = df,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">155</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    f_stat = f_statistic,</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">156</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    p_val = p_val</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">157</td>
+                    <td class="num">100</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">158</td>
+                    <td class="num">101</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' Calculation of Satterthwaite Degrees of Freedom for One-Dimensional Contrast</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/cov_struct.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">102</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Covariance Type Database</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">103</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @description Used in [df_1d()] if method is</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">104</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' An internal constant for covariance type information.</pre>
+                      <pre class="language-r">#' "Satterthwaite".</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">105</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">106</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @format A data frame with 5 variables and one record per covariance type:</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">107</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param contrast (`numeric`)\cr contrast vector. Note that this should not include</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">108</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \describe{</pre>
+                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">109</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   \item{name}{</pre>
+                      <pre class="language-r">#'   actually estimated coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">110</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     The long-form name of the covariance structure type</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">111</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   }</pre>
+                      <pre class="language-r">#' @return List with `est`, `se`, `df`, `t_stat` and `p_val`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">112</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   \item{abbr}{</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">113</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     The abbreviated name of the covariance structure type</pre>
+                      <pre class="language-r">h_df_1d_sat &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">13</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">114</td>
+                    <td class="coverage">456<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   }</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">14</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">115</td>
+                    <td class="coverage">456<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   \item{habbr}{</pre>
+                      <pre class="language-r">  contrast &lt;- as.numeric(contrast)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">15</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">116</td>
+                    <td class="coverage">456<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     The abbreviated name of the heterogeneous version of a covariance</pre>
+                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">117</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     structure type (The abbreviated name (`abbr`) with a trailing `"h"` if</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">17</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">118</td>
+                    <td class="coverage">456<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     the structure has a heterogeneous implementation or `NA` otherwise).</pre>
+                      <pre class="language-r">  df &lt;- if (identical(object$vcov, "Asymptotic")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">18</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">119</td>
+                    <td class="coverage">444<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   }</pre>
+                      <pre class="language-r">    grad &lt;- h_gradient(component(object, "jac_list"), contrast)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">19</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">120</td>
+                    <td class="coverage">444<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   \item{heterogeneous}{</pre>
+                      <pre class="language-r">    v_num &lt;- 2 * h_quad_form_vec(contrast, component(object, "beta_vcov"))^2</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">20</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">121</td>
+                    <td class="coverage">444<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     A logical value indicating whether the covariance structure has a</pre>
+                      <pre class="language-r">    v_denom &lt;- h_quad_form_vec(grad, component(object, "theta_vcov"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">21</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">122</td>
+                    <td class="coverage">444<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     heterogeneous counterpart.</pre>
+                      <pre class="language-r">    v_num / v_denom</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">22</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">123</td>
+                    <td class="coverage">456<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   }</pre>
+                      <pre class="language-r">  } else if (object$vcov %in% c("Empirical", "Empirical-Jackknife", "Empirical-Bias-Reduced")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">23</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">124</td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   \item{spatial}{</pre>
+                      <pre class="language-r">    contrast_matrix &lt;- Matrix::.bdiag(rep(list(matrix(contrast, nrow = 1)), component(object, "n_subjects")))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">24</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">125</td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'     A logical value indicating whether the covariance structure is spatial.</pre>
+                      <pre class="language-r">    contrast_matrix &lt;- as.matrix(contrast_matrix)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">25</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">126</td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   }</pre>
+                      <pre class="language-r">    g_matrix &lt;- h_quad_form_mat(contrast_matrix, object$empirical_df_mat)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">26</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">127</td>
+                    <td class="coverage">12<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' }</pre>
+                      <pre class="language-r">    h_tr(g_matrix)^2 / sum(g_matrix^2)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">27</td>
+                    <td class="num">128</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">28</td>
+                    <td class="num">129</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">130</td>
+                    <td class="coverage">456<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  h_test_1d(object, contrast, df)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">29</td>
+                    <td class="num">131</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">COV_TYPES &lt;- local({ # nolint</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">30</td>
+                    <td class="num">132</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type &lt;- function(name, abbr, habbr, heterogeneous, spatial) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">31</td>
+                    <td class="num">133</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    args &lt;- as.list(match.call()[-1])</pre>
+                      <pre class="language-r">#' Calculating Denominator Degrees of Freedom for the Multi-Dimensional Case</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">32</td>
+                    <td class="num">134</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    do.call(data.frame, args)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">33</td>
+                    <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @description Calculates the degrees of freedom for multi-dimensional contrast.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">34</td>
+                    <td class="num">136</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">35</td>
+                    <td class="num">137</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  as.data.frame(</pre>
+                      <pre class="language-r">#' @param t_stat_df (`numeric`)\cr `n` t-statistic derived degrees of freedom.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">36</td>
+                    <td class="num">138</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    col.names = names(formals(type)),</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">37</td>
+                    <td class="num">139</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    rbind(</pre>
+                      <pre class="language-r">#' @return Usually the calculation is returning `2 * E / (E - n)` where</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">38</td>
+                    <td class="num">140</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      type("unstructured", "us", NA, FALSE, FALSE),</pre>
+                      <pre class="language-r">#'   `E` is the sum of `t / (t - 2)` over all `t_stat_df` values `t`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">39</td>
+                    <td class="num">141</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      type("Toeplitz", "toep", "toeph", TRUE, FALSE),</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">40</td>
+                    <td class="num">142</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      type("auto-regressive order one", "ar1", "ar1h", TRUE, FALSE),</pre>
+                      <pre class="language-r">#' @note If the input values are two similar to each other then just the average</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">41</td>
+                    <td class="num">143</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      type("ante-dependence", "ad", "adh", TRUE, FALSE),</pre>
+                      <pre class="language-r">#'   of them is returned. If any of the inputs is not larger than 2 then 2 is</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">42</td>
+                    <td class="num">144</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      type("compound symmetry", "cs", "csh", TRUE, FALSE),</pre>
+                      <pre class="language-r">#'   returned.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">43</td>
+                    <td class="num">145</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      type("spatial exponential", "sp_exp", NA, FALSE, TRUE)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">44</td>
+                    <td class="num">146</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">45</td>
+                    <td class="num">147</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">h_md_denom_df &lt;- function(t_stat_df) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">46</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">148</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">})</pre>
+                      <pre class="language-r">  assert_numeric(t_stat_df, min.len = 1L, lower = .Machine$double.xmin, any.missing = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">47</td>
+                    <td class="num">149</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">48</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">150</td>
+                    <td class="coverage">24<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Covariance Types</pre>
+                      <pre class="language-r">  if (test_scalar(t_stat_df)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">49</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">151</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    t_stat_df</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">50</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">152</td>
+                    <td class="coverage">23<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">  } else if (all(abs(diff(t_stat_df)) &lt; sqrt(.Machine$double.eps))) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">51</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">153</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    mean(t_stat_df)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">52</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">154</td>
+                    <td class="coverage">22<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param form (`character`)\cr covariance structure type name form. One or</pre>
+                      <pre class="language-r">  } else if (any(t_stat_df &lt;= 2)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">53</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">155</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   more of `"name"`, `"abbr"` (abbreviation), or `"habbr"` (heterogeneous</pre>
+                      <pre class="language-r">    2</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">156</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   abbreviation).</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">55</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">157</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param filter (`character`)\cr covariance structure type filter. One or</pre>
+                      <pre class="language-r">    e &lt;- sum(t_stat_df / (t_stat_df - 2))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">56</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">158</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   more of `"heterogeneous"` or `"spatial"`.</pre>
+                      <pre class="language-r">    2 * e / (e - (length(t_stat_df)))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">57</td>
+                    <td class="num">159</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">58</td>
+                    <td class="num">160</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A character vector of accepted covariance structure type names and</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">59</td>
+                    <td class="num">161</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   abbreviations.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">60</td>
+                    <td class="num">162</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Creating F-Statistic Results from One-Dimensional Contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">61</td>
+                    <td class="num">163</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @section Abbreviations for Covariance Structures:</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">62</td>
+                    <td class="num">164</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @description Creates multi-dimensional result from one-dimensional contrast from [df_1d()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">63</td>
+                    <td class="num">165</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ## Common Covariance Structures:</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">64</td>
+                    <td class="num">166</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr model fit.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">65</td>
+                    <td class="num">167</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tabular{clll}{</pre>
+                      <pre class="language-r">#' @param contrast (`numeric`)\cr one-dimensional contrast.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">66</td>
+                    <td class="num">168</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">67</td>
+                    <td class="num">169</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \strong{Structure}</pre>
+                      <pre class="language-r">#' @return The one-dimensional degrees of freedom are calculated and then</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">68</td>
+                    <td class="num">170</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \strong{Description}</pre>
+                      <pre class="language-r">#'   based on that the p-value is calculated.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">69</td>
+                    <td class="num">171</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \strong{Parameters}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">70</td>
+                    <td class="num">172</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \strong{\eqn{(i, j)} element}</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">71</td>
+                    <td class="num">173</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">h_df_md_from_1d &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">72</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">174</td>
+                    <td class="coverage">134<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  res_1d &lt;- h_df_1d_sat(object, contrast)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">73</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">175</td>
+                    <td class="coverage">134<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ad</pre>
+                      <pre class="language-r">  list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">74</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">176</td>
+                    <td class="coverage">134<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Ante-dependence</pre>
+                      <pre class="language-r">    num_df = 1,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">75</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">177</td>
+                    <td class="coverage">134<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{m}</pre>
+                      <pre class="language-r">    denom_df = res_1d$df,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">76</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">178</td>
+                    <td class="coverage">134<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\prod_{k=i}^{j-1}\rho_{k}}</pre>
+                      <pre class="language-r">    f_stat = res_1d$t_stat^2,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">77</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">179</td>
+                    <td class="coverage">134<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">    p_val = stats::pf(q = res_1d$t_stat^2, df1 = 1, df2 = res_1d$df, lower.tail = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">78</td>
+                    <td class="num">180</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">79</td>
+                    <td class="num">181</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' adh</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">80</td>
+                    <td class="num">182</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Heterogeneous ante-dependence</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">81</td>
+                    <td class="num">183</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{2m-1}</pre>
+                      <pre class="language-r">#' Calculation of Satterthwaite Degrees of Freedom for Multi-Dimensional Contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">82</td>
+                    <td class="num">184</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\prod_{k=i}^{j-1}\rho_{k}}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">83</td>
+                    <td class="num">185</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">#' @description Used in [df_md()] if method is "Satterthwaite".</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">84</td>
+                    <td class="num">186</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">85</td>
+                    <td class="num">187</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ar1</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">86</td>
+                    <td class="num">188</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab First-order auto-regressive</pre>
+                      <pre class="language-r">#' @param contrast (`matrix`)\cr numeric contrast matrix, if given a `numeric`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">87</td>
+                    <td class="num">189</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{2}</pre>
+                      <pre class="language-r">#'   then this is coerced to a row vector. Note that this should not include</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">88</td>
+                    <td class="num">190</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\rho^{\left \vert {i-j} \right \vert}}</pre>
+                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">89</td>
+                    <td class="num">191</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">#'   actually estimated coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">90</td>
+                    <td class="num">192</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">91</td>
+                    <td class="num">193</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ar1h</pre>
+                      <pre class="language-r">#' @return List with `num_df`, `denom_df`, `f_stat` and `p_val` (2-sided p-value).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">92</td>
+                    <td class="num">194</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Heterogeneous first-order auto-regressive</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">93</td>
+                    <td class="num">195</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{m+1}</pre>
+                      <pre class="language-r">h_df_md_sat &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">94</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">196</td>
+                    <td class="coverage">151<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\rho^{\left \vert {i-j} \right \vert}}</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">95</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">197</td>
+                    <td class="coverage">151<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">96</td>
+                    <td class="num">198</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  # Early return if we are in the one-dimensional case.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">97</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">199</td>
+                    <td class="coverage">151<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' cs</pre>
+                      <pre class="language-r">  if (identical(nrow(contrast), 1L)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">98</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">200</td>
+                    <td class="coverage">132<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Compound symmetry</pre>
+                      <pre class="language-r">    return(h_df_md_from_1d(object, contrast))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">99</td>
+                    <td class="num">201</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{2}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">100</td>
+                    <td class="num">202</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\left[ \rho I(i \neq j)+I(i=j) \right]}</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">101</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">203</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">  contrast_cov &lt;- h_quad_form_mat(contrast, component(object, "beta_vcov"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">102</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">204</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  eigen_cont_cov &lt;- eigen(contrast_cov)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">103</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">205</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' csh</pre>
+                      <pre class="language-r">  eigen_cont_cov_vctrs &lt;- eigen_cont_cov$vectors</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">104</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">206</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Heterogeneous compound symmetry</pre>
+                      <pre class="language-r">  eigen_cont_cov_vals &lt;- eigen_cont_cov$values</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">105</td>
+                    <td class="num">207</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{m+1}</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">106</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">208</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\left[ \rho I(i \neq j)+I(i=j) \right]}</pre>
+                      <pre class="language-r">  eps &lt;- sqrt(.Machine$double.eps)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">107</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">209</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">  tol &lt;- max(eps * eigen_cont_cov_vals[1], 0)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">108</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">210</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  rank_cont_cov &lt;- sum(eigen_cont_cov_vals &gt; tol)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">109</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">211</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' toep</pre>
+                      <pre class="language-r">  assert_number(rank_cont_cov, lower = .Machine$double.xmin)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">110</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">212</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Toeplitz</pre>
+                      <pre class="language-r">  rank_seq &lt;- seq_len(rank_cont_cov)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">111</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">213</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{m}</pre>
+                      <pre class="language-r">  vctrs_cont_prod &lt;- crossprod(eigen_cont_cov_vctrs, contrast)[rank_seq, , drop = FALSE]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">112</td>
+                    <td class="num">214</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma_{\left \vert {i-j} \right \vert +1}}</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">113</td>
+                    <td class="num">215</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">  # Early return if rank 1.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">114</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">216</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  if (identical(rank_cont_cov, 1L)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">115</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">217</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' toeph</pre>
+                      <pre class="language-r">    return(h_df_md_from_1d(object, vctrs_cont_prod))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">116</td>
+                    <td class="num">218</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Heterogeneous Toeplitz</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">117</td>
+                    <td class="num">219</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{2m-1}</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">118</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">220</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma_{i}\sigma_{j}\rho_{\left \vert {i-j} \right \vert}}</pre>
+                      <pre class="language-r">  t_squared_nums &lt;- drop(vctrs_cont_prod %*% object$beta_est)^2</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">119</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">221</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r">  t_squared_denoms &lt;- eigen_cont_cov_vals[rank_seq]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">120</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">222</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  t_squared &lt;- t_squared_nums / t_squared_denoms</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">121</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">223</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' us</pre>
+                      <pre class="language-r">  f_stat &lt;- sum(t_squared) / rank_cont_cov</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">122</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">224</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab Unstructured</pre>
+                      <pre class="language-r">  t_stat_df_nums &lt;- 2 * eigen_cont_cov_vals^2</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">123</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">225</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{m(m+1)/2}</pre>
+                      <pre class="language-r">  t_stat_df &lt;- if (identical(object$vcov, "Asymptotic")) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">124</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">226</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma_{ij}}</pre>
+                      <pre class="language-r">    grads_vctrs_cont_prod &lt;- lapply(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">125</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">227</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      rank_seq,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">126</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">228</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' }</pre>
+                      <pre class="language-r">      function(m) h_gradient(component(object, "jac_list"), contrast = vctrs_cont_prod[m, ])</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">127</td>
+                    <td class="num">229</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    )</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">230</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    t_stat_df_denoms &lt;- vapply(</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">231</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      grads_vctrs_cont_prod,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">232</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      h_quad_form_vec,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">128</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">233</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' where \eqn{i} and \eqn{j} denote \eqn{i}-th and \eqn{j}-th time points,</pre>
+                      <pre class="language-r">      center = component(object, "theta_vcov"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">129</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">234</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' respectively, out of total \eqn{m} time points, \eqn{1 \leq i, j \leq m}.</pre>
+                      <pre class="language-r">      numeric(1)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">130</td>
+                    <td class="num">235</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">131</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">236</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @note The **ante-dependence** covariance structure in this package refers to</pre>
+                      <pre class="language-r">    t_stat_df_nums / t_stat_df_denoms</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">132</td>
+                    <td class="num">237</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' homogeneous ante-dependence, while the ante-dependence covariance structure</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">133</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">238</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' from SAS `PROC MIXED` refers to heterogeneous ante-dependence and the</pre>
+                      <pre class="language-r">    vapply(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">134</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">239</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' homogeneous version is not available in SAS.</pre>
+                      <pre class="language-r">      rank_seq,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">135</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">240</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      function(m) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">136</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">241</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @note For all non-spatial covariance structures, the time variable must</pre>
+                      <pre class="language-r">        contrast_matrix &lt;- Matrix::.bdiag(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">137</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">242</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' be coded as a factor.</pre>
+                      <pre class="language-r">          rep(list(vctrs_cont_prod[m, , drop = FALSE]), component(object, "n_subjects"))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">138</td>
+                    <td class="num">243</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">139</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">244</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ## Spatial Covariance structures:</pre>
+                      <pre class="language-r">        contrast_matrix &lt;- as.matrix(contrast_matrix)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">140</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">245</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        g_matrix &lt;- h_quad_form_mat(contrast_matrix, object$empirical_df_mat)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">141</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">246</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tabular{clll}{</pre>
+                      <pre class="language-r">        h_tr(g_matrix)^2 / sum(g_matrix^2)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">142</td>
+                    <td class="num">247</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      },</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">143</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">248</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \strong{Structure}</pre>
+                      <pre class="language-r">      FUN.VALUE = 0</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">144</td>
+                    <td class="num">249</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \strong{Description}</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">145</td>
+                    <td class="num">250</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \strong{Parameters}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">146</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">251</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \strong{\eqn{(i, j)} element}</pre>
+                      <pre class="language-r">  denom_df &lt;- h_md_denom_df(t_stat_df)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">147</td>
+                    <td class="num">252</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \cr</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">148</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">253</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">149</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">254</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' sp_exp</pre>
+                      <pre class="language-r">    num_df = rank_cont_cov,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">150</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">255</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab spatial exponential</pre>
+                      <pre class="language-r">    denom_df = denom_df,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">151</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">256</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{2}</pre>
+                      <pre class="language-r">    f_stat = f_stat,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">152</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">257</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \tab \eqn{\sigma^{2}\rho^{-d_{ij}}}</pre>
+                      <pre class="language-r">    p_val = stats::pf(q = f_stat, df1 = rank_cont_cov, df2 = denom_df, lower.tail = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">153</td>
+                    <td class="num">258</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">154</td>
+                    <td class="num">259</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' }</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/component.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">155</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Component Access for `mmrm_tmb` Objects</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">156</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' where \eqn{d_{ij}} denotes the Euclidean distance between time points</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">157</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' \eqn{i} and \eqn{j}.</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">158</td>
+                    <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">159</td>
+                    <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @family covariance types</pre>
+                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">160</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @name covariance_types</pre>
+                      <pre class="language-r">#' @param name (`character`)\cr the component(s) to be retrieved.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">161</td>
+                    <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#' @return The corresponding component of the object, see details.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">162</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">cov_types &lt;- function(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">163</td>
+                    <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    form = c("name", "abbr", "habbr"),</pre>
+                      <pre class="language-r">#' @details Available `component()` names are as follows:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">164</td>
+                    <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    filter = c("heterogeneous", "spatial")) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">165</td>
-                    <td class="coverage">1666<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  form &lt;- match.arg(form, several.ok = TRUE)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">166</td>
-                    <td class="coverage">1666<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  filter &lt;- if (missing(filter)) c() else match.arg(filter, several.ok = TRUE)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">167</td>
-                    <td class="coverage">1666<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- COV_TYPES[form][rowSums(!COV_TYPES[filter]) == 0, ]</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">168</td>
-                    <td class="coverage">1666<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  Filter(Negate(is.na), unlist(t(df), use.names = FALSE))</pre>
+                      <pre class="language-r">#' - `call`: low-level function call which generated the model.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">169</td>
+                    <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' - `formula`: model formula.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">170</td>
+                    <td class="num">12</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' - `dataset`: data set name.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">171</td>
+                    <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Retrieve Associated Abbreviated Covariance Structure Type Name</pre>
+                      <pre class="language-r">#' - `cov_type`: covariance structure type.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">172</td>
+                    <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `n_theta`: number of parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">173</td>
+                    <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param type (`string`)\cr either a full name or abbreviate covariance</pre>
+                      <pre class="language-r">#' - `n_subjects`: number of subjects.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">174</td>
+                    <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   structure type name to collapse into an abbreviated type.</pre>
+                      <pre class="language-r">#' - `n_timepoints`: number of modeled time points.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">175</td>
+                    <td class="num">17</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `n_obs`: total number of observations.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">176</td>
+                    <td class="num">18</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The corresponding abbreviated covariance type name.</pre>
+                      <pre class="language-r">#' - `reml`: was REML used (ML was used if `FALSE`).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">177</td>
+                    <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `neg_log_lik`: negative log likelihood.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">178</td>
+                    <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' - `convergence`: convergence code from optimizer.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">179</td>
+                    <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">cov_type_abbr &lt;- function(type) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">180</td>
-                    <td class="coverage">299<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  row &lt;- which(COV_TYPES == type, arr.ind = TRUE)[, 1]</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">181</td>
-                    <td class="coverage">299<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  COV_TYPES$abbr[row]</pre>
+                      <pre class="language-r">#' - `conv_message`: message accompanying the convergence code.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">182</td>
+                    <td class="num">22</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' - `evaluations`: number of function evaluations for optimization.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">183</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' - `method`: Adjustment method which was used (for `mmrm` objects),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">184</td>
+                    <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Retrieve Associated Full Covariance Structure Type Name</pre>
+                      <pre class="language-r">#'      otherwise `NULL` (for `mmrm_tmb` objects).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">185</td>
+                    <td class="num">25</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `beta_vcov`: estimated variance-covariance matrix of coefficients</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">186</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param type (`string`)\cr either a full name or abbreviate covariance</pre>
+                      <pre class="language-r">#'      (excluding aliased coefficients). When Kenward-Roger/Empirical adjusted</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">187</td>
+                    <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   structure type name to convert to a long-form type.</pre>
+                      <pre class="language-r">#'      coefficients covariance matrix is used, the adjusted covariance matrix is returned (to still obtain the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">188</td>
+                    <td class="num">28</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'      original asymptotic covariance matrix use `object$beta_vcov`).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">189</td>
+                    <td class="num">29</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The corresponding abbreviated covariance type name.</pre>
+                      <pre class="language-r">#' - `beta_vcov_complete`: estimated variance-covariance matrix including</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">190</td>
+                    <td class="num">30</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'      aliased coefficients with entries set to `NA`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">191</td>
+                    <td class="num">31</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' - `varcor`: estimated covariance matrix for residuals. If there are multiple</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">192</td>
+                    <td class="num">32</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">cov_type_name &lt;- function(type) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">193</td>
-                    <td class="coverage">6<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  row &lt;- which(COV_TYPES == type, arr.ind = TRUE)[, 1]</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">194</td>
-                    <td class="coverage">6<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  COV_TYPES$name[row]</pre>
+                      <pre class="language-r">#'      groups, a named list of estimated covariance matrices for residuals will be</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">195</td>
+                    <td class="num">33</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'      returned. The names are the group levels.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">196</td>
+                    <td class="num">34</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' - `theta_est`: estimated variance parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">197</td>
+                    <td class="num">35</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Produce A Covariance Identifier Passing to TMB</pre>
+                      <pre class="language-r">#' - `beta_est`: estimated coefficients (excluding aliased coefficients).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">198</td>
+                    <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `beta_est_complete`: estimated coefficients including aliased coefficients</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">199</td>
+                    <td class="num">37</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param cov (`cov_struct`)\cr a covariance structure object.</pre>
+                      <pre class="language-r">#'      set to `NA`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">200</td>
+                    <td class="num">38</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `beta_aliased`: whether each coefficient was aliased (i.e. cannot be estimated)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">201</td>
+                    <td class="num">39</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A string used for method dispatch when passed to TMB.</pre>
+                      <pre class="language-r">#'      or not.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">202</td>
+                    <td class="num">40</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `theta_vcov`:  estimated variance-covariance matrix of variance parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">203</td>
+                    <td class="num">41</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' - `x_matrix`: design matrix used (excluding aliased columns).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">204</td>
+                    <td class="num">42</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">tmb_cov_type &lt;- function(cov) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">205</td>
-                    <td class="coverage">266<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  paste0(cov$type, if (cov$heterogeneous) "h")</pre>
+                      <pre class="language-r">#' - `xlev`: 	a named list of character vectors giving the full set of levels to be assumed for each factor.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">206</td>
+                    <td class="num">43</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' - `contrasts`: a list of contrasts used for each factor.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">207</td>
+                    <td class="num">44</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' - `y_vector`: response vector used.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">208</td>
+                    <td class="num">45</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Define a Covariance Structure</pre>
+                      <pre class="language-r">#' - `jac_list`: Jacobian, see  [h_jac_list()] for details.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">209</td>
+                    <td class="num">46</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `full_frame`: `data.frame` with `n` rows containing all variables needed in the model.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">210</td>
+                    <td class="num">47</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">211</td>
+                    <td class="num">48</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @seealso In the `lme4` package there is a similar function `getME()`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">212</td>
+                    <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param type (`string`)\cr the name of the covariance structure type to use.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">213</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   For available options, see `cov_types()`. If a type abbreviation is used</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">214</td>
+                    <td class="num">51</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   that implies heterogeneity (e.g. `cph`) and no value is provided to</pre>
+                      <pre class="language-r">#' fit &lt;- mmrm(</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">215</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `heterogeneous`, then the heterogeneity is derived from the type name.</pre>
+                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">216</td>
+                    <td class="num">53</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param visits (`character`)\cr a vector of variable names to use for the</pre>
+                      <pre class="language-r">#'   data = fev_data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">217</td>
+                    <td class="num">54</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   longitudinal terms of the covariance structure. Multiple terms are only</pre>
+                      <pre class="language-r">#' )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">218</td>
+                    <td class="num">55</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   permitted for the `"spatial"` covariance type.</pre>
+                      <pre class="language-r">#' # Get all available components.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">219</td>
+                    <td class="num">56</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param subject (`string`)\cr the name of the variable that encodes a subject</pre>
+                      <pre class="language-r">#' component(fit)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">220</td>
+                    <td class="num">57</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   identifier.</pre>
+                      <pre class="language-r">#' # Get convergence code and message.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">221</td>
+                    <td class="num">58</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param group (`string`)\cr optionally, the name of the variable that encodes</pre>
+                      <pre class="language-r">#' component(fit, c("convergence", "conv_message"))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">222</td>
+                    <td class="num">59</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   a grouping variable for subjects.</pre>
+                      <pre class="language-r">#' # Get modeled formula as a string.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">223</td>
+                    <td class="num">60</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param heterogeneous (`flag`)\cr</pre>
+                      <pre class="language-r">#' component(fit, c("formula"))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">224</td>
+                    <td class="num">61</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">225</td>
+                    <td class="num">62</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A `cov_struct` object.</pre>
+                      <pre class="language-r">#' @export</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">226</td>
+                    <td class="num">63</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">component &lt;- function(object,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">227</td>
+                    <td class="num">64</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">                      name = c(</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">228</td>
+                    <td class="num">65</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' cov_struct("csh", "AVISITN", "USUBJID")</pre>
+                      <pre class="language-r">                        "cov_type", "subject_var", "n_theta", "n_subjects", "n_timepoints",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">229</td>
+                    <td class="num">66</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' cov_struct("spatial", c("VISITA", "VISITB"), group = "GRP", subject = "SBJ")</pre>
+                      <pre class="language-r">                        "n_obs", "beta_vcov", "beta_vcov_complete",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">230</td>
+                    <td class="num">67</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                        "varcor", "formula", "dataset", "n_groups",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">231</td>
+                    <td class="num">68</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @family covariance types</pre>
+                      <pre class="language-r">                        "reml", "convergence", "evaluations", "method", "optimizer",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">232</td>
+                    <td class="num">69</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">                        "conv_message", "call", "theta_est",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">233</td>
+                    <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">cov_struct &lt;- function(</pre>
+                      <pre class="language-r">                        "beta_est", "beta_est_complete", "beta_aliased",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">234</td>
+                    <td class="num">71</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    type = cov_types(), visits, subject, group = character(),</pre>
+                      <pre class="language-r">                        "x_matrix", "y_vector", "neg_log_lik", "jac_list", "theta_vcov",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">235</td>
+                    <td class="num">72</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    heterogeneous = FALSE) {</pre>
+                      <pre class="language-r">                        "full_frame", "xlev", "contrasts"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">236</td>
+                    <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # if heterogeneous isn't provided, derive from provided type</pre>
+                      <pre class="language-r">                      )) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">237</td>
-                    <td class="coverage">296<em>x</em></td>
+                    <td class="num">74</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(heterogeneous)) {</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">238</td>
-                    <td class="coverage">294<em>x</em></td>
+                    <td class="num">75</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    heterogeneous &lt;- switch(type,</pre>
+                      <pre class="language-r">  name &lt;- match.arg(name, several.ok = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">239</td>
-                    <td class="coverage">294<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">76</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      toeph = ,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">240</td>
-                    <td class="coverage">294<em>x</em></td>
+                    <td class="num">77</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ar1h = ,</pre>
+                      <pre class="language-r">  list_components &lt;- sapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">241</td>
-                    <td class="coverage">294<em>x</em></td>
+                    <td class="num">78</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      adh = ,</pre>
+                      <pre class="language-r">    X = name,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">242</td>
-                    <td class="coverage">294<em>x</em></td>
+                    <td class="num">79</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      csh = TRUE,</pre>
+                      <pre class="language-r">    FUN = switch,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">243</td>
-                    <td class="coverage">294<em>x</em></td>
+                    <td class="num">80</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      heterogeneous</pre>
+                      <pre class="language-r">    "call" = object$call,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">244</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">    # Strings.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">245</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">82</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    "cov_type" = object$formula_parts$cov_type,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">246</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">83</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    "subject_var" = object$formula_parts$subject_var,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">247</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">84</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # coerce all type options into abbreviated form</pre>
+                      <pre class="language-r">    "formula" = deparse(object$call$formula),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">248</td>
-                    <td class="coverage">296<em>x</em></td>
+                    <td class="num">85</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
+                      <pre class="language-r">    "dataset" = object$call$data,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">249</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">86</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type &lt;- cov_type_abbr(type)</pre>
+                      <pre class="language-r">    "reml" = object$reml,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">87</td>
+                    <td class="coverage">5115<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    "conv_message" = object$opt_details$message,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">250</td>
+                    <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    # Numeric of length 1.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">251</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">89</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x &lt;- structure(</pre>
+                      <pre class="language-r">    "convergence" = object$opt_details$convergence,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">252</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">90</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    list(</pre>
+                      <pre class="language-r">    "neg_log_lik" = object$neg_log_lik,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">253</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">91</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      type = type,</pre>
+                      <pre class="language-r">    "n_theta" = length(object$theta_est),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">254</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">92</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      heterogeneous = heterogeneous,</pre>
+                      <pre class="language-r">    "n_subjects" = object$tmb_data$n_subjects,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">255</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">93</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      visits = visits,</pre>
+                      <pre class="language-r">    "n_timepoints" = object$tmb_data$n_visits,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">256</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">94</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      subject = subject,</pre>
+                      <pre class="language-r">    "n_obs" = length(object$tmb_data$y_vector),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">257</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">95</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      group = group</pre>
+                      <pre class="language-r">    "n_groups" = ifelse(is.list(object$cov), length(object$cov), 1L),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">258</td>
+                    <td class="num">96</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ),</pre>
+                      <pre class="language-r">    # Numeric of length &gt; 1.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">259</td>
-                    <td class="coverage">295<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    class = c("cov_struct", "mmrm_cov_struct", "list")</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">260</td>
-                    <td class="coverage"></td>
+                    <td class="num">97</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    "evaluations" = unlist(ifelse(is.null(object$opt_details$evaluations),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">261</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">98</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      list(object$opt_details$counts),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">262</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">99</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  validate_cov_struct(x)</pre>
+                      <pre class="language-r">      list(object$opt_details$evaluations)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">263</td>
+                    <td class="num">100</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    )),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">264</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">101</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    "method" = object$method,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">265</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">102</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Reconcile Possible Covariance Structure Inputs</pre>
+                      <pre class="language-r">    "optimizer" = object$optimizer,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">266</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">103</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    "beta_est" = object$beta_est,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">267</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">104</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams mmrm</pre>
+                      <pre class="language-r">    "beta_est_complete" =</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">268</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">105</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      if (any(object$tmb_data$x_cols_aliased)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">269</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">106</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The value `covariance` if it's provided or a covariance structure</pre>
+                      <pre class="language-r">        stats::setNames(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">270</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">107</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   derived from the provided `formula` otherwise. An error is raised of both</pre>
+                      <pre class="language-r">          object$beta_est[names(object$tmb_data$x_cols_aliased)],</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">271</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">108</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   are provided.</pre>
+                      <pre class="language-r">          names(object$tmb_data$x_cols_aliased)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">272</td>
+                    <td class="num">109</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">273</td>
+                    <td class="num">110</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">      } else {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">111</td>
+                    <td class="coverage">54<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">        object$beta_est</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">274</td>
+                    <td class="num">112</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_reconcile_cov_struct &lt;- function(formula = NULL, covariance = NULL) {</pre>
+                      <pre class="language-r">      },</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">275</td>
-                    <td class="coverage">238<em>x</em></td>
+                    <td class="num">113</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_multi_class(covariance, c("formula", "cov_struct"), null.ok = TRUE)</pre>
+                      <pre class="language-r">    "beta_aliased" = object$tmb_data$x_cols_aliased,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">276</td>
-                    <td class="coverage">238<em>x</em></td>
+                    <td class="num">114</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_formula(formula, null.ok = FALSE)</pre>
+                      <pre class="language-r">    "theta_est" = object$theta_est,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">277</td>
-                    <td class="coverage">238<em>x</em></td>
+                    <td class="num">115</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (inherits(covariance, "formula")) {</pre>
+                      <pre class="language-r">    "y_vector" = object$tmb_data$y_vector,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">278</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">116</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    covariance &lt;- as.cov_struct(covariance)</pre>
+                      <pre class="language-r">    "jac_list" = object$jac_list,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">279</td>
+                    <td class="num">117</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">280</td>
-                    <td class="coverage">238<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(covariance) &amp;&amp; length(h_extract_covariance_terms(formula)) &gt; 0) {</pre>
+                      <pre class="language-r">    # Matrices.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">281</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">118</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop(paste0(</pre>
+                      <pre class="language-r">    "beta_vcov" =</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">282</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">119</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Redundant covariance structure definition in `formula` and ",</pre>
+                      <pre class="language-r">      if (is.null(object$vcov) || identical(object$vcov, "Asymptotic")) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">283</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">120</td>
+                    <td class="coverage">985<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "`covariance` arguments"</pre>
+                      <pre class="language-r">        object$beta_vcov</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">284</td>
+                    <td class="num">121</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ))</pre>
+                      <pre class="language-r">      } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">285</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">122</td>
+                    <td class="coverage">66<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">        object$beta_vcov_adj</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">286</td>
+                    <td class="num">123</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      },</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">287</td>
-                    <td class="coverage">236<em>x</em></td>
+                    <td class="num">124</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(covariance)) {</pre>
+                      <pre class="language-r">    "beta_vcov_complete" =</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">288</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">125</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(covariance)</pre>
+                      <pre class="language-r">      if (any(object$tmb_data$x_cols_aliased)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">289</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">126</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">        stats::.vcov.aliased(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">290</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">127</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">          aliased = object$tmb_data$x_cols_aliased,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">291</td>
-                    <td class="coverage">231<em>x</em></td>
+                    <td class="num">128</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  as.cov_struct(formula, warn_partial = FALSE)</pre>
+                      <pre class="language-r">          vc = component(object, "beta_vcov"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">292</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">129</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">          complete = TRUE</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">293</td>
+                    <td class="num">130</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">        )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">294</td>
+                    <td class="num">131</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Validate Covariance Structure Data</pre>
+                      <pre class="language-r">      } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">295</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">132</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        object$beta_vcov</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">296</td>
+                    <td class="num">133</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Run checks against relational integrity of covariance definition</pre>
+                      <pre class="language-r">      },</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">297</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">134</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    "varcor" = object$cov,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">298</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">135</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`cov_struct`)\cr a covariance structure object.</pre>
+                      <pre class="language-r">    "x_matrix" = object$tmb_data$x_matrix,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">299</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">136</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    "xlev" = stats::.getXlevels(terms(object), object$tmb_data$full_frame),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">300</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">137</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return `x` if successful, or an error is thrown otherwise.</pre>
+                      <pre class="language-r">    "contrasts" = attr(object$tmb_data$x_matrix, "contrasts"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">301</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">138</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    "theta_vcov" = object$theta_vcov,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">302</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">139</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    "full_frame" = object$tmb_data$full_frame,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">303</td>
+                    <td class="num">140</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">validate_cov_struct &lt;- function(x) {</pre>
+                      <pre class="language-r">    # If not found.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">304</td>
-                    <td class="coverage">295<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  checks &lt;- checkmate::makeAssertCollection()</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">305</td>
-                    <td class="coverage"></td>
+                    <td class="num">141</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    "..foo.." =</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">306</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">142</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  with(x, {</pre>
+                      <pre class="language-r">      stop(sprintf(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">307</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">143</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_character(subject, len = 1, add = checks)</pre>
+                      <pre class="language-r">        "component '%s' is not available",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">308</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">144</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_logical(heterogeneous, len = 1, add = checks)</pre>
+                      <pre class="language-r">        name, paste0(class(object), collapse = ", ")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">309</td>
+                    <td class="num">145</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      )),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">310</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">146</td>
+                    <td class="coverage">5115<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (length(group) &gt; 1 || length(visits) &lt; 1) {</pre>
+                      <pre class="language-r">    simplify = FALSE</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">147</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">311</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">148</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      checks$push(</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">312</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">149</td>
+                    <td class="coverage">23<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "Covariance structure must be of the form `time | (group /) subject`"</pre>
+                      <pre class="language-r">  if (length(name) == 1) list_components[[1]] else list_components</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">313</td>
+                    <td class="num">150</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/interop-parsnip.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">314</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#' Register `mmrm` For Use With `tidymodels`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">315</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">316</td>
-                    <td class="coverage">295<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">3</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (!type %in% cov_types(filter = "spatial") &amp;&amp; length(visits) &gt; 1) {</pre>
+                      <pre class="language-r">#' @inheritParams base::requireNamespace</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">317</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">4</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      checks$push(paste(</pre>
+                      <pre class="language-r">#' @return A logical value indicating whether registration was successful.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">318</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">5</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "Non-spatial covariance structures must have a single longitudinal",</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">319</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">6</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "variable"</pre>
+                      <pre class="language-r">#' @details We can use `parsnip::show_model_info("linear_reg")` to check the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">320</td>
+                    <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ))</pre>
+                      <pre class="language-r">#'   registration with `parsnip` and thus the wider `tidymodels` ecosystem.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">321</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">322</td>
+                    <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">323</td>
+                    <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">parsnip_add_mmrm &lt;- function(quietly = FALSE) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">324</td>
-                    <td class="coverage">295<em>x</em></td>
+                    <td class="num">11</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  reportAssertions(checks)</pre>
+                      <pre class="language-r">  if (!requireNamespace("parsnip", quietly = quietly)) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">325</td>
-                    <td class="coverage">289<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">12</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x</pre>
+                      <pre class="language-r">    return(FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">326</td>
+                    <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">327</td>
+                    <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">328</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">15</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Format Covariance Structure Object</pre>
+                      <pre class="language-r">  parsnip::set_model_engine(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">329</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">16</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    model = "linear_reg",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">330</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">17</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`cov_struct`)\cr a covariance structure object.</pre>
+                      <pre class="language-r">    eng = "mmrm",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">331</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">18</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... Additional arguments unused.</pre>
+                      <pre class="language-r">    mode = "regression"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">332</td>
+                    <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">333</td>
+                    <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A formatted string for `x`.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">334</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">21</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  parsnip::set_dependency(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">335</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">22</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">    pkg = "mmrm",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">336</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">23</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">format.cov_struct &lt;- function(x, ...) {</pre>
+                      <pre class="language-r">    model = "linear_reg",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">337</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">24</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  sprintf(</pre>
+                      <pre class="language-r">    eng = "mmrm",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">338</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">25</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "&lt;covariance structure&gt;\n%s%s:\n\n  %s | %s%s\n",</pre>
+                      <pre class="language-r">    mode = "regression"</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">26</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  )</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">27</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">339</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">28</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (x$heterogeneous) "heterogeneous " else "",</pre>
+                      <pre class="language-r">  parsnip::set_encoding(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">340</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">29</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cov_type_name(x$type),</pre>
+                      <pre class="language-r">    model = "linear_reg",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">341</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">30</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    format_symbols(x$visits),</pre>
+                      <pre class="language-r">    eng = "mmrm",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">342</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">31</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    if (length(x$group) &gt; 0) paste0(format_symbols(x$group), " / ") else "",</pre>
+                      <pre class="language-r">    mode = "regression",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">343</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">32</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    format_symbols(x$subject)</pre>
+                      <pre class="language-r">    options = list(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">344</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">33</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">      predictor_indicators = "none",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">345</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">34</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      compute_intercept = FALSE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">346</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">35</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      remove_intercept = FALSE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">347</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">36</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Print a Covariance Structure Object</pre>
+                      <pre class="language-r">      allow_sparse_x = TRUE</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">348</td>
+                    <td class="num">37</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">349</td>
+                    <td class="num">38</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`cov_struct`)\cr a covariance structure object.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">350</td>
+                    <td class="num">39</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... Additional arguments unused.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">351</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">40</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  parsnip::set_fit(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">352</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">41</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return `x` invisibly.</pre>
+                      <pre class="language-r">    model = "linear_reg",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">353</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">42</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    eng = "mmrm",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">354</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">43</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">    mode = "regression",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">355</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">44</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">print.cov_struct &lt;- function(x, ...) {</pre>
+                      <pre class="language-r">    value = list(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">356</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">45</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat(format(x, ...), "\n")</pre>
+                      <pre class="language-r">      interface = "formula",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">357</td>
-                    <td class="coverage">3<em>x</em></td>
+                    <td class="num">46</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  invisible(x)</pre>
+                      <pre class="language-r">      protect = c("formula", "data", "weights"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">358</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">47</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      data = c(formula = "formula", data = "data", weights = "weights"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">359</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">48</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      func = c(pkg = "mmrm", fun = "mmrm"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">360</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">49</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Coerce into a Covariance Structure Definition</pre>
+                      <pre class="language-r">      defaults = list()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">361</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">362</td>
+                    <td class="num">51</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">363</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">364</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">53</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">  parsnip::set_pred(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">365</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">54</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' A covariance structure can be parsed from a model definition formula or call.</pre>
+                      <pre class="language-r">    model = "linear_reg",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">366</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">55</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Generally, covariance structures defined using non-standard evaluation take</pre>
+                      <pre class="language-r">    eng = "mmrm",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">367</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">56</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' the following form:</pre>
+                      <pre class="language-r">    mode = "regression",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">368</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">57</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    type = "numeric",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">369</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">58</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ```</pre>
+                      <pre class="language-r">    value = parsnip::pred_value_template(</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">370</td>
+                    <td class="num">59</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' type( (visit, )* visit | (group /)? subject )</pre>
+                      <pre class="language-r">      # This is boilerplate.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">371</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">60</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ```</pre>
+                      <pre class="language-r">      func = c(fun = "predict"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">372</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">61</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      object = quote(object$fit),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">373</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">62</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' For example, formulas may include terms such as</pre>
+                      <pre class="language-r">      newdata = quote(new_data)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">374</td>
+                    <td class="num">63</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">375</td>
+                    <td class="num">64</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ```r</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">376</td>
+                    <td class="num">65</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' us(time | subject)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">377</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">66</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' cp(time | group / subject)</pre>
+                      <pre class="language-r">  parsnip::set_pred(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">378</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">67</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' sp_exp(coord1, coord2 | group / subject)</pre>
+                      <pre class="language-r">    model = "linear_reg",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">379</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">68</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ```</pre>
+                      <pre class="language-r">    eng = "mmrm",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">380</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">69</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    mode = "regression",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">381</td>
+                    <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Note that only `sp_exp` (spatial) covariance structures may provide multiple</pre>
+                      <pre class="language-r">    # This type allows to pass arguments via `opts` to `parsnip::predict.model_fit`.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">382</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">71</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' coordinates, which identify the Euclidean distance between the time points.</pre>
+                      <pre class="language-r">    type = "raw",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">383</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">72</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    value = parsnip::pred_value_template(</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">384</td>
+                    <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x an object from which to derive a covariance structure. See object</pre>
+                      <pre class="language-r">      # This is boilerplate.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">385</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">74</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   specific sections for details.</pre>
+                      <pre class="language-r">      func = c(fun = "predict"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">386</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">75</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param warn_partial (`flag`)\cr whether to emit a warning when parts of the</pre>
+                      <pre class="language-r">      object = quote(object$fit),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">387</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">76</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   formula are disregarded.</pre>
+                      <pre class="language-r">      newdata = quote(new_data)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">388</td>
+                    <td class="num">77</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... additional arguments unused.</pre>
+                      <pre class="language-r">      # We don't specify additional argument defaults here since otherwise</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">389</td>
+                    <td class="num">78</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      # the user is not able to change them (they will be fixed).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">390</td>
+                    <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A [cov_struct()] object.</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">391</td>
+                    <td class="num">80</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">392</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">393</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">82</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # provide a covariance structure as a right-sided formula</pre>
+                      <pre class="language-r">  TRUE</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">394</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' as.cov_struct(~ csh(visit | group / subject))</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/tmb-methods.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">395</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Methods for `mmrm_tmb` Objects</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">396</td>
+                    <td class="num">2</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # when part of a full formula, suppress warnings using `warn_partial = FALSE`</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">397</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' as.cov_struct(y ~ x + csh(visit | group / subject), warn_partial = FALSE)</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">398</td>
+                    <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">399</td>
+                    <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @family covariance types</pre>
+                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">400</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#' @param x (`mmrm_tmb`)\cr same as `object`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">401</td>
+                    <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">as.cov_struct &lt;- function(x, ...) { # nolint</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">402</td>
-                    <td class="coverage">278<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  UseMethod("as.cov_struct")</pre>
+                      <pre class="language-r">#' @param formula (`mmrm_tmb`)\cr same as `object`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">403</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @param complete (`flag`)\cr whether to include potential non-estimable</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">404</td>
+                    <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">405</td>
+                    <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#' @param ... mostly not used;</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">406</td>
+                    <td class="num">11</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">as.cov_struct.cov_struct &lt;- function(x, ...) {</pre>
+                      <pre class="language-r">#'   Exception is `model.matrix()` passing `...` to the default method.</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">407</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">12</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x</pre>
+                      <pre class="language-r">#' @return Depends on the method, see Functions.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">408</td>
+                    <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">409</td>
+                    <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @name mmrm_tmb_methods</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">410</td>
+                    <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn as.cov_struct</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">411</td>
+                    <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' When provided a formula, any specialized functions are assumed to be</pre>
+                      <pre class="language-r">#' @seealso [`mmrm_methods`], [`mmrm_tidiers`] for additional methods.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">412</td>
+                    <td class="num">17</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' covariance structure definitions and must follow the form:</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">413</td>
+                    <td class="num">18</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">414</td>
+                    <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ```</pre>
+                      <pre class="language-r">#' formula &lt;- FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">415</td>
+                    <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' y ~ xs + type( (visit, )* visit | (group /)? subject )</pre>
+                      <pre class="language-r">#' object &lt;- fit_mmrm(formula, fev_data, weights = rep(1, nrow(fev_data)))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">416</td>
+                    <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' ```</pre>
+                      <pre class="language-r">NULL</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">417</td>
+                    <td class="num">22</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">418</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Any component on the right hand side of a formula is considered when</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the estimated coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">419</td>
+                    <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' searching for a covariance definition.</pre>
+                      <pre class="language-r">#' @importFrom stats coef</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">420</td>
+                    <td class="num">25</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">421</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">422</td>
+                    <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">as.cov_struct.formula &lt;- function(x, warn_partial = TRUE, ...) {</pre>
+                      <pre class="language-r">#' # Estimated coefficients:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">423</td>
-                    <td class="coverage">278<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">28</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x_calls &lt;- h_extract_covariance_terms(x)</pre>
+                      <pre class="language-r">#' coef(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">424</td>
+                    <td class="num">29</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">coef.mmrm_tmb &lt;- function(object, complete = TRUE, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">425</td>
-                    <td class="coverage">278<em>x</em></td>
+                    <td class="num">30</td>
+                    <td class="coverage">58<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (length(x_calls) &lt; 1) {</pre>
+                      <pre class="language-r">  assert_flag(complete)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">426</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">31</td>
+                    <td class="coverage">58<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop(</pre>
+                      <pre class="language-r">  nm &lt;- if (complete) "beta_est_complete" else "beta_est"</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">427</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">32</td>
+                    <td class="coverage">58<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Covariance structure must be specified in formula. ",</pre>
+                      <pre class="language-r">  component(object, name = nm)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">428</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">33</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Possible covariance structures include: ",</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">429</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">34</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      paste0(cov_types(c("abbr", "habbr")), collapse = ", ")</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">430</td>
+                    <td class="num">35</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the fitted values.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">431</td>
+                    <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @importFrom stats fitted</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">432</td>
+                    <td class="num">37</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">433</td>
-                    <td class="coverage">274<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">38</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (length(x_calls) &gt; 1) {</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">434</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">39</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cov_struct_types &lt;- as.character(lapply(x_calls, `[[`, 1L))</pre>
+                      <pre class="language-r">#' # Fitted values:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">435</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">40</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop(</pre>
+                      <pre class="language-r">#' fitted(object)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">436</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">41</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Only one covariance structure can be specified. ",</pre>
+                      <pre class="language-r">fitted.mmrm_tmb &lt;- function(object, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">437</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">42</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Currently specified covariance structures are: ",</pre>
+                      <pre class="language-r">  fitted_col &lt;- component(object, "x_matrix") %*% component(object, "beta_est")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">438</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">43</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      paste0(cov_struct_types, collapse = ", ")</pre>
+                      <pre class="language-r">  fitted_col[, 1L, drop = TRUE]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">439</td>
+                    <td class="num">44</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">440</td>
+                    <td class="num">45</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">441</td>
+                    <td class="num">46</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods predict conditional means for new data;</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">442</td>
+                    <td class="num">47</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # flatten into list of infix operators, calls and names/atomics</pre>
+                      <pre class="language-r">#'  optionally with standard errors and confidence or prediction intervals.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">443</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">48</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x &lt;- flatten_call(x_calls[[1]])</pre>
+                      <pre class="language-r">#'  Returns a vector of predictions if `se.fit == FALSE` and</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">444</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">49</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type &lt;- as.character(x[[1]])</pre>
+                      <pre class="language-r">#'  `interval == "none"`; otherwise it returns a data.frame with multiple</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">445</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">50</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x &lt;- drop_elements(x, 1)</pre>
+                      <pre class="language-r">#'  columns and one row per input data row.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">446</td>
+                    <td class="num">51</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">447</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # take visits until "|"</pre>
+                      <pre class="language-r">#' @param newdata (`data.frame`)\cr optional new data, otherwise data from `object` is used.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">448</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">53</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n &lt;- position_symbol(x, "|", nomatch = 0)</pre>
+                      <pre class="language-r">#' @param se.fit (`flag`)\cr indicator if standard errors are required.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">449</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">54</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  visits &lt;- as.character(utils::head(x, max(n - 1, 0)))</pre>
+                      <pre class="language-r">#' @param interval (`string`)\cr type of interval calculation. Can be abbreviated.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">450</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">55</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x &lt;- drop_elements(x, n)</pre>
+                      <pre class="language-r">#' @param level (`number`)\cr tolerance/confidence level.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">451</td>
+                    <td class="num">56</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @param nsim (`count`)\cr number of simulations to use.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">452</td>
+                    <td class="num">57</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # take group until "/"</pre>
+                      <pre class="language-r">#' @param conditional (`flag`)\cr indicator if the prediction is conditional on the observation or not.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">453</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">58</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n &lt;- position_symbol(x, "/", nomatch = 0)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">454</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">59</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  group &lt;- as.character(utils::head(x, max(n - 1, 0)))</pre>
+                      <pre class="language-r">#' @importFrom stats predict</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">455</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">60</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x &lt;- drop_elements(x, n)</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">456</td>
+                    <td class="num">61</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">457</td>
+                    <td class="num">62</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # remainder is subject</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">458</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">63</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  subject &lt;- as.character(x)</pre>
+                      <pre class="language-r">#' predict(object, newdata = fev_data)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">459</td>
+                    <td class="num">64</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">predict.mmrm_tmb &lt;- function(object,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">460</td>
-                    <td class="coverage">273<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">65</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cov_struct(type = type, visits = visits, group = group, subject = subject)</pre>
+                      <pre class="language-r">                             newdata,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">461</td>
+                    <td class="num">66</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">                             se.fit = FALSE, # nolint</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/interop-emmeans.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">67</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Support for `emmeans`</pre>
+                      <pre class="language-r">                             interval = c("none", "confidence", "prediction"),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">68</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                             level = 0.95,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">69</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">                             nsim = 1000L,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                             conditional = FALSE,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">71</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This package includes methods that allow `mmrm` objects to be used</pre>
+                      <pre class="language-r">                             ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">6</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">72</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' with the `emmeans` package. `emmeans` computes estimated marginal means</pre>
+                      <pre class="language-r">  if (missing(newdata)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">7</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">73</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' (also called least-square means) for the coefficients of the MMRM.</pre>
+                      <pre class="language-r">    newdata &lt;- object$data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">74</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' We can also e.g. obtain differences between groups by applying</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">9</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">75</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' [`pairs()`][emmeans::pairs.emmGrid()] on the object returned</pre>
+                      <pre class="language-r">  assert_data_frame(newdata)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">10</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">76</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' by [emmeans::emmeans()].</pre>
+                      <pre class="language-r">  orig_row_names &lt;- row.names(newdata)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">77</td>
+                    <td class="coverage">45<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_flag(se.fit)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">78</td>
+                    <td class="coverage">45<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_number(level, lower = 0, upper = 1)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">79</td>
+                    <td class="coverage">45<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">80</td>
+                    <td class="coverage">45<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_flag(conditional)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">81</td>
+                    <td class="coverage">45<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  interval &lt;- match.arg(interval)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">82</td>
+                    <td class="coverage">45<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  formula_parts &lt;- object$formula_parts</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">83</td>
+                    <td class="coverage">45<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  if (any(object$tmb_data$x_cols_aliased)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">11</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">84</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    warning(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">12</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">85</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">      "In fitted object there are co-linear variables and therefore dropped terms, ",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">13</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">86</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit &lt;- mmrm(</pre>
+                      <pre class="language-r">      "and this could lead to incorrect prediction on new data."</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">14</td>
+                    <td class="num">87</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">15</td>
+                    <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   data = fev_data</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">16</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">89</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' )</pre>
+                      <pre class="language-r">  colnames &lt;- names(Filter(isFALSE, object$tmb_data$x_cols_aliased))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">17</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">90</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' if (require(emmeans)) {</pre>
+                      <pre class="language-r">  if (!conditional &amp;&amp; interval %in% c("none", "confidence")) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">18</td>
+                    <td class="num">91</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   emmeans(fit, ~ ARMCD | AVISIT)</pre>
+                      <pre class="language-r">    # model.matrix always return a complete matrix (no NA allowed)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">19</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">92</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   pairs(emmeans(fit, ~ ARMCD | AVISIT), reverse = TRUE)</pre>
+                      <pre class="language-r">    x_mat &lt;- stats::model.matrix(object, data = newdata, use_response = FALSE)[, colnames, drop = FALSE]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">20</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">93</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' }</pre>
+                      <pre class="language-r">    x_mat_full &lt;- matrix(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">21</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">94</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @name emmeans_support</pre>
+                      <pre class="language-r">      NA,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">22</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">95</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">NULL</pre>
+                      <pre class="language-r">      nrow = nrow(newdata), ncol = ncol(x_mat),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">23</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">96</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      dimnames = list(row.names(newdata), colnames(x_mat))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">24</td>
+                    <td class="num">97</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Returns a `data.frame` for `emmeans` Purposes</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">25</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">98</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    x_mat_full[row.names(x_mat), ] &lt;- x_mat</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">26</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">99</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @seealso See [emmeans::recover_data()] for background.</pre>
+                      <pre class="language-r">    predictions &lt;- (x_mat_full %*% component(object, "beta_est"))[, 1]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">27</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">100</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    predictions_raw &lt;- stats::setNames(rep(NA_real_, nrow(newdata)), row.names(newdata))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">28</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">101</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @noRd</pre>
+                      <pre class="language-r">    predictions_raw[names(predictions)] &lt;- predictions</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">29</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">102</td>
+                    <td class="coverage">27<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">recover_data.mmrm &lt;- function(object, ...) { # nolint</pre>
+                      <pre class="language-r">    if (identical(interval, "none")) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">30</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">103</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fun_call &lt;- stats::getCall(object)</pre>
+                      <pre class="language-r">      return(predictions_raw)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">31</td>
+                    <td class="num">104</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # subject_var is excluded because it should not contain fixed effect.</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">32</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">105</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # visit_var is not excluded because emmeans can provide marginal mean</pre>
+                      <pre class="language-r">    se &lt;- switch(interval,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">33</td>
+                    <td class="num">106</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # by each visit if visit_var is not spatial.</pre>
+                      <pre class="language-r">      # can be NA if there are aliased cols</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">34</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">107</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  model_frame &lt;- stats::model.frame(</pre>
+                      <pre class="language-r">      "confidence" = diag(x_mat_full %*% component(object, "beta_vcov") %*% t(x_mat_full)),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">35</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">108</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    object,</pre>
+                      <pre class="language-r">      "none" = NA_real_</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">109</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">36</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">110</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    include = c(</pre>
+                      <pre class="language-r">    res &lt;- cbind(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">37</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">111</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (!object$formula_parts$is_spatial) "visit_var" else NULL,</pre>
+                      <pre class="language-r">      fit = predictions, se = se,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">38</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">112</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "response_var", "group_var"</pre>
+                      <pre class="language-r">      lwr = predictions - stats::qnorm(1 - level / 2) * se, upr = predictions + stats::qnorm(1 - level / 2) * se</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">39</td>
+                    <td class="num">113</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">40</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">114</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    if (!se.fit) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">41</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">115</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  model_terms &lt;- stats::delete.response(stats::terms(model_frame))</pre>
+                      <pre class="language-r">      res &lt;- res[, setdiff(colnames(res), "se")]</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">42</td>
-                    <td class="coverage">13<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">116</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  emmeans::recover_data(</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">43</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">117</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    fun_call,</pre>
+                      <pre class="language-r">    res_raw &lt;- matrix(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">44</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">118</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    trms = model_terms,</pre>
+                      <pre class="language-r">      NA_real_,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">45</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">119</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    na.action = "na.omit",</pre>
+                      <pre class="language-r">      ncol = ncol(res), nrow = nrow(newdata),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">46</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">120</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    frame = model_frame,</pre>
+                      <pre class="language-r">      dimnames = list(row.names(newdata), colnames(res))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">47</td>
+                    <td class="num">121</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ...</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">48</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">122</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    res_raw[row.names(res), ] &lt;- res</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">49</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">123</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    return(res_raw)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">124</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">51</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">125</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Returns a List of Model Details for `emmeans` Purposes</pre>
+                      <pre class="language-r">  tmb_data &lt;- h_mmrm_tmb_data(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">52</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">126</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    formula_parts, newdata,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">53</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">127</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @seealso See [emmeans::emm_basis()] for background.</pre>
+                      <pre class="language-r">    weights = rep(1, nrow(newdata)),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">54</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">128</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    reml = TRUE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">55</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">129</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @noRd</pre>
+                      <pre class="language-r">    singular = "keep",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">56</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">130</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">emm_basis.mmrm &lt;- function(object, # nolint</pre>
+                      <pre class="language-r">    drop_visit_levels = FALSE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">57</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">131</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                           trms,</pre>
+                      <pre class="language-r">    allow_na_response = TRUE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">58</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">132</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                           xlev,</pre>
+                      <pre class="language-r">    drop_levels = FALSE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">59</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">133</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                           grid,</pre>
+                      <pre class="language-r">    xlev = component(object, "xlev"),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">134</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    contrasts = component(object, "contrasts")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">60</td>
+                    <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                           ...) {</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">61</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">136</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  model_frame &lt;- stats::model.frame(trms, grid, na.action = stats::na.pass, xlev = xlev)</pre>
+                      <pre class="language-r">  tmb_data$x_matrix &lt;- tmb_data$x_matrix[, colnames, drop = FALSE]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">62</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">137</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  contrasts &lt;- component(object, "contrasts")</pre>
+                      <pre class="language-r">  predictions &lt;- h_get_prediction(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">63</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">138</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  model_mat &lt;- stats::model.matrix(trms, model_frame, contrasts.arg = contrasts)</pre>
+                      <pre class="language-r">    tmb_data, object$theta_est, object$beta_est, component(object, "beta_vcov")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">64</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">139</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  beta_hat &lt;- component(object, "beta_est")</pre>
+                      <pre class="language-r">  )$prediction</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">65</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">140</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  nbasis &lt;- if (length(beta_hat) &lt; ncol(model_mat)) {</pre>
+                      <pre class="language-r">  res &lt;- cbind(fit = rep(NA_real_, nrow(newdata)))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">66</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">141</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    kept &lt;- match(names(beta_hat), colnames(model_mat))</pre>
+                      <pre class="language-r">  new_order &lt;- match(row.names(tmb_data$full_frame), orig_row_names)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">67</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">142</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    beta_hat &lt;- NA * model_mat[1L, ]</pre>
+                      <pre class="language-r">  res[new_order, "fit"] &lt;- predictions[, "fit"]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">68</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">143</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    beta_hat[kept] &lt;- component(object, "beta_est")</pre>
+                      <pre class="language-r">  se &lt;- switch(interval,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">69</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">144</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    orig_model_mat &lt;- stats::model.matrix(</pre>
+                      <pre class="language-r">    "confidence" = sqrt(predictions[, "conf_var"]),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">70</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">145</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      trms,</pre>
+                      <pre class="language-r">    "prediction" = sqrt(h_get_prediction_variance(object, nsim, tmb_data)),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">71</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">146</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      stats::model.frame(</pre>
+                      <pre class="language-r">    "none" = NULL</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">147</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">72</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">148</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        object,</pre>
+                      <pre class="language-r">  if (interval != "none") {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">73</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">149</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        include = c(</pre>
+                      <pre class="language-r">    res &lt;- cbind(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">74</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">150</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          if (!object$formula_parts$is_spatial) "visit_var" else NULL,</pre>
+                      <pre class="language-r">      res,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">75</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">151</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          "response_var", "group_var"</pre>
+                      <pre class="language-r">      se = NA_real_</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">76</td>
+                    <td class="num">152</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        )</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">77</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">153</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ),</pre>
+                      <pre class="language-r">    res[new_order, "se"] &lt;- se</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">78</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">154</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      contrasts.arg = contrasts</pre>
+                      <pre class="language-r">    alpha &lt;- 1 - level</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">79</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">155</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">    z &lt;- stats::qnorm(1 - alpha / 2) * res[, "se"]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">80</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">156</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    estimability::nonest.basis(orig_model_mat)</pre>
+                      <pre class="language-r">    res &lt;- cbind(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">81</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">157</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">      res,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">82</td>
+                    <td class="num">158</td>
                     <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    estimability::all.estble</pre>
+                      <pre class="language-r">      lwr = res[, "fit"] - z,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">159</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      upr = res[, "fit"] + z</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">83</td>
+                    <td class="num">160</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">84</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">161</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    if (!se.fit) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="missed">
+                    <td class="num">162</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  dfargs &lt;- list(object = object)</pre>
+                      <pre class="language-r">      res &lt;- res[, setdiff(colnames(res), "se")]</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">85</td>
-                    <td class="coverage">13<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">163</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  dffun &lt;- function(k, dfargs) {</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">86</td>
-                    <td class="coverage">113<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">164</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mmrm::df_md(dfargs$object, contrast = k)$denom_df</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">87</td>
+                    <td class="num">165</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  # Use original names.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">88</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">166</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">  row.names(res) &lt;- orig_row_names</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">89</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">167</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    X = model_mat,</pre>
+                      <pre class="language-r">  if (ncol(res) == 1) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">168</td>
+                    <td class="coverage">11<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    bhat = beta_hat,</pre>
+                      <pre class="language-r">    res &lt;- res[, "fit"]</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">13<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">169</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    nbasis = nbasis,</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">170</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    V = component(object, "beta_vcov"),</pre>
+                      <pre class="language-r">  return(res)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">13<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">171</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    dffun = dffun,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">94</td>
-                    <td class="coverage">13<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">172</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    dfargs = dfargs</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">95</td>
+                    <td class="num">173</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' Get Prediction</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">96</td>
+                    <td class="num">174</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/satterthwaite.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">175</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Obtain List of Jacobian Matrix Entries for Covariance Matrix</pre>
+                      <pre class="language-r">#' @description Get predictions with given `data`, `theta`, `beta`, `beta_vcov`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">176</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">177</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Obtain the Jacobian matrices given the covariance function and variance parameters.</pre>
+                      <pre class="language-r">#' @details See `predict` function in `predict.cpp` which is called internally.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">178</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">179</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr produced by [h_mmrm_tmb_data()].</pre>
+                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">180</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param theta_est (`numeric`)\cr variance parameters point estimate.</pre>
+                      <pre class="language-r">#' @param theta (`numeric`)\cr theta value.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">181</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param beta_vcov (`matrix`)\cr vairance covariance matrix of coefficients.</pre>
+                      <pre class="language-r">#' @param beta (`numeric`)\cr beta value.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">182</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param beta_vcov (`matrix`)\cr beta_vcov matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">183</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with one element per variance parameter containing a matrix</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">184</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   of the same dimensions as the covariance matrix. The values are the derivatives</pre>
+                      <pre class="language-r">#' @return List with:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">185</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   with regards to this variance parameter.</pre>
+                      <pre class="language-r">#' - `prediction`: Matrix with columns `fit`, `conf_var`, and `var`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">186</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' - `covariance`: List with subject specific covariance matrices.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">13</td>
+                    <td class="num">187</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' - `index`: List of zero-based subject indices.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">14</td>
+                    <td class="num">188</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_jac_list &lt;- function(tmb_data,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">15</td>
+                    <td class="num">189</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                       theta_est,</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">190</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                       beta_vcov) {</pre>
+                      <pre class="language-r">h_get_prediction &lt;- function(tmb_data, theta, beta, beta_vcov) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">17</td>
-                    <td class="coverage">82<em>x</em></td>
+                    <td class="num">191</td>
+                    <td class="coverage">1696<em>x</em></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">18</td>
-                    <td class="coverage">82<em>x</em></td>
+                    <td class="num">192</td>
+                    <td class="coverage">1696<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(theta_est)</pre>
+                      <pre class="language-r">  assert_numeric(theta)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">19</td>
-                    <td class="coverage">82<em>x</em></td>
+                    <td class="num">193</td>
+                    <td class="coverage">1696<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(beta_vcov)</pre>
+                      <pre class="language-r">  n_beta &lt;- ncol(tmb_data$x_matrix)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">20</td>
-                    <td class="coverage">82<em>x</em></td>
+                    <td class="num">194</td>
+                    <td class="coverage">1696<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  .Call(`_mmrm_get_jacobian`, PACKAGE = "mmrm", tmb_data, theta_est, beta_vcov)</pre>
+                      <pre class="language-r">  assert_numeric(beta, finite = TRUE, any.missing = FALSE, len = n_beta)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">195</td>
+                    <td class="coverage">1696<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_matrix(beta_vcov, mode = "numeric", any.missing = FALSE, nrows = n_beta, ncols = n_beta)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">196</td>
+                    <td class="coverage">1696<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  .Call(`_mmrm_predict`, PACKAGE = "mmrm", tmb_data, theta, beta, beta_vcov)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">21</td>
+                    <td class="num">197</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">22</td>
+                    <td class="num">198</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">23</td>
+                    <td class="num">199</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Quadratic Form Calculations</pre>
+                      <pre class="language-r">#' Get Prediction Variance</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">24</td>
+                    <td class="num">200</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">25</td>
+                    <td class="num">201</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description These helpers are mainly for easier readability and slightly better efficiency</pre>
+                      <pre class="language-r">#' @description Get prediction variance with given fit, `tmb_data` with the Monte Carlo sampling method.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">26</td>
+                    <td class="num">202</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' of the quadratic forms used in the Satterthwaite calculations.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">27</td>
+                    <td class="num">203</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">28</td>
+                    <td class="num">204</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param center (`matrix`)\cr square numeric matrix with the same dimensions as</pre>
+                      <pre class="language-r">#' @param nsim (`count`)\cr number of samples.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">29</td>
+                    <td class="num">205</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `x` as the center of the quadratic form.</pre>
+                      <pre class="language-r">#' @param tmb_data (`mmrm_tmb_data`)\cr object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">30</td>
+                    <td class="num">206</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">31</td>
+                    <td class="num">207</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @name h_quad_form</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">32</td>
+                    <td class="num">208</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">NULL</pre>
+                      <pre class="language-r">h_get_prediction_variance &lt;- function(object, nsim, tmb_data) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">33</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">209</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">34</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">210</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn h_quad_form calculates the number `vec %*% center %*% t(vec)`</pre>
+                      <pre class="language-r">  assert_class(tmb_data, "mmrm_tmb_data")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">35</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">211</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   as a numeric (not a matrix).</pre>
+                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">36</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">212</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  theta_chol &lt;- chol(object$theta_vcov)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">37</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">213</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param vec (`numeric`)\cr interpreted as a row vector.</pre>
+                      <pre class="language-r">  n_theta &lt;- length(object$theta_est)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">38</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">214</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  res &lt;- replicate(nsim, {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">39</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">215</td>
+                    <td class="coverage">1150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    z &lt;- stats::rnorm(n = n_theta)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">40</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">216</td>
+                    <td class="coverage">1150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_quad_form_vec &lt;- function(vec, center) {</pre>
+                      <pre class="language-r">    theta_sample &lt;- object$theta_est + z %*% theta_chol</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">41</td>
-                    <td class="coverage">5607<em>x</em></td>
+                    <td class="num">217</td>
+                    <td class="coverage">1150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  vec &lt;- as.vector(vec)</pre>
+                      <pre class="language-r">    cond_beta_results &lt;- object$tmb_object$report(theta_sample)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">42</td>
-                    <td class="coverage">5607<em>x</em></td>
+                    <td class="num">218</td>
+                    <td class="coverage">1150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(vec, any.missing = FALSE)</pre>
+                      <pre class="language-r">    beta_mean &lt;- cond_beta_results$beta</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">43</td>
-                    <td class="coverage">5607<em>x</em></td>
+                    <td class="num">219</td>
+                    <td class="coverage">1150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(</pre>
+                      <pre class="language-r">    beta_cov &lt;- cond_beta_results$beta_vcov</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">44</td>
-                    <td class="coverage">5607<em>x</em></td>
+                    <td class="num">220</td>
+                    <td class="coverage">1150<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    center,</pre>
+                      <pre class="language-r">    h_get_prediction(tmb_data, theta_sample, beta_mean, beta_cov)$prediction</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">45</td>
-                    <td class="coverage">5607<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">221</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "numeric",</pre>
+                      <pre class="language-r">  })</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">46</td>
-                    <td class="coverage">5607<em>x</em></td>
+                    <td class="num">222</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    any.missing = FALSE,</pre>
+                      <pre class="language-r">  mean_of_var &lt;- rowMeans(res[, "var", ])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">47</td>
-                    <td class="coverage">5607<em>x</em></td>
+                    <td class="num">223</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    nrows = length(vec),</pre>
+                      <pre class="language-r">  var_of_mean &lt;- apply(res[, "fit", ], 1, stats::var)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">48</td>
-                    <td class="coverage">5607<em>x</em></td>
+                    <td class="num">224</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ncols = length(vec)</pre>
+                      <pre class="language-r">  mean_of_var + var_of_mean</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">49</td>
+                    <td class="num">225</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">226</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">51</td>
-                    <td class="coverage">5607<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">227</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  sum(vec * (center %*% vec))</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the model frame.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">52</td>
+                    <td class="num">228</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @param data (`data.frame`)\cr object in which to construct the frame.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">53</td>
+                    <td class="num">229</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @param include (`character`)\cr names of variable types to include.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">230</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn h_quad_form calculates the quadratic form `mat %*% center %*% t(mat)`</pre>
+                      <pre class="language-r">#'   Must be `NULL` or one or more of `c("subject_var", "visit_var", "group_var", "response_var")`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">55</td>
+                    <td class="num">231</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   as a matrix, the result is square and has dimensions identical to the number</pre>
+                      <pre class="language-r">#' @param full (`flag`)\cr indicator whether to return full model frame (deprecated).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">56</td>
+                    <td class="num">232</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   of rows in `mat`.</pre>
+                      <pre class="language-r">#' @param na.action (`string`)\cr na action.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">57</td>
+                    <td class="num">233</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @importFrom stats model.frame</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">58</td>
+                    <td class="num">234</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param mat (`matrix`)\cr numeric matrix to be multiplied left and right of</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">59</td>
+                    <td class="num">235</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `center`, therefore needs to have as many columns as there are rows and columns</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">60</td>
+                    <td class="num">236</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   in `center`.</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">61</td>
+                    <td class="num">237</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' `include` argument controls the variables the returned model frame will include.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">62</td>
+                    <td class="num">238</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' Possible options are "response_var", "subject_var", "visit_var" and "group_var", representing the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">63</td>
+                    <td class="num">239</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_quad_form_mat &lt;- function(mat, center) {</pre>
+                      <pre class="language-r">#' response variable, subject variable, visit variable or group variable.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">64</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">240</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(mat, mode = "numeric", any.missing = FALSE, min.cols = 1L)</pre>
+                      <pre class="language-r">#' `character` values in new data will always be factorized according to the data in the fit</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">65</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">241</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(</pre>
+                      <pre class="language-r">#' to avoid mismatched in levels or issues in `model.matrix`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">66</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">242</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    center,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">67</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">243</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "numeric",</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">68</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">244</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    any.missing = FALSE,</pre>
+                      <pre class="language-r">#' # Model frame:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">69</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">245</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    nrows = ncol(center),</pre>
+                      <pre class="language-r">#' model.frame(object)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">70</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">246</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ncols = ncol(center)</pre>
+                      <pre class="language-r">#' model.frame(object, include = "subject_var")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">71</td>
+                    <td class="num">247</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">model.frame.mmrm_tmb &lt;- function(formula, data, include = c("subject_var", "visit_var", "group_var", "response_var"),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">72</td>
-                    <td class="coverage">119<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">248</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  mat %*% tcrossprod(center, mat)</pre>
+                      <pre class="language-r">                                 full, na.action = "na.omit", ...) { # nolint</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">73</td>
+                    <td class="num">249</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  # Construct updated formula and data arguments.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">74</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">250</td>
+                    <td class="coverage">46<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  lst_formula_and_data &lt;-</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">75</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">251</td>
+                    <td class="coverage">46<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Computation of a Gradient Given Jacobian and Contrast Vector</pre>
+                      <pre class="language-r">    h_construct_model_frame_inputs(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">76</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">252</td>
+                    <td class="coverage">46<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      formula = formula,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">77</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">253</td>
+                    <td class="coverage">46<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      data = data,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">254</td>
+                    <td class="coverage">46<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">      include = include,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">255</td>
+                    <td class="coverage">46<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Computes the gradient of a linear combination of `beta` given the Jacobian matrix and</pre>
+                      <pre class="language-r">      full = full</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">78</td>
+                    <td class="num">256</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' variance parameters.</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">79</td>
+                    <td class="num">257</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  # Only if include is default (full) and also data is missing, and also na.action is na.omit we will</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">80</td>
+                    <td class="num">258</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param jac_list (`list`)\cr Jacobian list produced e.g. by [h_jac_list()].</pre>
+                      <pre class="language-r">  # use the model frame from the tmb_data.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">81</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">259</td>
+                    <td class="coverage">46<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param contrast (`numeric`)\cr contrast vector, which needs to have the</pre>
+                      <pre class="language-r">  include_choice &lt;- c("subject_var", "visit_var", "group_var", "response_var")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">82</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">260</td>
+                    <td class="coverage">46<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   same number of elements as there are rows and columns in each element of</pre>
+                      <pre class="language-r">  if (missing(data) &amp;&amp; setequal(include, include_choice) &amp;&amp; identical(h_get_na_action(na.action), stats::na.omit)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">83</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">261</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `jac_list`.</pre>
+                      <pre class="language-r">    ret &lt;- formula$tmb_data$full_frame</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">84</td>
+                    <td class="num">262</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    # Remove weights column.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">85</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">263</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Numeric vector which contains the quadratic forms of each element of</pre>
+                      <pre class="language-r">    ret[, "(weights)"] &lt;- NULL</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">86</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">264</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `jac_list` with the `contrast` vector.</pre>
+                      <pre class="language-r">    ret</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">87</td>
+                    <td class="num">265</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">88</td>
+                    <td class="num">266</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    # Construct data frame to return to users.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">89</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">267</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_gradient &lt;- function(jac_list, contrast) {</pre>
+                      <pre class="language-r">    ret &lt;-</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">491<em>x</em></td>
+                    <td class="num">268</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(jac_list)</pre>
+                      <pre class="language-r">      stats::model.frame(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">491<em>x</em></td>
+                    <td class="num">269</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast)</pre>
+                      <pre class="language-r">        formula = lst_formula_and_data$formula,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">92</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">270</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">        data = h_get_na_action(na.action)(lst_formula_and_data$data),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">491<em>x</em></td>
+                    <td class="num">271</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  vapply(</pre>
+                      <pre class="language-r">        na.action = na.action,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">94</td>
-                    <td class="coverage">491<em>x</em></td>
+                    <td class="num">272</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    jac_list,</pre>
+                      <pre class="language-r">        xlev = stats::.getXlevels(terms(formula), formula$tmb_data$full_frame)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">95</td>
-                    <td class="coverage">491<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">273</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_quad_form_vec,</pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">96</td>
-                    <td class="coverage">491<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">274</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    vec = contrast,</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">97</td>
-                    <td class="coverage">491<em>x</em></td>
+                    <td class="num">275</td>
+                    <td class="coverage">45<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    numeric(1L)</pre>
+                      <pre class="language-r">  ret</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">98</td>
+                    <td class="num">276</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">99</td>
+                    <td class="num">277</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">100</td>
+                    <td class="num">278</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">101</td>
+                    <td class="num">279</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Satterthwaite Degrees of Freedom for One-Dimensional Contrast</pre>
+                      <pre class="language-r">#' Construction of Model Frame Formula and Data Inputs</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">102</td>
+                    <td class="num">280</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">103</td>
+                    <td class="num">281</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_1d()] if method is</pre>
+                      <pre class="language-r">#' @description</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">104</td>
+                    <td class="num">282</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' "Satterthwaite".</pre>
+                      <pre class="language-r">#' Input formulas are converted from mmrm-style to a style compatible</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">105</td>
+                    <td class="num">283</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' with default [stats::model.frame()] and [stats::model.matrix()] methods.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">106</td>
+                    <td class="num">284</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">107</td>
+                    <td class="num">285</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param contrast (`numeric`)\cr contrast vector. Note that this should not include</pre>
+                      <pre class="language-r">#' The full formula is returned so we can construct, for example, the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">108</td>
+                    <td class="num">286</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
+                      <pre class="language-r">#' `model.frame()` including all columns as well as the requested subset.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">109</td>
+                    <td class="num">287</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   actually estimated coefficients.</pre>
+                      <pre class="language-r">#' The full set is used to identify rows to include in the reduced model frame.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">110</td>
+                    <td class="num">288</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">111</td>
+                    <td class="num">289</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with `est`, `se`, `df`, `t_stat` and `p_val`.</pre>
+                      <pre class="language-r">#' @param formula (`mmrm`)\cr mmrm fit object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">112</td>
+                    <td class="num">290</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' @param data optional data frame that will be</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">113</td>
+                    <td class="num">291</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_1d_sat &lt;- function(object, contrast) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">114</td>
-                    <td class="coverage">456<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">115</td>
-                    <td class="coverage">456<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  contrast &lt;- as.numeric(contrast)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">116</td>
-                    <td class="coverage">456<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">#'   passed to `model.frame()` or `model.matrix()`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">117</td>
+                    <td class="num">292</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">118</td>
-                    <td class="coverage">456<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- if (identical(object$vcov, "Asymptotic")) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">119</td>
-                    <td class="coverage">444<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    grad &lt;- h_gradient(component(object, "jac_list"), contrast)</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">120</td>
-                    <td class="coverage">444<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    v_num &lt;- 2 * h_quad_form_vec(contrast, component(object, "beta_vcov"))^2</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">121</td>
-                    <td class="coverage">444<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    v_denom &lt;- h_quad_form_vec(grad, component(object, "theta_vcov"))</pre>
+                      <pre class="language-r">#' @param include (`character`)\cr names of variable to include</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">122</td>
-                    <td class="coverage">444<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">293</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    v_num / v_denom</pre>
+                      <pre class="language-r">#' @param full (`flag`)\cr indicator whether to return full model frame (deprecated).</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">123</td>
-                    <td class="coverage">456<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">294</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else if (object$vcov %in% c("Empirical", "Empirical-Jackknife", "Empirical-Bias-Reduced")) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">124</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">295</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    contrast_matrix &lt;- Matrix::.bdiag(rep(list(matrix(contrast, nrow = 1)), component(object, "n_subjects")))</pre>
+                      <pre class="language-r">#' @return named list with four elements:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">125</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">296</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    contrast_matrix &lt;- as.matrix(contrast_matrix)</pre>
+                      <pre class="language-r">#' - `"formula"`: the formula including the columns requested in the `include=` argument.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">126</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">297</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    g_matrix &lt;- h_quad_form_mat(contrast_matrix, object$empirical_df_mat)</pre>
+                      <pre class="language-r">#' - `"data"`: a data frame including all columns needed in the formula.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">127</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">298</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    h_tr(g_matrix)^2 / sum(g_matrix^2)</pre>
+                      <pre class="language-r">#'   full formula are identical</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">128</td>
+                    <td class="num">299</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">129</td>
+                    <td class="num">300</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">h_construct_model_frame_inputs &lt;- function(formula,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">130</td>
-                    <td class="coverage">456<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">301</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_test_1d(object, contrast, df)</pre>
+                      <pre class="language-r">                                           data,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">131</td>
+                    <td class="num">302</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">                                           include,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">132</td>
+                    <td class="num">303</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">                                           include_choice = c("subject_var", "visit_var", "group_var", "response_var"),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">133</td>
+                    <td class="num">304</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculating Denominator Degrees of Freedom for the Multi-Dimensional Case</pre>
+                      <pre class="language-r">                                           full) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">134</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">305</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  if (!missing(full) &amp;&amp; identical(full, TRUE)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">135</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">306</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Calculates the degrees of freedom for multi-dimensional contrast.</pre>
+                      <pre class="language-r">    lifecycle::deprecate_warn("0.3", "model.frame.mmrm_tmb(full)")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">136</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">307</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    include &lt;- include_choice</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">137</td>
+                    <td class="num">308</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param t_stat_df (`numeric`)\cr `n` t-statistic derived degrees of freedom.</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">138</td>
+                    <td class="num">309</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">139</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">310</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Usually the calculation is returning `2 * E / (E - n)` where</pre>
+                      <pre class="language-r">  assert_class(formula, classes = "mmrm_tmb")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">140</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">311</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `E` is the sum of `t / (t - 2)` over all `t_stat_df` values `t`.</pre>
+                      <pre class="language-r">  assert_subset(include, include_choice)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">141</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">312</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  if (missing(data)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">142</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">313</td>
+                    <td class="coverage">256<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @note If the input values are two similar to each other then just the average</pre>
+                      <pre class="language-r">    data &lt;- formula$data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">143</td>
+                    <td class="num">314</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   of them is returned. If any of the inputs is not larger than 2 then 2 is</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">144</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">315</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   returned.</pre>
+                      <pre class="language-r">  assert_data_frame(data)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">145</td>
+                    <td class="num">316</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">146</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">317</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  drop_response &lt;- !"response_var" %in% include</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">147</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">318</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_md_denom_df &lt;- function(t_stat_df) {</pre>
+                      <pre class="language-r">  add_vars &lt;- unlist(formula$formula_parts[include])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">148</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="num">319</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(t_stat_df, min.len = 1L, lower = .Machine$double.xmin, any.missing = FALSE)</pre>
+                      <pre class="language-r">  new_formula &lt;- h_add_terms(formula$formula_parts$model_formula, add_vars, drop_response)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">149</td>
+                    <td class="num">320</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">150</td>
-                    <td class="coverage">24<em>x</em></td>
+                    <td class="num">321</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (test_scalar(t_stat_df)) {</pre>
+                      <pre class="language-r">  drop_response_full &lt;- !"response_var" %in% include_choice</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">151</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">322</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    t_stat_df</pre>
+                      <pre class="language-r">  add_vars_full &lt;- unlist(formula$formula_parts[include_choice])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">152</td>
-                    <td class="coverage">23<em>x</em></td>
+                    <td class="num">323</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else if (all(abs(diff(t_stat_df)) &lt; sqrt(.Machine$double.eps))) {</pre>
+                      <pre class="language-r">  new_formula_full &lt;-</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">153</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">324</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mean(t_stat_df)</pre>
+                      <pre class="language-r">    h_add_terms(formula$formula_parts$model_formula, add_vars_full, drop_response_full)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">154</td>
-                    <td class="coverage">22<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">325</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else if (any(t_stat_df &lt;= 2)) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">155</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">326</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    2</pre>
+                      <pre class="language-r">  # Update data based on the columns in the full formula return.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">156</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">327</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">  all_vars &lt;- all.vars(new_formula_full)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">157</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">328</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    e &lt;- sum(t_stat_df / (t_stat_df - 2))</pre>
+                      <pre class="language-r">  assert_names(colnames(data), must.include = all_vars)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">158</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">329</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    2 * e / (e - (length(t_stat_df)))</pre>
+                      <pre class="language-r">  data &lt;- data[, all_vars, drop = FALSE]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">159</td>
+                    <td class="num">330</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">160</td>
+                    <td class="num">331</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  # Return list with updated formula, data.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">161</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">332</td>
+                    <td class="coverage">280<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  list(</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">333</td>
+                    <td class="coverage">280<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    formula = new_formula,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">334</td>
+                    <td class="coverage">280<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    data = data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">162</td>
+                    <td class="num">335</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Creating F-Statistic Results from One-Dimensional Contrast</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">163</td>
+                    <td class="num">336</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">164</td>
+                    <td class="num">337</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Creates multi-dimensional result from one-dimensional contrast from [df_1d()].</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">165</td>
+                    <td class="num">338</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the model matrix.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">166</td>
+                    <td class="num">339</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr model fit.</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">167</td>
+                    <td class="num">340</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param contrast (`numeric`)\cr one-dimensional contrast.</pre>
+                      <pre class="language-r">#' @param use_response (`flag`)\cr whether to use the response for complete rows.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">168</td>
+                    <td class="num">341</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">169</td>
+                    <td class="num">342</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The one-dimensional degrees of freedom are calculated and then</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">170</td>
+                    <td class="num">343</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   based on that the p-value is calculated.</pre>
+                      <pre class="language-r">#' # Model matrix:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">171</td>
+                    <td class="num">344</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' model.matrix(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">172</td>
+                    <td class="num">345</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">model.matrix.mmrm_tmb &lt;- function(object, data, use_response = TRUE, ...) { # nolint</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">173</td>
+                    <td class="num">346</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_md_from_1d &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">  # Always return the utilized model matrix if data not provided.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">174</td>
-                    <td class="coverage">134<em>x</em></td>
+                    <td class="num">347</td>
+                    <td class="coverage">37<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  res_1d &lt;- h_df_1d_sat(object, contrast)</pre>
+                      <pre class="language-r">  if (missing(data)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">175</td>
-                    <td class="coverage">134<em>x</em></td>
+                    <td class="num">348</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">    return(object$tmb_data$x_matrix)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">176</td>
-                    <td class="coverage">134<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">349</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    num_df = 1,</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">177</td>
-                    <td class="coverage">134<em>x</em></td>
+                    <td class="num">350</td>
+                    <td class="coverage">34<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    denom_df = res_1d$df,</pre>
+                      <pre class="language-r">  stats::model.matrix(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">178</td>
-                    <td class="coverage">134<em>x</em></td>
+                    <td class="num">351</td>
+                    <td class="coverage">34<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    f_stat = res_1d$t_stat^2,</pre>
+                      <pre class="language-r">    h_add_terms(object$formula_parts$model_formula, NULL, drop_response = !use_response),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">179</td>
-                    <td class="coverage">134<em>x</em></td>
+                    <td class="num">352</td>
+                    <td class="coverage">34<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    p_val = stats::pf(q = res_1d$t_stat^2, df1 = 1, df2 = res_1d$df, lower.tail = FALSE)</pre>
+                      <pre class="language-r">    data = data,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">180</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">353</td>
+                    <td class="coverage">34<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    contrasts.arg = attr(object$tmb_data$x_matrix, "contrasts"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">181</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">354</td>
+                    <td class="coverage">34<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    xlev = component(object, "xlev"),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">182</td>
+                    <td class="num">355</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    ...</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">183</td>
+                    <td class="num">356</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Satterthwaite Degrees of Freedom for Multi-Dimensional Contrast</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">184</td>
+                    <td class="num">357</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">185</td>
+                    <td class="num">358</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_md()] if method is "Satterthwaite".</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">186</td>
+                    <td class="num">359</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the terms object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">187</td>
+                    <td class="num">360</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr the MMRM fit.</pre>
+                      <pre class="language-r">#' @importFrom stats model.frame</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">188</td>
+                    <td class="num">361</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param contrast (`matrix`)\cr numeric contrast matrix, if given a `numeric`</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">189</td>
+                    <td class="num">362</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   then this is coerced to a row vector. Note that this should not include</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">190</td>
+                    <td class="num">363</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   elements for singular coefficient estimates, i.e. only refer to the</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">191</td>
+                    <td class="num">364</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   actually estimated coefficients.</pre>
+                      <pre class="language-r">#' # terms:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">192</td>
+                    <td class="num">365</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' terms(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">193</td>
+                    <td class="num">366</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with `num_df`, `denom_df`, `f_stat` and `p_val` (2-sided p-value).</pre>
+                      <pre class="language-r">#' terms(object, include = "subject_var")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">194</td>
+                    <td class="num">367</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">terms.mmrm_tmb &lt;- function(x, include = "response_var", ...) { # nolint</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">195</td>
+                    <td class="num">368</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_md_sat &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">  # Construct updated formula and data arguments.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">196</td>
-                    <td class="coverage">151<em>x</em></td>
+                    <td class="num">369</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">  lst_formula_and_data &lt;-</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">197</td>
-                    <td class="coverage">151<em>x</em></td>
+                    <td class="num">370</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">    h_construct_model_frame_inputs(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">198</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">371</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Early return if we are in the one-dimensional case.</pre>
+                      <pre class="language-r">      formula = x,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">199</td>
-                    <td class="coverage">151<em>x</em></td>
+                    <td class="num">372</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (identical(nrow(contrast), 1L)) {</pre>
+                      <pre class="language-r">      include = include</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">200</td>
-                    <td class="coverage">132<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">373</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(h_df_md_from_1d(object, contrast))</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">201</td>
+                    <td class="num">374</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">202</td>
+                    <td class="num">375</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  # Use formula method for `terms()` to construct the mmrm terms object.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">203</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">376</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  contrast_cov &lt;- h_quad_form_mat(contrast, component(object, "beta_vcov"))</pre>
+                      <pre class="language-r">  stats::terms(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">204</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">377</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  eigen_cont_cov &lt;- eigen(contrast_cov)</pre>
+                      <pre class="language-r">    x = lst_formula_and_data$formula,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">205</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">378</td>
+                    <td class="coverage">231<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  eigen_cont_cov_vctrs &lt;- eigen_cont_cov$vectors</pre>
+                      <pre class="language-r">    data = lst_formula_and_data$data</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">206</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">379</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  eigen_cont_cov_vals &lt;- eigen_cont_cov$values</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">207</td>
+                    <td class="num">380</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">208</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">381</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  eps &lt;- sqrt(.Machine$double.eps)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">209</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">382</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tol &lt;- max(eps * eigen_cont_cov_vals[1], 0)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">210</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">383</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  rank_cont_cov &lt;- sum(eigen_cont_cov_vals &gt; tol)</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the attained log likelihood value.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">211</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">384</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(rank_cont_cov, lower = .Machine$double.xmin)</pre>
+                      <pre class="language-r">#' @importFrom stats logLik</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">212</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">385</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  rank_seq &lt;- seq_len(rank_cont_cov)</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">213</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">386</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  vctrs_cont_prod &lt;- crossprod(eigen_cont_cov_vctrs, contrast)[rank_seq, , drop = FALSE]</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">214</td>
+                    <td class="num">387</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' # Log likelihood given the estimated parameters:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">215</td>
+                    <td class="num">388</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Early return if rank 1.</pre>
+                      <pre class="language-r">#' logLik(object)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">216</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">389</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (identical(rank_cont_cov, 1L)) {</pre>
+                      <pre class="language-r">logLik.mmrm_tmb &lt;- function(object, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">217</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">390</td>
+                    <td class="coverage">50<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(h_df_md_from_1d(object, vctrs_cont_prod))</pre>
+                      <pre class="language-r">  -component(object, "neg_log_lik")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">218</td>
+                    <td class="num">391</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">219</td>
+                    <td class="num">392</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">220</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">393</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  t_squared_nums &lt;- drop(vctrs_cont_prod %*% object$beta_est)^2</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the used formula.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">221</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">394</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  t_squared_denoms &lt;- eigen_cont_cov_vals[rank_seq]</pre>
+                      <pre class="language-r">#' @importFrom stats formula</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">222</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">395</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  t_squared &lt;- t_squared_nums / t_squared_denoms</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">223</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">396</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  f_stat &lt;- sum(t_squared) / rank_cont_cov</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">224</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">397</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  t_stat_df_nums &lt;- 2 * eigen_cont_cov_vals^2</pre>
+                      <pre class="language-r">#' # Formula which was used:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">225</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">398</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  t_stat_df &lt;- if (identical(object$vcov, "Asymptotic")) {</pre>
+                      <pre class="language-r">#' formula(object)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">226</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">399</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    grads_vctrs_cont_prod &lt;- lapply(</pre>
+                      <pre class="language-r">formula.mmrm_tmb &lt;- function(x, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">227</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">400</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      rank_seq,</pre>
+                      <pre class="language-r">  x$formula_parts$formula</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">228</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">401</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      function(m) h_gradient(component(object, "jac_list"), contrast = vctrs_cont_prod[m, ])</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">229</td>
+                    <td class="num">402</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">230</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">403</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    t_stat_df_denoms &lt;- vapply(</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the variance-covariance matrix estimate</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">231</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">404</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      grads_vctrs_cont_prod,</pre>
+                      <pre class="language-r">#'   for the coefficients.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">232</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">405</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      h_quad_form_vec,</pre>
+                      <pre class="language-r">#' @importFrom stats vcov</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">233</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">406</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      center = component(object, "theta_vcov"),</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">234</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">407</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      numeric(1)</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">235</td>
+                    <td class="num">408</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' # Variance-covariance matrix estimate for coefficients:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">236</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">409</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    t_stat_df_nums / t_stat_df_denoms</pre>
+                      <pre class="language-r">#' vcov(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">237</td>
+                    <td class="num">410</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
+                      <pre class="language-r">vcov.mmrm_tmb &lt;- function(object, complete = TRUE, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">238</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">411</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    vapply(</pre>
+                      <pre class="language-r">  assert_flag(complete)</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">239</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">412</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      rank_seq,</pre>
+                      <pre class="language-r">  nm &lt;- if (complete) "beta_vcov_complete" else "beta_vcov"</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">240</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">413</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      function(m) {</pre>
+                      <pre class="language-r">  component(object, name = nm)</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">241</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">414</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        contrast_matrix &lt;- Matrix::.bdiag(</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">242</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">415</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          rep(list(vctrs_cont_prod[m, , drop = FALSE]), component(object, "n_subjects"))</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">243</td>
+                    <td class="num">416</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        )</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the variance-covariance matrix estimate</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">244</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">417</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        contrast_matrix &lt;- as.matrix(contrast_matrix)</pre>
+                      <pre class="language-r">#'   for the residuals.</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">245</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">418</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        g_matrix &lt;- h_quad_form_mat(contrast_matrix, object$empirical_df_mat)</pre>
+                      <pre class="language-r">#' @param sigma cannot be used (this parameter does not exist in MMRM).</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">246</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">419</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        h_tr(g_matrix)^2 / sum(g_matrix^2)</pre>
+                      <pre class="language-r">#' @importFrom nlme VarCorr</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">247</td>
+                    <td class="num">420</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      },</pre>
+                      <pre class="language-r">#' @export VarCorr</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">248</td>
-                    <td class="coverage">!</td>
+                  <tr class="never">
+                    <td class="num">421</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      FUN.VALUE = 0</pre>
+                      <pre class="language-r">#' @aliases VarCorr</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">249</td>
+                    <td class="num">422</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">250</td>
+                    <td class="num">423</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">251</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">424</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  denom_df &lt;- h_md_denom_df(t_stat_df)</pre>
+                      <pre class="language-r">#' # Variance-covariance matrix estimate for residuals:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">252</td>
+                    <td class="num">425</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' VarCorr(object)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">253</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">426</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">VarCorr.mmrm_tmb &lt;- function(x, sigma = NA, ...) { # nolint</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">254</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">427</td>
+                    <td class="coverage">10<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    num_df = rank_cont_cov,</pre>
+                      <pre class="language-r">  assert_scalar_na(sigma)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">255</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">428</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    denom_df = denom_df,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">256</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">429</td>
+                    <td class="coverage">10<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    f_stat = f_stat,</pre>
+                      <pre class="language-r">  component(x, name = "varcor")</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">257</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">430</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    p_val = stats::pf(q = f_stat, df1 = rank_cont_cov, df2 = denom_df, lower.tail = FALSE)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">258</td>
+                    <td class="num">431</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">259</td>
+                    <td class="num">432</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the deviance, which is defined here</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/component.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">433</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Component Access for `mmrm_tmb` Objects</pre>
+                      <pre class="language-r">#'   as twice the negative log likelihood, which can either be integrated</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">434</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   over the coefficients for REML fits or the usual one for ML fits.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">435</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">#' @importFrom stats deviance</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">436</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">437</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">438</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param name (`character`)\cr the component(s) to be retrieved.</pre>
+                      <pre class="language-r">#' # REML criterion (twice the negative log likelihood):</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">439</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The corresponding component of the object, see details.</pre>
+                      <pre class="language-r">#' deviance(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">440</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">deviance.mmrm_tmb &lt;- function(object, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">9</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">441</td>
+                    <td class="coverage">74<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details Available `component()` names are as follows:</pre>
+                      <pre class="language-r">  2 * component(object, "neg_log_lik")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">442</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `call`: low-level function call which generated the model.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">443</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `formula`: model formula.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">444</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `dataset`: data set name.</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the Akaike Information Criterion,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">13</td>
+                    <td class="num">445</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `cov_type`: covariance structure type.</pre>
+                      <pre class="language-r">#'   where the degrees of freedom are the number of variance parameters (`n_theta`).</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">14</td>
+                    <td class="num">446</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `n_theta`: number of parameters.</pre>
+                      <pre class="language-r">#'   If `corrected`, then this is multiplied with `m / (m - n_theta - 1)` where</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">15</td>
+                    <td class="num">447</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `n_subjects`: number of subjects.</pre>
+                      <pre class="language-r">#'   `m` is the number of observations minus the number of coefficients, or</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">448</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `n_timepoints`: number of modeled time points.</pre>
+                      <pre class="language-r">#'   `n_theta + 2` if it is smaller than that \insertCite{hurvich1989regression,burnham1998practical}{mmrm}.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">17</td>
+                    <td class="num">449</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `n_obs`: total number of observations.</pre>
+                      <pre class="language-r">#' @param corrected (`flag`)\cr whether corrected AIC should be calculated.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">18</td>
+                    <td class="num">450</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `reml`: was REML used (ML was used if `FALSE`).</pre>
+                      <pre class="language-r">#' @param k (`number`)\cr the penalty per parameter to be used; default `k = 2`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">19</td>
+                    <td class="num">451</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `neg_log_lik`: negative log likelihood.</pre>
+                      <pre class="language-r">#'   is the classical AIC.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">20</td>
+                    <td class="num">452</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `convergence`: convergence code from optimizer.</pre>
+                      <pre class="language-r">#' @importFrom stats AIC</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">21</td>
+                    <td class="num">453</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `conv_message`: message accompanying the convergence code.</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">22</td>
+                    <td class="num">454</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `evaluations`: number of function evaluations for optimization.</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">23</td>
+                    <td class="num">455</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `method`: Adjustment method which was used (for `mmrm` objects),</pre>
+                      <pre class="language-r">#' # AIC:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">24</td>
+                    <td class="num">456</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      otherwise `NULL` (for `mmrm_tmb` objects).</pre>
+                      <pre class="language-r">#' AIC(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">25</td>
+                    <td class="num">457</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `beta_vcov`: estimated variance-covariance matrix of coefficients</pre>
+                      <pre class="language-r">#' AIC(object, corrected = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">26</td>
+                    <td class="num">458</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      (excluding aliased coefficients). When Kenward-Roger/Empirical adjusted</pre>
+                      <pre class="language-r">#' @references</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">27</td>
+                    <td class="num">459</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      coefficients covariance matrix is used, the adjusted covariance matrix is returned (to still obtain the</pre>
+                      <pre class="language-r">#' - \insertRef{hurvich1989regression}{mmrm}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">28</td>
+                    <td class="num">460</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      original asymptotic covariance matrix use `object$beta_vcov`).</pre>
+                      <pre class="language-r">#' - \insertRef{burnham1998practical}{mmrm}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">29</td>
+                    <td class="num">461</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `beta_vcov_complete`: estimated variance-covariance matrix including</pre>
+                      <pre class="language-r">AIC.mmrm_tmb &lt;- function(object, corrected = FALSE, ..., k = 2) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">30</td>
+                    <td class="num">462</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      aliased coefficients with entries set to `NA`.</pre>
+                      <pre class="language-r">  # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">31</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">463</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `varcor`: estimated covariance matrix for residuals. If there are multiple</pre>
+                      <pre class="language-r">  assert_flag(corrected)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">32</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">464</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      groups, a named list of estimated covariance matrices for residuals will be</pre>
+                      <pre class="language-r">  assert_number(k, lower = 1)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">33</td>
+                    <td class="num">465</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      returned. The names are the group levels.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">34</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">466</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `theta_est`: estimated variance parameters.</pre>
+                      <pre class="language-r">  n_theta &lt;- length(component(object, "theta_est"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">35</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">467</td>
+                    <td class="coverage">44<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `beta_est`: estimated coefficients (excluding aliased coefficients).</pre>
+                      <pre class="language-r">  df &lt;- if (!corrected) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">36</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">468</td>
+                    <td class="coverage">43<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `beta_est_complete`: estimated coefficients including aliased coefficients</pre>
+                      <pre class="language-r">    n_theta</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">37</td>
+                    <td class="num">469</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      set to `NA`.</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">38</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">470</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `beta_aliased`: whether each coefficient was aliased (i.e. cannot be estimated)</pre>
+                      <pre class="language-r">    n_obs &lt;- length(component(object, "y_vector"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">39</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">471</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'      or not.</pre>
+                      <pre class="language-r">    n_beta &lt;- length(component(object, "beta_est"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">40</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">472</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `theta_vcov`:  estimated variance-covariance matrix of variance parameters.</pre>
+                      <pre class="language-r">    m &lt;- max(n_theta + 2, n_obs - n_beta)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">473</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    n_theta * (m / (m - n_theta - 1))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">41</td>
+                    <td class="num">474</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `x_matrix`: design matrix used (excluding aliased columns).</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">42</td>
+                    <td class="num">475</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `xlev`: 	a named list of character vectors giving the full set of levels to be assumed for each factor.</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">476</td>
+                    <td class="coverage">44<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  2 * component(object, "neg_log_lik") + k * df</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">43</td>
+                    <td class="num">477</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `contrasts`: a list of contrasts used for each factor.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">44</td>
+                    <td class="num">478</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `y_vector`: response vector used.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">45</td>
+                    <td class="num">479</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `jac_list`: Jacobian, see  [h_jac_list()] for details.</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods obtains the Bayesian Information Criterion,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">46</td>
+                    <td class="num">480</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' - `full_frame`: `data.frame` with `n` rows containing all variables needed in the model.</pre>
+                      <pre class="language-r">#'   which is using the natural logarithm of the number of subjects for the</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">47</td>
+                    <td class="num">481</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   penalty parameter `k`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">48</td>
+                    <td class="num">482</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @seealso In the `lme4` package there is a similar function `getME()`.</pre>
+                      <pre class="language-r">#' @importFrom stats BIC</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">49</td>
+                    <td class="num">483</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">484</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">51</td>
+                    <td class="num">485</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit &lt;- mmrm(</pre>
+                      <pre class="language-r">#' # BIC:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">52</td>
+                    <td class="num">486</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
+                      <pre class="language-r">#' BIC(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">53</td>
+                    <td class="num">487</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   data = fev_data</pre>
+                      <pre class="language-r">BIC.mmrm_tmb &lt;- function(object, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">488</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' )</pre>
+                      <pre class="language-r">  # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">55</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">489</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Get all available components.</pre>
+                      <pre class="language-r">  k &lt;- log(component(object, "n_subjects"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">56</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">490</td>
+                    <td class="coverage">21<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' component(fit)</pre>
+                      <pre class="language-r">  AIC(object, corrected = FALSE, k = k)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">57</td>
+                    <td class="num">491</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Get convergence code and message.</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">58</td>
+                    <td class="num">492</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' component(fit, c("convergence", "conv_message"))</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">59</td>
+                    <td class="num">493</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Get modeled formula as a string.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">60</td>
+                    <td class="num">494</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' component(fit, c("formula"))</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods prints the object.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">61</td>
+                    <td class="num">495</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">62</td>
+                    <td class="num">496</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @export</pre>
+                      <pre class="language-r">print.mmrm_tmb &lt;- function(x,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">63</td>
+                    <td class="num">497</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">component &lt;- function(object,</pre>
+                      <pre class="language-r">                           ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">64</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">498</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      name = c(</pre>
+                      <pre class="language-r">  cat("mmrm fit\n\n")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">65</td>
+                    <td class="num">499</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "cov_type", "subject_var", "n_theta", "n_subjects", "n_timepoints",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">66</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">500</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "n_obs", "beta_vcov", "beta_vcov_complete",</pre>
+                      <pre class="language-r">  h_print_call(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">67</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">501</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "varcor", "formula", "dataset", "n_groups",</pre>
+                      <pre class="language-r">    component(x, "call"), component(x, "n_obs"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">68</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">502</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "reml", "convergence", "evaluations", "method", "optimizer",</pre>
+                      <pre class="language-r">    component(x, "n_subjects"), component(x, "n_timepoints")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">69</td>
+                    <td class="num">503</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "conv_message", "call", "theta_est",</pre>
+                      <pre class="language-r">  )</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">504</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  h_print_cov(component(x, "cov_type"), component(x, "n_theta"), component(x, "n_groups"))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">70</td>
+                    <td class="num">505</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "beta_est", "beta_est_complete", "beta_aliased",</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">506</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cat("Inference:   ")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">507</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cat(ifelse(component(x, "reml"), "REML", "ML"))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">508</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cat("\n")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">509</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cat("Deviance:    ")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">510</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cat(deviance(x))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">71</td>
+                    <td class="num">511</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "x_matrix", "y_vector", "neg_log_lik", "jac_list", "theta_vcov",</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">512</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cat("\n\nCoefficients: ")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">513</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  n_singular_coefs &lt;- sum(component(x, "beta_aliased"))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">514</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  if (n_singular_coefs &gt; 0) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">515</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    cat("(", n_singular_coefs, " not defined because of singularities)", sep = "")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">72</td>
+                    <td class="num">516</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                        "full_frame", "xlev", "contrasts"</pre>
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">517</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  cat("\n")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">518</td>
+                    <td class="coverage">2<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  print(coef(x, complete = TRUE))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">73</td>
+                    <td class="num">519</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      )) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">74</td>
-                    <td class="coverage">5115<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">75</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">520</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  name &lt;- match.arg(name, several.ok = TRUE)</pre>
+                      <pre class="language-r">  cat("\nModel Inference Optimization:")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">76</td>
+                    <td class="num">521</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">77</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">522</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list_components &lt;- sapply(</pre>
+                      <pre class="language-r">  cat(ifelse(component(x, "convergence") == 0, "\nConverged", "\nFailed to converge"))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">78</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">523</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    X = name,</pre>
+                      <pre class="language-r">  cat(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">79</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">524</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    FUN = switch,</pre>
+                      <pre class="language-r">    " with code", component(x, "convergence"),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">80</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">525</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "call" = object$call,</pre>
+                      <pre class="language-r">    "and message:",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">81</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">526</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Strings.</pre>
+                      <pre class="language-r">    if (is.null(component(x, "conv_message"))) "No message provided." else tolower(component(x, "conv_message"))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">82</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">527</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "cov_type" = object$formula_parts$cov_type,</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">83</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">528</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "subject_var" = object$formula_parts$subject_var,</pre>
+                      <pre class="language-r">  cat("\n")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">84</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">529</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "formula" = deparse(object$call$formula),</pre>
+                      <pre class="language-r">  invisible(x)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">85</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">530</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "dataset" = object$call$data,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">86</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">531</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "reml" = object$reml,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">87</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">532</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "conv_message" = object$opt_details$message,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">88</td>
+                    <td class="num">533</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Numeric of length 1.</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods to obtain residuals - either unscaled ('response'), 'pearson' or 'normalized'.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">89</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">534</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "convergence" = object$opt_details$convergence,</pre>
+                      <pre class="language-r">#' @param type (`string`)\cr unscaled (`response`), `pearson` or `normalized`. Default is `response`,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">535</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "neg_log_lik" = object$neg_log_lik,</pre>
+                      <pre class="language-r">#' and this is the only type available for use with models with a spatial covariance structure.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">536</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "n_theta" = length(object$theta_est),</pre>
+                      <pre class="language-r">#' @importFrom stats residuals</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">537</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "n_subjects" = object$tmb_data$n_subjects,</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">538</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "n_timepoints" = object$tmb_data$n_visits,</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">94</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">539</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "n_obs" = length(object$tmb_data$y_vector),</pre>
+                      <pre class="language-r">#' # residuals:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">95</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">540</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "n_groups" = ifelse(is.list(object$cov), length(object$cov), 1L),</pre>
+                      <pre class="language-r">#' residuals(object, type = "response")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">96</td>
+                    <td class="num">541</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Numeric of length &gt; 1.</pre>
+                      <pre class="language-r">#' residuals(object, type = "pearson")</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">97</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">542</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "evaluations" = unlist(ifelse(is.null(object$opt_details$evaluations),</pre>
+                      <pre class="language-r">#' residuals(object, type = "normalized")</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">98</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">543</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      list(object$opt_details$counts),</pre>
+                      <pre class="language-r">#' @references</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">99</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">544</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      list(object$opt_details$evaluations)</pre>
+                      <pre class="language-r">#' - \insertRef{galecki2013linear}{mmrm}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">100</td>
+                    <td class="num">545</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )),</pre>
+                      <pre class="language-r">residuals.mmrm_tmb &lt;- function(object, type = c("response", "pearson", "normalized"), ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">101</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">546</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "method" = object$method,</pre>
+                      <pre class="language-r">  type &lt;- match.arg(type)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">102</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">547</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "optimizer" = object$optimizer,</pre>
+                      <pre class="language-r">  switch(type,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">103</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">548</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "beta_est" = object$beta_est,</pre>
+                      <pre class="language-r">    "response" = h_residuals_response(object),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">104</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">549</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "beta_est_complete" =</pre>
+                      <pre class="language-r">    "pearson" = h_residuals_pearson(object),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">105</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">550</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (any(object$tmb_data$x_cols_aliased)) {</pre>
+                      <pre class="language-r">    "normalized" = h_residuals_normalized(object)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">106</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">551</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        stats::setNames(</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">107</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">552</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          object$beta_est[names(object$tmb_data$x_cols_aliased)],</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">108</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">553</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          names(object$tmb_data$x_cols_aliased)</pre>
+                      <pre class="language-r">#' Calculate Pearson Residuals</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">109</td>
+                    <td class="num">554</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">110</td>
+                    <td class="num">555</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else {</pre>
+                      <pre class="language-r">#' This is used by [residuals.mmrm_tmb()] to calculate Pearson residuals.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">111</td>
-                    <td class="coverage">54<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">556</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        object$beta_est</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">112</td>
+                    <td class="num">557</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      },</pre>
+                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">113</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">558</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "beta_aliased" = object$tmb_data$x_cols_aliased,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">114</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">559</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "theta_est" = object$theta_est,</pre>
+                      <pre class="language-r">#' @return Vector of residuals.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">115</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">560</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "y_vector" = object$tmb_data$y_vector,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">116</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">561</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "jac_list" = object$jac_list,</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">117</td>
+                    <td class="num">562</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # Matrices.</pre>
+                      <pre class="language-r">h_residuals_pearson &lt;- function(object) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">118</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">563</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "beta_vcov" =</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">119</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">564</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (is.null(object$vcov) || identical(object$vcov, "Asymptotic")) {</pre>
+                      <pre class="language-r">  h_residuals_response(object) * object$tmb_object$report()$diag_cov_inv_sqrt</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">120</td>
-                    <td class="coverage">985<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">565</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        object$beta_vcov</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">121</td>
+                    <td class="num">566</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">122</td>
-                    <td class="coverage">66<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">567</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        object$beta_vcov_adj</pre>
+                      <pre class="language-r">#' Calculate normalized residuals</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">123</td>
+                    <td class="num">568</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      },</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">124</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">569</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "beta_vcov_complete" =</pre>
+                      <pre class="language-r">#' This is used by [residuals.mmrm_tmb()] to calculate normalized / scaled residuals.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">125</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">570</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (any(object$tmb_data$x_cols_aliased)) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">126</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">571</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        stats::.vcov.aliased(</pre>
+                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">127</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">572</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          aliased = object$tmb_data$x_cols_aliased,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">128</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">573</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          vc = component(object, "beta_vcov"),</pre>
+                      <pre class="language-r">#' @return Vector of residuals</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">129</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">574</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">          complete = TRUE</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">130</td>
+                    <td class="num">575</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        )</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">131</td>
+                    <td class="num">576</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else {</pre>
+                      <pre class="language-r">h_residuals_normalized &lt;- function(object) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">132</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">577</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        object$beta_vcov</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">578</td>
+                    <td class="coverage">8<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  object$tmb_object$report()$epsilonTilde</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">133</td>
+                    <td class="num">579</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      },</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">134</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">580</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "varcor" = object$cov,</pre>
+                      <pre class="language-r">#' Calculate response residuals.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">135</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">581</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "x_matrix" = object$tmb_data$x_matrix,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">136</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">582</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "xlev" = stats::.getXlevels(terms(object), object$tmb_data$full_frame),</pre>
+                      <pre class="language-r">#' This is used by [residuals.mmrm_tmb()] to calculate response residuals.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">137</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">583</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "contrasts" = attr(object$tmb_data$x_matrix, "contrasts"),</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">138</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">584</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "theta_vcov" = object$theta_vcov,</pre>
+                      <pre class="language-r">#' @param object (`mmrm_tmb`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">139</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">585</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "full_frame" = object$tmb_data$full_frame,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">140</td>
+                    <td class="num">586</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # If not found.</pre>
+                      <pre class="language-r">#' @return Vector of residuals</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">141</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">587</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "..foo.." =</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">142</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">588</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      stop(sprintf(</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">589</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">h_residuals_response &lt;- function(object) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">143</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">590</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "component '%s' is not available",</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm_tmb")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">144</td>
-                    <td class="coverage">5115<em>x</em></td>
+                    <td class="num">591</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        name, paste0(class(object), collapse = ", ")</pre>
+                      <pre class="language-r">  component(object, "y_vector") - unname(fitted(object))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">145</td>
+                    <td class="num">592</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )),</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">146</td>
-                    <td class="coverage">5115<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">593</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">594</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    simplify = FALSE</pre>
+                      <pre class="language-r">#' @describeIn mmrm_tmb_methods simulate responses from a fitted model according</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">147</td>
+                    <td class="num">595</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'   to the simulation `method`, returning a `data.frame` of dimension `[n, m]`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">148</td>
+                    <td class="num">596</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   where n is the number of rows in `newdata`,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">149</td>
-                    <td class="coverage">23<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">597</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (length(name) == 1) list_components[[1]] else list_components</pre>
+                      <pre class="language-r">#'   and m is the number `nsim` of simulated responses.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">150</td>
+                    <td class="num">598</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/between-within.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">599</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Determine Within or Between for each Design Matrix Column</pre>
+                      <pre class="language-r">#' @param seed unused argument from [stats::simulate()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">600</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param method (`string`)\cr simulation method to use. If "conditional",</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">601</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [h_df_bw_calc()] to determine whether a variable</pre>
+                      <pre class="language-r">#'   simulated values are sampled given the estimated covariance matrix of `object`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">602</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   differs only between subjects or also within subjects.</pre>
+                      <pre class="language-r">#'   If "marginal", the variance of the estimated covariance matrix is taken into account.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">603</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">604</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x_matrix (`matrix`)\cr the design matrix with column names.</pre>
+                      <pre class="language-r">#' @importFrom stats simulate</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">605</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param subject_ids (`factor`)\cr the subject IDs.</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">606</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">simulate.mmrm_tmb &lt;- function(object,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">607</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Character vector with "intercept", "within" or "between" for each</pre>
+                      <pre class="language-r">                              nsim = 1,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">608</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   design matrix column identified via the names of the vector.</pre>
+                      <pre class="language-r">                              seed = NULL,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">609</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">                              newdata,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">610</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">                              ...,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">13</td>
+                    <td class="num">611</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_within_or_between &lt;- function(x_matrix, subject_ids) {</pre>
+                      <pre class="language-r">                              method = c("conditional", "marginal")) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">14</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">612</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(x_matrix, col.names = "unique", min.cols = 1L)</pre>
+                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">15</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">613</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_factor(subject_ids, len = nrow(x_matrix))</pre>
+                      <pre class="language-r">  assert_null(seed)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">614</td>
+                    <td class="coverage">15<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  if (missing(newdata)) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">615</td>
+                    <td class="coverage">12<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    newdata &lt;- object$data</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">616</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">17</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">617</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_subjects &lt;- length(unique(subject_ids))</pre>
+                      <pre class="language-r">  assert_data_frame(newdata)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">18</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">618</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  vapply(</pre>
+                      <pre class="language-r">  method &lt;- match.arg(method)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">19</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">619</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    colnames(x_matrix),</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">20</td>
-                    <td class="coverage">19<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">620</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    function(x) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">21</td>
-                    <td class="coverage">112<em>x</em></td>
+                    <td class="num">621</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      if (x == "(Intercept)") {</pre>
+                      <pre class="language-r">  tmb_data &lt;- h_mmrm_tmb_data(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">22</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">622</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        "intercept"</pre>
+                      <pre class="language-r">    object$formula_parts, newdata,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">23</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">623</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      } else {</pre>
+                      <pre class="language-r">    weights = rep(1, nrow(newdata)),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">24</td>
-                    <td class="coverage">93<em>x</em></td>
+                    <td class="num">624</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        n_unique &lt;- nrow(unique(cbind(x_matrix[, x], subject_ids)))</pre>
+                      <pre class="language-r">    reml = TRUE,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">25</td>
-                    <td class="coverage">43<em>x</em></td>
+                    <td class="num">625</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        if (n_unique &gt; n_subjects) "within" else "between"</pre>
+                      <pre class="language-r">    singular = "keep",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">26</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">626</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      }</pre>
+                      <pre class="language-r">    drop_visit_levels = FALSE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">27</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">627</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    },</pre>
+                      <pre class="language-r">    allow_na_response = TRUE,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">28</td>
-                    <td class="coverage">19<em>x</em></td>
+                    <td class="num">628</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    character(1L)</pre>
+                      <pre class="language-r">    drop_levels = FALSE,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">29</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">629</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    xlev = component(object, "xlev"),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">30</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">630</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    contrasts = component(object, "contrasts")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">31</td>
+                    <td class="num">631</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">32</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">632</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Between-Within Degrees of Freedom</pre>
+                      <pre class="language-r">  ret &lt;- if (method == "conditional") {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">33</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">633</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    predict_res &lt;- h_get_prediction(tmb_data, object$theta_est, object$beta_est, object$beta_vcov)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">34</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">634</td>
+                    <td class="coverage">8<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [h_df_1d_bw()] and [h_df_md_bw()].</pre>
+                      <pre class="language-r">    as.data.frame(h_get_sim_per_subj(predict_res, tmb_data$n_subjects, nsim))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">35</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">635</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  } else if (method == "marginal") {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">36</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">636</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr the fitted MMRM.</pre>
+                      <pre class="language-r">    theta_chol &lt;- t(chol(object$theta_vcov))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">37</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">637</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    n_theta &lt;- length(object$theta_est)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">38</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">638</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return List with:</pre>
+                      <pre class="language-r">    as.data.frame(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">39</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">639</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `coefs_between_within` calculated via [h_within_or_between()]</pre>
+                      <pre class="language-r">      sapply(seq_len(nsim), function(x) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">40</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">640</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `ddf_between`</pre>
+                      <pre class="language-r">        newtheta &lt;- object$theta_est + theta_chol %*% matrix(stats::rnorm(n_theta), ncol = 1)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">41</td>
+                    <td class="num">641</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   - `ddf_within`</pre>
+                      <pre class="language-r">        # Recalculate betas with sampled thetas.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">42</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">642</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        hold &lt;- object$tmb_object$report(newtheta)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">43</td>
+                    <td class="num">643</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">        # Resample betas given new beta distribution.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">44</td>
+                    <td class="num">644</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_bw_calc &lt;- function(object) {</pre>
+                      <pre class="language-r">        # We first solve L^\top w = D^{-1/2}z_{sample}:</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">45</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">645</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">        w_sample &lt;- backsolve(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">46</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">646</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">          r = hold$XtWX_L,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">47</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">647</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_subjects &lt;- component(object, "n_subjects")</pre>
+                      <pre class="language-r">          x = stats::rnorm(length(hold$beta)) / sqrt(hold$XtWX_D),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">48</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">648</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_obs &lt;- component(object, "n_obs")</pre>
+                      <pre class="language-r">          upper.tri = FALSE,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">49</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">649</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  x_mat &lt;- component(object, "x_matrix")</pre>
+                      <pre class="language-r">          transpose = TRUE</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">50</td>
+                    <td class="num">650</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">        )</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">651</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">        # Then we add the mean vector, the beta estimate.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">51</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">652</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  subject_var &lt;- component(object, "subject_var")</pre>
+                      <pre class="language-r">        beta_sample &lt;- hold$beta + w_sample</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">52</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">653</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  full_frame &lt;- component(object, "full_frame")</pre>
+                      <pre class="language-r">        predict_res &lt;- h_get_prediction(tmb_data, newtheta, beta_sample, hold$beta_vcov)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">53</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">654</td>
+                    <td class="coverage">503<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  subject_ids &lt;- full_frame[[subject_var]]</pre>
+                      <pre class="language-r">        h_get_sim_per_subj(predict_res, tmb_data$n_subjects, 1L)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">54</td>
+                    <td class="num">655</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      })</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">55</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">656</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coefs_between_within &lt;- h_within_or_between(x_mat, subject_ids)</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">56</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">657</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_coefs_between &lt;- sum(coefs_between_within == "between")</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">57</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">658</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_intercept &lt;- sum(coefs_between_within == "intercept")</pre>
+                      <pre class="language-r">  orig_row_names &lt;- row.names(newdata)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">58</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">659</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  n_coefs_within &lt;- sum(coefs_between_within == "within")</pre>
+                      <pre class="language-r">  new_order &lt;- match(orig_row_names, row.names(tmb_data$full_frame))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">59</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">660</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ddf_between &lt;- n_subjects - n_coefs_between - n_intercept</pre>
+                      <pre class="language-r">  ret[new_order, , drop = FALSE]</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">60</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">661</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ddf_within &lt;- n_obs - n_subjects - n_coefs_within</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">61</td>
+                    <td class="num">662</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">62</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">663</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  list(</pre>
+                      <pre class="language-r">#' Get simulated values by patient.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">63</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">664</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    coefs_between_within = coefs_between_within,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">64</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">665</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ddf_between = ddf_between,</pre>
+                      <pre class="language-r">#' @param predict_res (`list`)\cr from [h_get_prediction()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">65</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">666</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ddf_within = ddf_within</pre>
+                      <pre class="language-r">#' @param nsub (`count`)\cr number of subjects.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">66</td>
+                    <td class="num">667</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' @param nsim (`count`)\cr number of values to simulate.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">67</td>
+                    <td class="num">668</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">68</td>
+                    <td class="num">669</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">69</td>
+                    <td class="num">670</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Assign Minimum Degrees of Freedom Given Involved Coefficients</pre>
+                      <pre class="language-r">h_get_sim_per_subj &lt;- function(predict_res, nsub, nsim) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">70</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">671</td>
+                    <td class="coverage">517<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_list(predict_res)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">71</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">672</td>
+                    <td class="coverage">517<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [h_df_1d_bw()] and [h_df_md_bw()].</pre>
+                      <pre class="language-r">  assert_count(nsub, positive = TRUE)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">673</td>
+                    <td class="coverage">516<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_count(nsim, positive = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">72</td>
+                    <td class="num">674</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">675</td>
+                    <td class="coverage">515<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  ret &lt;- matrix(</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">676</td>
+                    <td class="coverage">515<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    predict_res$prediction[, "fit"],</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">677</td>
+                    <td class="coverage">515<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    ncol = nsim,</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">678</td>
+                    <td class="coverage">515<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    nrow = nrow(predict_res$prediction)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">73</td>
+                    <td class="num">679</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param bw_calc (`list`)\cr from [h_df_bw_calc()].</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">74</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">680</td>
+                    <td class="coverage">515<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param is_coef_involved (`logical`)\cr whether each coefficient is involved</pre>
+                      <pre class="language-r">  for (i in seq_len(nsub)) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">75</td>
+                    <td class="num">681</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   in the contrast.</pre>
+                      <pre class="language-r">    # Skip subjects which are not included in predict_res.</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">76</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">682</td>
+                    <td class="coverage">82699<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    if (length(predict_res$index[[i]]) &gt; 0) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">77</td>
+                    <td class="num">683</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The minimum of the degrees of freedom assigned to each involved</pre>
+                      <pre class="language-r">      # Obtain indices of data.frame belonging to subject i</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">78</td>
+                    <td class="num">684</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   coefficient according to its between-within categorization.</pre>
+                      <pre class="language-r">      # (increment by 1, since indices from cpp are 0-order).</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">79</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">685</td>
+                    <td class="coverage">66631<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      inds &lt;- predict_res$index[[i]] + 1</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">80</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">686</td>
+                    <td class="coverage">66631<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">      obs &lt;- length(inds)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">81</td>
+                    <td class="num">687</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_min_bw &lt;- function(bw_calc, is_coef_involved) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">82</td>
-                    <td class="coverage">17<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">688</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_list(bw_calc)</pre>
+                      <pre class="language-r">      # Get relevant covariance matrix for subject i.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">83</td>
-                    <td class="coverage">17<em>x</em></td>
+                    <td class="num">689</td>
+                    <td class="coverage">66631<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_names(names(bw_calc), identical.to = c("coefs_between_within", "ddf_between", "ddf_within"))</pre>
+                      <pre class="language-r">      covmat_i &lt;- predict_res$covariance[[i]]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">84</td>
-                    <td class="coverage">17<em>x</em></td>
+                    <td class="num">690</td>
+                    <td class="coverage">66631<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_logical(is_coef_involved, len = length(bw_calc$coefs_between_within))</pre>
+                      <pre class="language-r">      theta_chol &lt;- t(chol(covmat_i))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">85</td>
-                    <td class="coverage">17<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">691</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_true(sum(is_coef_involved) &gt; 0)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">86</td>
+                    <td class="num">692</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      # Simulate epsilon from covariance matrix.</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">87</td>
-                    <td class="coverage">17<em>x</em></td>
+                    <td class="num">693</td>
+                    <td class="coverage">66631<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_categories &lt;- bw_calc$coefs_between_within[is_coef_involved]</pre>
+                      <pre class="language-r">      mus &lt;- ret[inds, , drop = FALSE]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">88</td>
-                    <td class="coverage">17<em>x</em></td>
+                    <td class="num">694</td>
+                    <td class="coverage">66631<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_dfs &lt;- vapply(</pre>
+                      <pre class="language-r">      epsilons &lt;- theta_chol %*% matrix(stats::rnorm(nsim * obs), ncol = nsim)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">89</td>
-                    <td class="coverage">17<em>x</em></td>
+                    <td class="num">695</td>
+                    <td class="coverage">66631<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    X = coef_categories,</pre>
+                      <pre class="language-r">      ret[inds, ] &lt;- mus + epsilons</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">17<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">696</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    FUN = switch,</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">17<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">697</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    intercept = bw_calc$ddf_within,</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">17<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">698</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    between = bw_calc$ddf_between,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">17<em>x</em></td>
+                    <td class="num">699</td>
+                    <td class="coverage">515<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    within = bw_calc$ddf_within,</pre>
+                      <pre class="language-r">  ret</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">94</td>
-                    <td class="coverage">17<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">700</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    FUN.VALUE = integer(1)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/between-within.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">95</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' Determine Within or Between for each Design Matrix Column</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">96</td>
-                    <td class="coverage">17<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">2</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  min(coef_dfs)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">97</td>
+                    <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @description Used in [h_df_bw_calc()] to determine whether a variable</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">98</td>
+                    <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   differs only between subjects or also within subjects.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">99</td>
+                    <td class="num">5</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Between-Within Degrees of Freedom for One-Dimensional Contrast</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">100</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @param x_matrix (`matrix`)\cr the design matrix with column names.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">101</td>
+                    <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_1d()] if method is "Between-Within".</pre>
+                      <pre class="language-r">#' @param subject_ids (`factor`)\cr the subject IDs.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">102</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">103</td>
+                    <td class="num">9</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_df_1d_sat</pre>
+                      <pre class="language-r">#' @return Character vector with "intercept", "within" or "between" for each</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">104</td>
+                    <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inherit h_df_1d_sat return</pre>
+                      <pre class="language-r">#'   design matrix column identified via the names of the vector.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">105</td>
+                    <td class="num">11</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#'</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">12</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">106</td>
+                    <td class="num">13</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_1d_bw &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">h_within_or_between &lt;- function(x_matrix, subject_ids) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">107</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">14</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">  assert_matrix(x_matrix, col.names = "unique", min.cols = 1L)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">108</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">15</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">  assert_factor(subject_ids, len = nrow(x_matrix))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">109</td>
+                    <td class="num">16</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">110</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">17</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  bw_calc &lt;- h_df_bw_calc(object)</pre>
+                      <pre class="language-r">  n_subjects &lt;- length(unique(subject_ids))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">111</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">18</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  is_coef_involved &lt;- contrast != 0</pre>
+                      <pre class="language-r">  vapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">112</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">19</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- h_df_min_bw(bw_calc, is_coef_involved)</pre>
+                      <pre class="language-r">    colnames(x_matrix),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">113</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">20</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_test_1d(object, contrast, df)</pre>
+                      <pre class="language-r">    function(x) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">114</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">21</td>
+                    <td class="coverage">112<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      if (x == "(Intercept)") {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">115</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">22</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">        "intercept"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">116</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Between-Within Degrees of Freedom for Multi-Dimensional Contrast</pre>
+                      <pre class="language-r">      } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">117</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">24</td>
+                    <td class="coverage">93<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">        n_unique &lt;- nrow(unique(cbind(x_matrix[, x], subject_ids)))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">118</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">25</td>
+                    <td class="coverage">43<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_md()] if method is "Between-Within".</pre>
+                      <pre class="language-r">        if (n_unique &gt; n_subjects) "within" else "between"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">119</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">120</td>
+                    <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_df_md_sat</pre>
+                      <pre class="language-r">    },</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">121</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">28</td>
+                    <td class="coverage">19<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inherit h_df_md_sat return</pre>
+                      <pre class="language-r">    character(1L)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">122</td>
+                    <td class="num">29</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">123</td>
+                    <td class="num">30</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_md_bw &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">124</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">31</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">125</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">32</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">#' Calculation of Between-Within Degrees of Freedom</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">126</td>
+                    <td class="num">33</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">127</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">34</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  bw_calc &lt;- h_df_bw_calc(object)</pre>
+                      <pre class="language-r">#' @description Used in [h_df_1d_bw()] and [h_df_md_bw()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">128</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">35</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  is_coef_involved &lt;- apply(X = contrast != 0, MARGIN = 2L, FUN = any)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">129</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">36</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- h_df_min_bw(bw_calc, is_coef_involved)</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr the fitted MMRM.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">130</td>
-                    <td class="coverage">7<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">37</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_test_md(object, contrast, df)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">131</td>
+                    <td class="num">38</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @return List with:</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/mmrm-methods.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">39</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Methods for `mmrm` Objects</pre>
+                      <pre class="language-r">#'   - `coefs_between_within` calculated via [h_within_or_between()]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">40</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   - `ddf_between`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">41</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">#'   - `ddf_within`</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">42</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">43</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr the fitted MMRM including Jacobian and call etc.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">44</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... not used.</pre>
+                      <pre class="language-r">h_df_bw_calc &lt;- function(object) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">7</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">45</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Depends on the method, see Details and Functions.</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">46</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">9</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">47</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details</pre>
+                      <pre class="language-r">  n_subjects &lt;- component(object, "n_subjects")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">10</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">48</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' While printing the summary of (`mmrm`)\cr object, the following will be displayed:</pre>
+                      <pre class="language-r">  n_obs &lt;- component(object, "n_obs")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">11</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">49</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 1. Formula. The formula used in the model.</pre>
+                      <pre class="language-r">  x_mat &lt;- component(object, "x_matrix")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 2. Data. The data used for analysis, including number of subjects, number of valid observations.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">13</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">51</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 3. Covariance. The covariance structure and number of variance parameters.</pre>
+                      <pre class="language-r">  subject_var &lt;- component(object, "subject_var")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">14</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">52</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 4. Method. Restricted maximum likelihood(REML) or maximum likelihood(ML).</pre>
+                      <pre class="language-r">  full_frame &lt;- component(object, "full_frame")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">15</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">53</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 5. Model selection criteria. AIC, BIC, log likelihood and deviance.</pre>
+                      <pre class="language-r">  subject_ids &lt;- full_frame[[subject_var]]</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">54</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 6. Coefficients. Coefficients of the covariates.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">17</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">55</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' 7. Covariance estimate. The covariance estimate(for each group).</pre>
+                      <pre class="language-r">  coefs_between_within &lt;- h_within_or_between(x_mat, subject_ids)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">18</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">56</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'    1. If the covariance structure is non-spatial, the covariance matrix of all categorical time points available</pre>
+                      <pre class="language-r">  n_coefs_between &lt;- sum(coefs_between_within == "between")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">19</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">57</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'       in data will be displayed.</pre>
+                      <pre class="language-r">  n_intercept &lt;- sum(coefs_between_within == "intercept")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">20</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">58</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'    2. If the covariance structure is spatial, the covariance matrix of two time points with unit distance</pre>
+                      <pre class="language-r">  n_coefs_within &lt;- sum(coefs_between_within == "within")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">21</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">59</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'       will be displayed.</pre>
+                      <pre class="language-r">  ddf_between &lt;- n_subjects - n_coefs_between - n_intercept</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">22</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">60</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  ddf_within &lt;- n_obs - n_subjects - n_coefs_within</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">23</td>
+                    <td class="num">61</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' `confint` is used to obtain the confidence intervals for the coefficients.</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">62</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  list(</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">63</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    coefs_between_within = coefs_between_within,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">24</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">64</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Please note that this is different from the confidence interval of difference</pre>
+                      <pre class="language-r">    ddf_between = ddf_between,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">25</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">65</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' of least square means from `emmeans`.</pre>
+                      <pre class="language-r">    ddf_within = ddf_within</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">26</td>
+                    <td class="num">66</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">27</td>
+                    <td class="num">67</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @name mmrm_methods</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">28</td>
+                    <td class="num">68</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">29</td>
+                    <td class="num">69</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @seealso [`mmrm_tmb_methods`], [`mmrm_tidiers`] for additional methods.</pre>
+                      <pre class="language-r">#' Assign Minimum Degrees of Freedom Given Involved Coefficients</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">30</td>
+                    <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">31</td>
+                    <td class="num">71</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @description Used in [h_df_1d_bw()] and [h_df_md_bw()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">32</td>
+                    <td class="num">72</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' formula &lt;- FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">33</td>
+                    <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' object &lt;- mmrm(formula, fev_data)</pre>
+                      <pre class="language-r">#' @param bw_calc (`list`)\cr from [h_df_bw_calc()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">34</td>
+                    <td class="num">74</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">NULL</pre>
+                      <pre class="language-r">#' @param is_coef_involved (`logical`)\cr whether each coefficient is involved</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">35</td>
+                    <td class="num">75</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'   in the contrast.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">36</td>
+                    <td class="num">76</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Coefficients Table for MMRM Fit</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">37</td>
+                    <td class="num">77</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @return The minimum of the degrees of freedom assigned to each involved</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">38</td>
+                    <td class="num">78</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used by [summary.mmrm()] to obtain the coefficients table.</pre>
+                      <pre class="language-r">#'   coefficient according to its between-within categorization.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">39</td>
+                    <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">40</td>
+                    <td class="num">80</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param object (`mmrm`)\cr model fit.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">41</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">h_df_min_bw &lt;- function(bw_calc, is_coef_involved) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">42</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">82</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return Matrix with one row per coefficient and columns</pre>
+                      <pre class="language-r">  assert_list(bw_calc)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">43</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">83</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `Estimate`, `Std. Error`, `df`, `t value` and `Pr(&gt;|t|)`.</pre>
+                      <pre class="language-r">  assert_names(names(bw_calc), identical.to = c("coefs_between_within", "ddf_between", "ddf_within"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">44</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">84</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  assert_logical(is_coef_involved, len = length(bw_calc$coefs_between_within))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">45</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">85</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  assert_true(sum(is_coef_involved) &gt; 0)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">46</td>
+                    <td class="num">86</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_coef_table &lt;- function(object) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">47</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">87</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">  coef_categories &lt;- bw_calc$coefs_between_within[is_coef_involved]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">48</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">88</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  coef_dfs &lt;- vapply(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">49</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">89</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_est &lt;- component(object, "beta_est")</pre>
+                      <pre class="language-r">    X = coef_categories,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">50</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">90</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_contrasts &lt;- diag(x = rep(1, length(coef_est)))</pre>
+                      <pre class="language-r">    FUN = switch,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">51</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">91</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  rownames(coef_contrasts) &lt;- names(coef_est)</pre>
+                      <pre class="language-r">    intercept = bw_calc$ddf_within,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">52</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">92</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_table &lt;- t(apply(</pre>
+                      <pre class="language-r">    between = bw_calc$ddf_between,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">53</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">93</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    coef_contrasts,</pre>
+                      <pre class="language-r">    within = bw_calc$ddf_within,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">54</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">94</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    MARGIN = 1L,</pre>
+                      <pre class="language-r">    FUN.VALUE = integer(1)</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">95</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">55</td>
-                    <td class="coverage">40<em>x</em></td>
+                    <td class="num">96</td>
+                    <td class="coverage">17<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    FUN = function(contrast) unlist(df_1d(object, contrast))</pre>
+                      <pre class="language-r">  min(coef_dfs)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">56</td>
+                    <td class="num">97</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ))</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">57</td>
-                    <td class="coverage">40<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">98</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_names(</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">58</td>
-                    <td class="coverage">40<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">99</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    colnames(coef_table),</pre>
+                      <pre class="language-r">#' Calculation of Between-Within Degrees of Freedom for One-Dimensional Contrast</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">59</td>
-                    <td class="coverage">40<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">100</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    identical.to = c("est", "se", "df", "t_stat", "p_val")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">60</td>
+                    <td class="num">101</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' @description Used in [df_1d()] if method is "Between-Within".</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">61</td>
-                    <td class="coverage">40<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">102</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  colnames(coef_table) &lt;- c("Estimate", "Std. Error", "df", "t value", "Pr(&gt;|t|)")</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">62</td>
+                    <td class="num">103</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @inheritParams h_df_1d_sat</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">63</td>
-                    <td class="coverage">40<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">104</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_aliased &lt;- component(object, "beta_aliased")</pre>
+                      <pre class="language-r">#' @inherit h_df_1d_sat return</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">64</td>
-                    <td class="coverage">40<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">105</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (any(coef_aliased)) {</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">65</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">106</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    names_coef_na &lt;- names(which(coef_aliased))</pre>
+                      <pre class="language-r">h_df_1d_bw &lt;- function(object, contrast) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">66</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">107</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    coef_na_table &lt;- matrix(</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">67</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">108</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      data = NA,</pre>
+                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">68</td>
-                    <td class="coverage">2<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">109</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      nrow = length(names_coef_na),</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">69</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">110</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      ncol = ncol(coef_table),</pre>
+                      <pre class="language-r">  bw_calc &lt;- h_df_bw_calc(object)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">70</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">111</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      dimnames = list(names_coef_na, colnames(coef_table))</pre>
+                      <pre class="language-r">  is_coef_involved &lt;- contrast != 0</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">71</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">112</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">  df &lt;- h_df_min_bw(bw_calc, is_coef_involved)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">72</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">113</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    coef_table &lt;- rbind(coef_table, coef_na_table)[names(coef_aliased), ]</pre>
+                      <pre class="language-r">  h_test_1d(object, contrast, df)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">73</td>
+                    <td class="num">114</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">74</td>
+                    <td class="num">115</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">75</td>
-                    <td class="coverage">40<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">116</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_table</pre>
+                      <pre class="language-r">#' Calculation of Between-Within Degrees of Freedom for Multi-Dimensional Contrast</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">76</td>
+                    <td class="num">117</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">77</td>
+                    <td class="num">118</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @description Used in [df_md()] if method is "Between-Within".</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">78</td>
+                    <td class="num">119</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_methods summarizes the MMRM fit results.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">79</td>
+                    <td class="num">120</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">#' @inheritParams h_df_md_sat</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">80</td>
+                    <td class="num">121</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @inherit h_df_md_sat return</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">81</td>
+                    <td class="num">122</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Summary:</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">82</td>
+                    <td class="num">123</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' summary(object)</pre>
+                      <pre class="language-r">h_df_md_bw &lt;- function(object, contrast) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">124</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">125</td>
+                    <td class="coverage">7<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">83</td>
+                    <td class="num">126</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">summary.mmrm &lt;- function(object, ...) {</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">84</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">127</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  aic_list &lt;- list(</pre>
+                      <pre class="language-r">  bw_calc &lt;- h_df_bw_calc(object)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">85</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">128</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    AIC = AIC(object),</pre>
+                      <pre class="language-r">  is_coef_involved &lt;- apply(X = contrast != 0, MARGIN = 2L, FUN = any)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">86</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">129</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    BIC = BIC(object),</pre>
+                      <pre class="language-r">  df &lt;- h_df_min_bw(bw_calc, is_coef_involved)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">87</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">130</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    logLik = logLik(object),</pre>
+                      <pre class="language-r">  h_test_md(object, contrast, df)</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">88</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">131</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    deviance = deviance(object)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/mmrm-methods.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
                   <tr class="never">
-                    <td class="num">89</td>
+                    <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' Methods for `mmrm` Objects</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">2</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coefficients &lt;- h_coef_table(object)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">3</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  call &lt;- stats::getCall(object)</pre>
+                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">4</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  components &lt;- component(object, c(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">5</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "cov_type", "reml", "n_groups", "n_theta",</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr the fitted MMRM including Jacobian and call etc.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">94</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">6</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "n_subjects", "n_timepoints", "n_obs",</pre>
+                      <pre class="language-r">#' @param ... not used.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">95</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">7</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "beta_vcov", "varcor"</pre>
+                      <pre class="language-r">#' @return Depends on the method, see Details and Functions.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">96</td>
+                    <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">97</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">9</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  components$method &lt;- object$method</pre>
+                      <pre class="language-r">#' @details</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">98</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">10</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  components$vcov &lt;- object$vcov</pre>
+                      <pre class="language-r">#' While printing the summary of (`mmrm`)\cr object, the following will be displayed:</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">99</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">11</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  structure(</pre>
+                      <pre class="language-r">#' 1. Formula. The formula used in the model.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">100</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">12</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    c(</pre>
+                      <pre class="language-r">#' 2. Data. The data used for analysis, including number of subjects, number of valid observations.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">101</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">13</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' 3. Covariance. The covariance structure and number of variance parameters.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">14</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' 4. Method. Restricted maximum likelihood(REML) or maximum likelihood(ML).</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">15</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' 5. Model selection criteria. AIC, BIC, log likelihood and deviance.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">16</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' 6. Coefficients. Coefficients of the covariates.</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">17</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      components,</pre>
+                      <pre class="language-r">#' 7. Covariance estimate. The covariance estimate(for each group).</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">102</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">18</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      list(</pre>
+                      <pre class="language-r">#'    1. If the covariance structure is non-spatial, the covariance matrix of all categorical time points available</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">103</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">19</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        coefficients = coefficients,</pre>
+                      <pre class="language-r">#'       in data will be displayed.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">104</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">20</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        n_singular_coefs = sum(component(object, "beta_aliased")),</pre>
+                      <pre class="language-r">#'    2. If the covariance structure is spatial, the covariance matrix of two time points with unit distance</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">105</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">21</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        aic_list = aic_list,</pre>
+                      <pre class="language-r">#'       will be displayed.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">106</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">22</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">        call = call</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">107</td>
+                    <td class="num">23</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      )</pre>
+                      <pre class="language-r">#' `confint` is used to obtain the confidence intervals for the coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">108</td>
+                    <td class="num">24</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ),</pre>
+                      <pre class="language-r">#' Please note that this is different from the confidence interval of difference</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">109</td>
-                    <td class="coverage">20<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">25</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    class = "summary.mmrm"</pre>
+                      <pre class="language-r">#' of least square means from `emmeans`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">110</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">111</td>
+                    <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' @name mmrm_methods</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">112</td>
+                    <td class="num">28</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">113</td>
+                    <td class="num">29</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Printing MMRM Function Call</pre>
+                      <pre class="language-r">#' @seealso [`mmrm_tmb_methods`], [`mmrm_tidiers`] for additional methods.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">114</td>
+                    <td class="num">30</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">115</td>
+                    <td class="num">31</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used in [print.summary.mmrm()].</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">116</td>
+                    <td class="num">32</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' formula &lt;- FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">117</td>
+                    <td class="num">33</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param call (`call`)\cr original [mmrm()] function call.</pre>
+                      <pre class="language-r">#' object &lt;- mmrm(formula, fev_data)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">118</td>
+                    <td class="num">34</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_obs (`int`)\cr number of observations.</pre>
+                      <pre class="language-r">NULL</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">119</td>
+                    <td class="num">35</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_subjects (`int`)\cr number of subjects.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">120</td>
+                    <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_timepoints (`int`)\cr number of timepoints.</pre>
+                      <pre class="language-r">#' Coefficients Table for MMRM Fit</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">121</td>
+                    <td class="num">37</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">122</td>
+                    <td class="num">38</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">#' This is used by [summary.mmrm()] to obtain the coefficients table.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">123</td>
+                    <td class="num">39</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_print_call &lt;- function(call, n_obs, n_subjects, n_timepoints) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">124</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">40</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  pass &lt;- 0</pre>
+                      <pre class="language-r">#' @param object (`mmrm`)\cr model fit.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">125</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">41</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(tmp &lt;- call$formula)) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">126</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">42</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cat("Formula:    ", deparse(tmp), fill = TRUE)</pre>
+                      <pre class="language-r">#' @return Matrix with one row per coefficient and columns</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">127</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">43</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    rhs &lt;- tmp[[2]]</pre>
+                      <pre class="language-r">#'   `Estimate`, `Std. Error`, `df`, `t value` and `Pr(&gt;|t|)`.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">128</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">44</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    pass &lt;- nchar(deparse(rhs))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">129</td>
+                    <td class="num">45</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">130</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">46</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(call$data)) {</pre>
+                      <pre class="language-r">h_coef_table &lt;- function(object) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">131</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">47</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cat(</pre>
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">132</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">48</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "Data:       ", deparse(call$data), "(used", n_obs, "observations from",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">133</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">49</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      n_subjects, "subjects with maximum", n_timepoints, "timepoints)",</pre>
+                      <pre class="language-r">  coef_est &lt;- component(object, "beta_est")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">134</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">50</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      fill = TRUE</pre>
+                      <pre class="language-r">  coef_contrasts &lt;- diag(x = rep(1, length(coef_est)))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">135</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">51</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">  rownames(coef_contrasts) &lt;- names(coef_est)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">136</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">52</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  coef_table &lt;- t(apply(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">137</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">53</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  # Display the expression of weights</pre>
+                      <pre class="language-r">    coef_contrasts,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">138</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">54</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(call$weights)) {</pre>
+                      <pre class="language-r">    MARGIN = 1L,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">139</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">55</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cat("Weights:    ", deparse(call$weights), fill = TRUE)</pre>
+                      <pre class="language-r">    FUN = function(contrast) unlist(df_1d(object, contrast))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">140</td>
+                    <td class="num">56</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  ))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">141</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">57</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  assert_names(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">142</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">58</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    colnames(coef_table),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">143</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">59</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Printing MMRM Covariance Type</pre>
+                      <pre class="language-r">    identical.to = c("est", "se", "df", "t_stat", "p_val")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">144</td>
+                    <td class="num">60</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">145</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">61</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used in [print.summary.mmrm()].</pre>
+                      <pre class="language-r">  colnames(coef_table) &lt;- c("Estimate", "Std. Error", "df", "t value", "Pr(&gt;|t|)")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">146</td>
+                    <td class="num">62</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">147</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">63</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param cov_type (`string`)\cr covariance structure abbreviation.</pre>
+                      <pre class="language-r">  coef_aliased &lt;- component(object, "beta_aliased")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">148</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">64</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_theta (`count`)\cr number of variance parameters.</pre>
+                      <pre class="language-r">  if (any(coef_aliased)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">149</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">65</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param n_groups (`count`)\cr number of groups.</pre>
+                      <pre class="language-r">    names_coef_na &lt;- names(which(coef_aliased))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">150</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">66</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    coef_na_table &lt;- matrix(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">151</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">67</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_print_cov &lt;- function(cov_type, n_theta, n_groups) {</pre>
+                      <pre class="language-r">      data = NA,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">152</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">68</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(cov_type)</pre>
+                      <pre class="language-r">      nrow = length(names_coef_na),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">153</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">69</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(n_theta, positive = TRUE)</pre>
+                      <pre class="language-r">      ncol = ncol(coef_table),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">154</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">70</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_count(n_groups, positive = TRUE)</pre>
+                      <pre class="language-r">      dimnames = list(names_coef_na, colnames(coef_table))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">155</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">71</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cov_definition &lt;- switch(cov_type,</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">156</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">72</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    us = "unstructured",</pre>
+                      <pre class="language-r">    coef_table &lt;- rbind(coef_table, coef_na_table)[names(coef_aliased), ]</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">157</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">73</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    toep = "Toeplitz",</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">158</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">74</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    toeph = "heterogeneous Toeplitz",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">159</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">75</td>
+                    <td class="coverage">40<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ar1 = "auto-regressive order one",</pre>
+                      <pre class="language-r">  coef_table</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">160</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">76</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ar1h = "heterogeneous auto-regressive order one",</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">161</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">77</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ad = "ante-dependence",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">162</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">78</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    adh = "heterogeneous ante-dependence",</pre>
+                      <pre class="language-r">#' @describeIn mmrm_methods summarizes the MMRM fit results.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">163</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">79</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cs = "compound symmetry",</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">164</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">80</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    csh = "heterogeneous compound symmetry",</pre>
+                      <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">165</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">81</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    sp_exp = "spatial exponential"</pre>
+                      <pre class="language-r">#' # Summary:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">166</td>
+                    <td class="num">82</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' summary(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">167</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">summary.mmrm &lt;- function(object, ...) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">168</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">84</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  catstr &lt;- sprintf(</pre>
+                      <pre class="language-r">  aic_list &lt;- list(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">169</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">85</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    "Covariance:  %s (%d variance parameters%s)\n",</pre>
+                      <pre class="language-r">    AIC = AIC(object),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">170</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">86</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cov_definition,</pre>
+                      <pre class="language-r">    BIC = BIC(object),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">171</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">87</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    n_theta,</pre>
+                      <pre class="language-r">    logLik = logLik(object),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">172</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">88</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ifelse(n_groups == 1L, "", sprintf(" of %d groups", n_groups))</pre>
+                      <pre class="language-r">    deviance = deviance(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">173</td>
+                    <td class="num">89</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">174</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">90</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat(catstr)</pre>
+                      <pre class="language-r">  coefficients &lt;- h_coef_table(object)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">175</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">91</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  call &lt;- stats::getCall(object)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">176</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">92</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  components &lt;- component(object, c(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">177</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">93</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Printing AIC and other Model Fit Criteria</pre>
+                      <pre class="language-r">    "cov_type", "reml", "n_groups", "n_theta",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">178</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">94</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    "n_subjects", "n_timepoints", "n_obs",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">179</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">95</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used in [print.summary.mmrm()].</pre>
+                      <pre class="language-r">    "beta_vcov", "varcor"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">180</td>
+                    <td class="num">96</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  ))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">181</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">97</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param aic_list (`list`)\cr list as part of from [summary.mmrm()].</pre>
+                      <pre class="language-r">  components$method &lt;- object$method</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">182</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">98</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param digits (`number`)\cr number of decimal places used with [round()].</pre>
+                      <pre class="language-r">  components$vcov &lt;- object$vcov</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">183</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">99</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  structure(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">184</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">100</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    c(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">185</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">101</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_print_aic_list &lt;- function(aic_list,</pre>
+                      <pre class="language-r">      components,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">186</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">102</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                             digits = 1) {</pre>
+                      <pre class="language-r">      list(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">187</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">103</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  diag_vals &lt;- round(unlist(aic_list), digits)</pre>
+                      <pre class="language-r">        coefficients = coefficients,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">188</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">104</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  diag_vals &lt;- format(diag_vals)</pre>
+                      <pre class="language-r">        n_singular_coefs = sum(component(object, "beta_aliased")),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">189</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">105</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  print(diag_vals, quote = FALSE)</pre>
+                      <pre class="language-r">        aic_list = aic_list,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">190</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">106</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">        call = call</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">191</td>
+                    <td class="num">107</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">192</td>
+                    <td class="num">108</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_methods prints the MMRM fit summary.</pre>
+                      <pre class="language-r">    ),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">193</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">109</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">    class = "summary.mmrm"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">194</td>
+                    <td class="num">110</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">195</td>
+                    <td class="num">111</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">print.summary.mmrm &lt;- function(x,</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">196</td>
+                    <td class="num">112</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                               digits = max(3, getOption("digits") - 3),</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">197</td>
+                    <td class="num">113</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                               signif.stars = getOption("show.signif.stars"), # nolint</pre>
+                      <pre class="language-r">#' Printing MMRM Function Call</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">198</td>
+                    <td class="num">114</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                               ...) {</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">199</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">115</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("mmrm fit\n\n")</pre>
+                      <pre class="language-r">#' This is used in [print.summary.mmrm()].</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">200</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">116</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_print_call(x$call, x$n_obs, x$n_subjects, x$n_timepoints)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">201</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">117</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_print_cov(x$cov_type, x$n_theta, x$n_groups)</pre>
+                      <pre class="language-r">#' @param call (`call`)\cr original [mmrm()] function call.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">202</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">118</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("Method:      ", x$method, "\n", sep = "")</pre>
+                      <pre class="language-r">#' @param n_obs (`int`)\cr number of observations.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">203</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">119</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("Vcov Method: ", x$vcov, "\n", sep = "")</pre>
+                      <pre class="language-r">#' @param n_subjects (`int`)\cr number of subjects.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">204</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">120</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("Inference:   ")</pre>
+                      <pre class="language-r">#' @param n_timepoints (`int`)\cr number of timepoints.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">205</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">121</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat(ifelse(x$reml, "REML", "ML"))</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">206</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">122</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n\n")</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">207</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">123</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("Model selection criteria:\n")</pre>
+                      <pre class="language-r">h_print_call &lt;- function(call, n_obs, n_subjects, n_timepoints) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">208</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">124</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_print_aic_list(x$aic_list)</pre>
+                      <pre class="language-r">  pass &lt;- 0</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">209</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">125</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n")</pre>
+                      <pre class="language-r">  if (!is.null(tmp &lt;- call$formula)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">210</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">126</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cat("Coefficients: ")</pre>
+                      <pre class="language-r">    cat("Formula:    ", deparse(tmp), fill = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">211</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">127</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (x$n_singular_coefs &gt; 0) {</pre>
+                      <pre class="language-r">    rhs &lt;- tmp[[2]]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">212</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">128</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    cat("(", x$n_singular_coefs, " not defined because of singularities)", sep = "")</pre>
+                      <pre class="language-r">    pass &lt;- nchar(deparse(rhs))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">213</td>
+                    <td class="num">129</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">214</td>
-                    <td class="coverage">5<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n")</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">215</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">130</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  stats::printCoefmat(</pre>
+                      <pre class="language-r">  if (!is.null(call$data)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">216</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">131</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x$coefficients,</pre>
+                      <pre class="language-r">    cat(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">217</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">132</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    zap.ind = 3,</pre>
+                      <pre class="language-r">      "Data:       ", deparse(call$data), "(used", n_obs, "observations from",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">218</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">133</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    digits = digits,</pre>
+                      <pre class="language-r">      n_subjects, "subjects with maximum", n_timepoints, "timepoints)",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">219</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">134</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    signif.stars = signif.stars</pre>
+                      <pre class="language-r">      fill = TRUE</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">220</td>
+                    <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">221</td>
-                    <td class="coverage">5<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n")</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">222</td>
-                    <td class="coverage">5<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  cat("Covariance estimate:\n")</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">223</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">136</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.list(x$varcor)) {</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">224</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">137</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    for (v in names(x$varcor)) {</pre>
+                      <pre class="language-r">  # Display the expression of weights</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">225</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">138</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      cat(sprintf("Group: %s\n", v))</pre>
+                      <pre class="language-r">  if (!is.null(call$weights)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">226</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">139</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      print(round(x$varcor[[v]], digits = digits))</pre>
+                      <pre class="language-r">    cat("Weights:    ", deparse(call$weights), fill = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">227</td>
+                    <td class="num">140</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    }</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">228</td>
+                    <td class="num">141</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">229</td>
-                    <td class="coverage">4<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    print(round(x$varcor, digits = digits))</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">230</td>
+                    <td class="num">142</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">231</td>
-                    <td class="coverage">5<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  cat("\n")</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">232</td>
-                    <td class="coverage">5<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  invisible(x)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">233</td>
+                    <td class="num">143</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#' Printing MMRM Covariance Type</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">234</td>
+                    <td class="num">144</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">235</td>
+                    <td class="num">145</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' This is used in [print.summary.mmrm()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">236</td>
+                    <td class="num">146</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_methods obtain the confidence intervals for the coefficients.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">237</td>
+                    <td class="num">147</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">#' @param cov_type (`string`)\cr covariance structure abbreviation.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">238</td>
+                    <td class="num">148</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @param n_theta (`count`)\cr number of variance parameters.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">239</td>
+                    <td class="num">149</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Confidence Interval:</pre>
+                      <pre class="language-r">#' @param n_groups (`count`)\cr number of groups.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">240</td>
+                    <td class="num">150</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' confint(object)</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">241</td>
+                    <td class="num">151</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">confint.mmrm &lt;- function(object, parm, level = 0.95, ...) {</pre>
+                      <pre class="language-r">h_print_cov &lt;- function(cov_type, n_theta, n_groups) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">242</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">152</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  cf &lt;- coef(object)</pre>
+                      <pre class="language-r">  assert_string(cov_type)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">243</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">153</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  pnames &lt;- names(cf)</pre>
+                      <pre class="language-r">  assert_count(n_theta, positive = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">244</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">154</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (missing(parm)) {</pre>
+                      <pre class="language-r">  assert_count(n_groups, positive = TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">245</td>
-                    <td class="coverage">15<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    parm &lt;- pnames</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">246</td>
-                    <td class="coverage"></td>
+                    <td class="num">155</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  cov_definition &lt;- switch(cov_type,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">247</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">156</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert(</pre>
+                      <pre class="language-r">    us = "unstructured",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">248</td>
-                    <td class="coverage">20<em>x</em></td>
+                    <td class="num">157</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    check_subset(parm, pnames),</pre>
+                      <pre class="language-r">    toep = "Toeplitz",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">249</td>
-                    <td class="coverage">20<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    check_integerish(parm, lower = 1L, upper = length(cf))</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">250</td>
-                    <td class="coverage"></td>
+                    <td class="num">158</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    toeph = "heterogeneous Toeplitz",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">251</td>
-                    <td class="coverage">2<em>x</em></td>
+                    <td class="num">159</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.numeric(parm)) parm &lt;- pnames[parm]</pre>
+                      <pre class="language-r">    ar1 = "auto-regressive order one",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">252</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">160</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(level, lower = 0, upper = 1)</pre>
+                      <pre class="language-r">    ar1h = "heterogeneous auto-regressive order one",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">253</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">161</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  a &lt;- (1 - level) / 2</pre>
+                      <pre class="language-r">    ad = "ante-dependence",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">254</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">162</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  pct &lt;- paste(format(100 * c(a, 1 - a), trim = TRUE, scientific = FALSE, digits = 3), "%")</pre>
+                      <pre class="language-r">    adh = "heterogeneous ante-dependence",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">255</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">163</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coef_table &lt;- h_coef_table(object)</pre>
+                      <pre class="language-r">    cs = "compound symmetry",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">256</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">164</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- coef_table[parm, "df"]</pre>
+                      <pre class="language-r">    csh = "heterogeneous compound symmetry",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">257</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">165</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ses &lt;- coef_table[parm, "Std. Error"]</pre>
+                      <pre class="language-r">    sp_exp = "spatial exponential"</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">258</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">166</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  fac &lt;- stats::qt(a, df = df)</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">259</td>
-                    <td class="coverage">18<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">167</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ci &lt;- array(NA, dim = c(length(parm), 2L), dimnames = list(parm, pct))</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">260</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">168</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  sefac &lt;- ses * fac</pre>
+                      <pre class="language-r">  catstr &lt;- sprintf(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">261</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">169</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ci[] &lt;- cf[parm] + c(sefac, -sefac)</pre>
+                      <pre class="language-r">    "Covariance:  %s (%d variance parameters%s)\n",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">262</td>
-                    <td class="coverage">18<em>x</em></td>
+                    <td class="num">170</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  ci</pre>
+                      <pre class="language-r">    cov_definition,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">263</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">171</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">    n_theta,</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/tidiers.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
-                  <tr class="never">
-                    <td class="num">1</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">172</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Tidying Methods for `mmrm` Objects</pre>
+                      <pre class="language-r">    ifelse(n_groups == 1L, "", sprintf(" of %d groups", n_groups))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">173</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">3</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">174</td>
+                    <td class="coverage">9<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description `r lifecycle::badge("stable")`</pre>
+                      <pre class="language-r">  cat(catstr)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">175</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">176</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' These methods tidy the estimates from an `mmrm` object into a</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">177</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' summary.</pre>
+                      <pre class="language-r">#' Printing AIC and other Model Fit Criteria</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">178</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">179</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`mmrm`)\cr fitted model.</pre>
+                      <pre class="language-r">#' This is used in [print.summary.mmrm()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">180</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param conf.int (`flag`)\cr if `TRUE` columns for the lower (`conf.low`) and upper bounds</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">10</td>
+                    <td class="num">181</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   (`conf.high`) of coefficient estimates are included.</pre>
+                      <pre class="language-r">#' @param aic_list (`list`)\cr list as part of from [summary.mmrm()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">11</td>
+                    <td class="num">182</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param conf.level (`number`)\cr defines the range of the optional confidence internal.</pre>
+                      <pre class="language-r">#' @param digits (`number`)\cr number of decimal places used with [round()].</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">183</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param newdata (`data.frame` or `NULL`)\cr optional new data frame.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">13</td>
+                    <td class="num">184</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param se_fit (`flag`)\cr whether to return standard errors of fit.</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">14</td>
+                    <td class="num">185</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param interval (`string`)\cr type of interval calculation.</pre>
+                      <pre class="language-r">h_print_aic_list &lt;- function(aic_list,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">15</td>
+                    <td class="num">186</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param type.residuals (`string`)\cr passed on to [residuals.mmrm_tmb()].</pre>
+                      <pre class="language-r">                             digits = 1) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">16</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">187</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... only used by `augment()` to pass arguments to the [predict.mmrm_tmb()] method.</pre>
+                      <pre class="language-r">  diag_vals &lt;- round(unlist(aic_list), digits)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">17</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">188</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  diag_vals &lt;- format(diag_vals)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">18</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">189</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @name mmrm_tidiers</pre>
+                      <pre class="language-r">  print(diag_vals, quote = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">19</td>
+                    <td class="num">190</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @aliases mmrm_tidiers</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">20</td>
+                    <td class="num">191</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">21</td>
+                    <td class="num">192</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @seealso [`mmrm_methods`], [`mmrm_tmb_methods`] for additional methods.</pre>
+                      <pre class="language-r">#' @describeIn mmrm_methods prints the MMRM fit summary.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">22</td>
+                    <td class="num">193</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">23</td>
+                    <td class="num">194</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">24</td>
+                    <td class="num">195</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit &lt;- mmrm(</pre>
+                      <pre class="language-r">print.summary.mmrm &lt;- function(x,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">25</td>
+                    <td class="num">196</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   formula = FEV1 ~ RACE + SEX + ARMCD * AVISIT + us(AVISIT | USUBJID),</pre>
+                      <pre class="language-r">                               digits = max(3, getOption("digits") - 3),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">26</td>
+                    <td class="num">197</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   data = fev_data</pre>
+                      <pre class="language-r">                               signif.stars = getOption("show.signif.stars"), # nolint</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">27</td>
+                    <td class="num">198</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' )</pre>
+                      <pre class="language-r">                               ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">28</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">199</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">NULL</pre>
+                      <pre class="language-r">  cat("mmrm fit\n\n")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">29</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">200</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  h_print_call(x$call, x$n_obs, x$n_subjects, x$n_timepoints)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">30</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">201</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tidiers derives tidy `tibble` from an `mmrm` object.</pre>
+                      <pre class="language-r">  h_print_cov(x$cov_type, x$n_theta, x$n_groups)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">31</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">202</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">  cat("Method:      ", x$method, "\n", sep = "")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">32</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">203</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">  cat("Vcov Method: ", x$vcov, "\n", sep = "")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">33</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">204</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Applying tidy method to return summary table of covariate estimates.</pre>
+                      <pre class="language-r">  cat("Inference:   ")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">34</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">205</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit |&gt; tidy()</pre>
+                      <pre class="language-r">  cat(ifelse(x$reml, "REML", "ML"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">35</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">206</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit |&gt; tidy(conf.int = TRUE, conf.level = 0.9)</pre>
+                      <pre class="language-r">  cat("\n\n")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">36</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">207</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">tidy.mmrm &lt;- function(x, # nolint</pre>
+                      <pre class="language-r">  cat("Model selection criteria:\n")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">37</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">208</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      conf.int = FALSE, # nolint</pre>
+                      <pre class="language-r">  h_print_aic_list(x$aic_list)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">38</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">209</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      conf.level = 0.95, # nolint</pre>
+                      <pre class="language-r">  cat("\n")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">39</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">210</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                      ...) {</pre>
+                      <pre class="language-r">  cat("Coefficients: ")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">40</td>
+                    <td class="num">211</td>
                     <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(conf.int)</pre>
+                      <pre class="language-r">  if (x$n_singular_coefs &gt; 0) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">41</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">212</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_number(conf.level, lower = 0, upper = 1)</pre>
+                      <pre class="language-r">    cat("(", x$n_singular_coefs, " not defined because of singularities)", sep = "")</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">213</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">42</td>
+                    <td class="num">214</td>
                     <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl &lt;- tibble::as_tibble(summary(x)$coefficients, rownames = "term")</pre>
+                      <pre class="language-r">  cat("\n")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">43</td>
+                    <td class="num">215</td>
                     <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  colnames(tbl) &lt;- c("term", "estimate", "std.error", "df", "statistic", "p.value")</pre>
+                      <pre class="language-r">  stats::printCoefmat(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">44</td>
+                    <td class="num">216</td>
                     <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  coefs &lt;- coef(x)</pre>
+                      <pre class="language-r">    x$coefficients,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">45</td>
+                    <td class="num">217</td>
                     <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (length(coefs) != nrow(tbl)) {</pre>
+                      <pre class="language-r">    zap.ind = 3,</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">46</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">218</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    coefs &lt;- tibble::enframe(coefs, name = "term", value = "estimate")</pre>
+                      <pre class="language-r">    digits = digits,</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
-                    <td class="num">47</td>
-                    <td class="coverage">!</td>
+                  <tr class="covered">
+                    <td class="num">219</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl &lt;- merge(coefs, tbl, by = c("term", "estimate"))</pre>
+                      <pre class="language-r">    signif.stars = signif.stars</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">48</td>
+                    <td class="num">220</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">49</td>
+                    <td class="num">221</td>
                     <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (conf.int) {</pre>
+                      <pre class="language-r">  cat("\n")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">50</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">222</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ci &lt;- h_tbl_confint_terms(x, level = conf.level)</pre>
+                      <pre class="language-r">  cat("Covariance estimate:\n")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">51</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">223</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl &lt;- tibble::as_tibble(merge(tbl, ci, by = "term"))</pre>
+                      <pre class="language-r">  if (is.list(x$varcor)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">52</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">224</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    for (v in names(x$varcor)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">53</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">225</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl</pre>
+                      <pre class="language-r">      cat(sprintf("Group: %s\n", v))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">54</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">226</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">      print(round(x$varcor[[v]], digits = digits))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">55</td>
+                    <td class="num">227</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">56</td>
+                    <td class="num">228</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tidiers derives `glance` `tibble` from an `mmrm` object.</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">57</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">229</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @exportS3Method</pre>
+                      <pre class="language-r">    print(round(x$varcor, digits = digits))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">58</td>
+                    <td class="num">230</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @examples</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">59</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">231</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Applying glance method to return summary table of goodness of fit statistics.</pre>
+                      <pre class="language-r">  cat("\n")</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">60</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">232</td>
+                    <td class="coverage">5<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit |&gt; glance()</pre>
+                      <pre class="language-r">  invisible(x)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">61</td>
+                    <td class="num">233</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">glance.mmrm &lt;- function(x, ...) { # nolint</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">62</td>
-                    <td class="coverage">1<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  tibble::as_tibble(summary(x)$aic_list)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">63</td>
+                    <td class="num">234</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">64</td>
+                    <td class="num">235</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">65</td>
+                    <td class="num">236</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @describeIn mmrm_tidiers derives `augment` `tibble` from an `mmrm` object.</pre>
+                      <pre class="language-r">#' @describeIn mmrm_methods obtain the confidence intervals for the coefficients.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">66</td>
+                    <td class="num">237</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @exportS3Method</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">67</td>
+                    <td class="num">238</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @examples</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">68</td>
+                    <td class="num">239</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' # Applying augment method to return merged `tibble` of model data, fitted and residuals.</pre>
+                      <pre class="language-r">#' # Confidence Interval:</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">69</td>
+                    <td class="num">240</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit |&gt; augment()</pre>
+                      <pre class="language-r">#' confint(object)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">70</td>
+                    <td class="num">241</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit |&gt; augment(interval = "confidence")</pre>
+                      <pre class="language-r">confint.mmrm &lt;- function(object, parm, level = 0.95, ...) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">71</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">242</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' fit |&gt; augment(type.residuals = "pearson")</pre>
+                      <pre class="language-r">  cf &lt;- coef(object)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">72</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">243</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">augment.mmrm &lt;- function(x, # nolint</pre>
+                      <pre class="language-r">  pnames &lt;- names(cf)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">73</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">244</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                         newdata = NULL,</pre>
+                      <pre class="language-r">  if (missing(parm)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">74</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">245</td>
+                    <td class="coverage">15<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                         interval = c("none", "confidence", "prediction"),</pre>
+                      <pre class="language-r">    parm &lt;- pnames</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">75</td>
+                    <td class="num">246</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                         se_fit = (interval != "none"),</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">76</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">247</td>
+                    <td class="coverage">20<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                         type.residuals = c("response", "pearson", "normalized"), # nolint</pre>
+                      <pre class="language-r">  assert(</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">248</td>
+                    <td class="coverage">20<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    check_subset(parm, pnames),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">249</td>
+                    <td class="coverage">20<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    check_integerish(parm, lower = 1L, upper = length(cf))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">77</td>
+                    <td class="num">250</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                         ...) {</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">78</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">251</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  type.residuals &lt;- match.arg(type.residuals) # nolint</pre>
+                      <pre class="language-r">  if (is.numeric(parm)) parm &lt;- pnames[parm]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">79</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">252</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  resid_df &lt;- NULL</pre>
+                      <pre class="language-r">  assert_number(level, lower = 0, upper = 1)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">80</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">253</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (is.null(newdata)) {</pre>
+                      <pre class="language-r">  a &lt;- (1 - level) / 2</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">81</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">254</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    newdata &lt;- stats::get_all_vars(x, data = stats::na.omit(x$data))</pre>
+                      <pre class="language-r">  pct &lt;- paste(format(100 * c(a, 1 - a), trim = TRUE, scientific = FALSE, digits = 3), "%")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">82</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">255</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    resid_df &lt;- data.frame(</pre>
+                      <pre class="language-r">  coef_table &lt;- h_coef_table(object)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">83</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">256</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      .rownames = rownames(newdata),</pre>
+                      <pre class="language-r">  df &lt;- coef_table[parm, "df"]</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">84</td>
-                    <td class="coverage">4<em>x</em></td>
+                    <td class="num">257</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      .resid = unname(residuals(x, type = type.residuals))</pre>
+                      <pre class="language-r">  ses &lt;- coef_table[parm, "Std. Error"]</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">85</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">258</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">  fac &lt;- stats::qt(a, df = df)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">86</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">259</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  ci &lt;- array(NA, dim = c(length(parm), 2L), dimnames = list(parm, pct))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">260</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  sefac &lt;- ses * fac</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">261</td>
+                    <td class="coverage">18<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  ci[] &lt;- cf[parm] + c(sefac, -sefac)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">87</td>
-                    <td class="coverage">9<em>x</em></td>
+                    <td class="num">262</td>
+                    <td class="coverage">18<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  interval &lt;- match.arg(interval)</pre>
+                      <pre class="language-r">  ci</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">88</td>
+                    <td class="num">263</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">89</td>
-                    <td class="coverage">9<em>x</em></td>
+                </tbody>
+              </table>
+            </div>
+            <div id="R/zzz.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
+                  <tr class="never">
+                    <td class="num">1</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl &lt;- h_newdata_add_pred(</pre>
+                      <pre class="language-r">#' Dynamic Registration for Package Interoperability</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">90</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">2</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">91</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">3</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    newdata = newdata,</pre>
+                      <pre class="language-r">#' @seealso See `vignette("xtending", package = "emmeans")` for background.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">92</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">4</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    se_fit = se_fit,</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">93</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">5</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    interval = interval,</pre>
+                      <pre class="language-r">#' @noRd</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">94</td>
+                    <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ...</pre>
+                      <pre class="language-r">.onLoad &lt;- function(libname, pkgname) { # nolint</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">95</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">7</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  if (!h_tmb_version_sufficient()) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">96</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">8</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!is.null(resid_df)) {</pre>
+                      <pre class="language-r">    msg &lt;- paste(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">97</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">9</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl &lt;- merge(tbl, resid_df, by = ".rownames")</pre>
+                      <pre class="language-r">      "TMB below version 1.9.15 has been used to compile the mmrm package.",</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">98</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">10</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl$.rownames &lt;- as.numeric(tbl$.rownames)</pre>
+                      <pre class="language-r">      "Reproducible model fits are not guaranteed.",</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">99</td>
-                    <td class="coverage">4<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">11</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl &lt;- tbl[order(tbl$.rownames), , drop = FALSE]</pre>
+                      <pre class="language-r">      "Please consider recompiling the package with TMB version 1.9.15 or higher."</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">100</td>
+                    <td class="num">12</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">101</td>
-                    <td class="coverage">9<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">13</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tibble::as_tibble(tbl)</pre>
+                      <pre class="language-r">    warning(msg, call. = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">102</td>
+                    <td class="num">14</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">103</td>
+                    <td class="num">15</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">104</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">16</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Extract `tibble` with Confidence Intervals and Term Names</pre>
+                      <pre class="language-r">  register_on_load(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">105</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">17</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    "emmeans", c("1.6", NA),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">106</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">18</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used in [tidy.mmrm()].</pre>
+                      <pre class="language-r">    callback = function() emmeans::.emm_register("mmrm", pkgname),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">107</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">19</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    message = "mmrm() registered as emmeans extension"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">108</td>
+                    <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`mmrm`)\cr fit object.</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">109</td>
+                    <td class="num">21</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... passed to [stats::confint()], hence not used at the moment.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">110</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">22</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  register_on_load(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">111</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">23</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A `tibble` with `term`, `conf.low`, `conf.high` columns.</pre>
+                      <pre class="language-r">    "parsnip", c("1.1.0", NA),</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">112</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">24</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    callback = parsnip_add_mmrm,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">113</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">25</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    message = emit_tidymodels_register_msg</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">114</td>
+                    <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_tbl_confint_terms &lt;- function(x, ...) {</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">115</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">27</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- stats::confint(x, ...)</pre>
+                      <pre class="language-r">  register_on_load(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">116</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">28</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl &lt;- tibble::as_tibble(df, rownames = "term", .name_repair = "minimal")</pre>
+                      <pre class="language-r">    "car", c("3.1.2", NA),</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">117</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">29</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  names(tbl) &lt;- c("term", "conf.low", "conf.high")</pre>
+                      <pre class="language-r">    callback = car_add_mmrm,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">118</td>
-                    <td class="coverage">8<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">30</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl</pre>
+                      <pre class="language-r">    message = "mmrm() registered as car::Anova extension"</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">119</td>
+                    <td class="num">31</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">120</td>
+                    <td class="num">32</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">121</td>
+                    <td class="num">33</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Add Prediction Results to New Data</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">122</td>
+                    <td class="num">34</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' Helper Function for Registering Functionality With Suggests Packages</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">123</td>
+                    <td class="num">35</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used in [augment.mmrm()].</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">124</td>
+                    <td class="num">36</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @inheritParams check_package_version</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">125</td>
+                    <td class="num">37</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param x (`mmrm`)\cr fit.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">126</td>
+                    <td class="num">38</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param newdata (`data.frame`)\cr data to predict.</pre>
+                      <pre class="language-r">#' @param callback (`function(...) ANY`)\cr a callback to execute upon package</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">127</td>
+                    <td class="num">39</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param se_fit (`flag`)\cr whether to return standard error of prediction,</pre>
+                      <pre class="language-r">#'   load. Note that no arguments are passed to this function. Any necessary</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">128</td>
+                    <td class="num">40</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   can only be used when `interval` is not "none".</pre>
+                      <pre class="language-r">#'   data must be provided upon construction.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">129</td>
+                    <td class="num">41</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param interval (`string`)\cr type of interval.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">130</td>
+                    <td class="num">42</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param ... passed to [predict.mmrm_tmb()].</pre>
+                      <pre class="language-r">#' @param message (`NULL` or `string`)\cr an optional message to print after</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">131</td>
+                    <td class="num">43</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#'   the callback is executed upon successful registration.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">132</td>
+                    <td class="num">44</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The `newdata` as a `tibble` with additional columns `.fitted`,</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">133</td>
+                    <td class="num">45</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   `.lower`, `.upper` (if interval is not `none`) and `.se.fit` (if `se_fit`</pre>
+                      <pre class="language-r">#' @return A logical (invisibly) indicating whether registration was successful.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">134</td>
+                    <td class="num">46</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   requested).</pre>
+                      <pre class="language-r">#'  If not, a onLoad hook was set for the next time the package is loaded.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">135</td>
+                    <td class="num">47</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">136</td>
+                    <td class="num">48</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">137</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">h_newdata_add_pred &lt;- function(x,</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">138</td>
+                    <td class="num">49</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                               newdata,</pre>
+                      <pre class="language-r">register_on_load &lt;- function(pkg,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">139</td>
+                    <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                               se_fit,</pre>
+                      <pre class="language-r">                             ver = c(NA_character_, NA_character_),</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">140</td>
+                    <td class="num">51</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                               interval,</pre>
+                      <pre class="language-r">                             callback,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">141</td>
+                    <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">                               ...) {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">142</td>
-                    <td class="coverage">13<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(x, "mmrm")</pre>
+                      <pre class="language-r">                             message = NULL) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">143</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">53</td>
+                    <td class="coverage">4<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_data_frame(newdata)</pre>
+                      <pre class="language-r">  if (isNamespaceLoaded(pkg) &amp;&amp; check_package_version(pkg, ver)) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">144</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">54</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_flag(se_fit)</pre>
+                      <pre class="language-r">    callback()</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">145</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">55</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_string(interval)</pre>
+                      <pre class="language-r">    if (is.character(message)) packageStartupMessage(message)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">146</td>
-                    <td class="coverage">13<em>x</em></td>
+                    <td class="num">56</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (interval == "none") {</pre>
+                      <pre class="language-r">    if (is.function(message)) packageStartupMessage(message())</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">147</td>
-                    <td class="coverage">7<em>x</em></td>
+                    <td class="num">57</td>
+                    <td class="coverage">3<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_false(se_fit)</pre>
+                      <pre class="language-r">    return(invisible(TRUE))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">148</td>
+                    <td class="num">58</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">149</td>
+                    <td class="num">59</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">150</td>
-                    <td class="coverage">12<em>x</em></td>
+                    <td class="num">60</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl &lt;- h_df_to_tibble(newdata)</pre>
+                      <pre class="language-r">  setHook(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">151</td>
-                    <td class="coverage">12<em>x</em></td>
+                    <td class="num">61</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  pred_results &lt;- predict(</pre>
+                      <pre class="language-r">    packageEvent(pkg, event = "onLoad"),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">152</td>
-                    <td class="coverage">12<em>x</em></td>
+                    <td class="num">62</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    x,</pre>
+                      <pre class="language-r">    action = "append",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">153</td>
-                    <td class="coverage">12<em>x</em></td>
+                    <td class="num">63</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    newdata = newdata,</pre>
+                      <pre class="language-r">    function(...) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">154</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">64</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    na.action = stats::na.pass,</pre>
+                      <pre class="language-r">      register_on_load(</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">155</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">65</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    se.fit = se_fit,</pre>
+                      <pre class="language-r">        pkg = pkg,</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">156</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">66</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    interval = interval,</pre>
+                      <pre class="language-r">        ver = ver,</pre>
+                    </td>
+                  </tr>
+                  <tr class="missed">
+                    <td class="num">67</td>
+                    <td class="coverage">!</td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">        callback = callback,</pre>
+                    </td>
+                  </tr>
+                  <tr class="missed">
+                    <td class="num">68</td>
+                    <td class="coverage">!</td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">        message = message</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">157</td>
+                    <td class="num">69</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    ...</pre>
+                      <pre class="language-r">      )</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">158</td>
+                    <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">    }</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">159</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">71</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (interval == "none") {</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">160</td>
-                    <td class="coverage">6<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">72</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    assert_numeric(pred_results)</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">161</td>
-                    <td class="coverage">6<em>x</em></td>
+                    <td class="num">73</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl$.fitted &lt;- unname(pred_results)</pre>
+                      <pre class="language-r">  invisible(FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">162</td>
+                    <td class="num">74</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  } else {</pre>
-                    </td>
-                  </tr>
-                  <tr class="covered">
-                    <td class="num">163</td>
-                    <td class="coverage">6<em>x</em></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r">    assert_matrix(pred_results)</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">164</td>
-                    <td class="coverage">6<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">75</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl$.fitted &lt;- unname(pred_results[, "fit"])</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">165</td>
-                    <td class="coverage">6<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">76</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl$.lower &lt;- unname(pred_results[, "lwr"])</pre>
+                      <pre class="language-r">#' Check Suggested Dependency Against Version Requirements</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">166</td>
-                    <td class="coverage">6<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">77</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl$.upper &lt;- unname(pred_results[, "upr"])</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">167</td>
+                    <td class="num">78</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#' @param pkg (`string`)\cr package name.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">168</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">79</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (se_fit) {</pre>
+                      <pre class="language-r">#' @param ver (`character`)\cr of length 2 whose elements can be provided to</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">169</td>
-                    <td class="coverage">5<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">80</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl$.se.fit &lt;- unname(pred_results[, "se"])</pre>
+                      <pre class="language-r">#'   [numeric_version()], representing a minimum and maximum (inclusive) version</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">170</td>
+                    <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">#'   requirement for interoperability. When `NA`, no version requirement is</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">171</td>
-                    <td class="coverage">12<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">82</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl</pre>
+                      <pre class="language-r">#'   imposed. Defaults to no version requirement.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">172</td>
+                    <td class="num">83</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">173</td>
+                    <td class="num">84</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @return A logical (invisibly) indicating whether the loaded package meets</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">174</td>
+                    <td class="num">85</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Coerce a Data Frame to a `tibble`</pre>
+                      <pre class="language-r">#'   the version requirements. A warning is emitted otherwise.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">175</td>
+                    <td class="num">86</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">176</td>
+                    <td class="num">87</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' This is used in [h_newdata_add_pred()].</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">177</td>
+                    <td class="num">88</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">check_package_version &lt;- function(pkg, ver = c(NA_character_, NA_character_)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">178</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">89</td>
+                    <td class="coverage">7<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details This is only a thin wrapper around [tibble::as_tibble()], except</pre>
+                      <pre class="language-r">  assert_character(ver, len = 2L)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">179</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">90</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' giving a useful error message and it checks for `rownames` and adds them</pre>
+                      <pre class="language-r">  pkg_ver &lt;- utils::packageVersion(pkg)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">180</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">91</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' as a new column `.rownames` if they are not just a numeric sequence as</pre>
+                      <pre class="language-r">  ver &lt;- numeric_version(ver, strict = FALSE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">181</td>
+                    <td class="num">92</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' per the [tibble::has_rownames()] decision.</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">182</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">93</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  warn_version &lt;- function(pkg, pkg_ver, ver) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">183</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">94</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @param data (`data.frame`)\cr what to coerce.</pre>
+                      <pre class="language-r">    ver_na &lt;- is.na(ver)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">184</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">95</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    warning(sprintf(</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">185</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">96</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return The `data` as a `tibble`, potentially with a `.rownames` column.</pre>
+                      <pre class="language-r">      "Cannot register mmrm for use with %s (v%s). Version %s required.",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">186</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">97</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">      pkg, pkg_ver,</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">187</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">98</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">      if (!any(ver_na)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">188</td>
-                    <td class="coverage"></td>
+                  <tr class="missed">
+                    <td class="num">99</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_to_tibble &lt;- function(data) {</pre>
+                      <pre class="language-r">        sprintf("%s to %s", ver[1], ver[2])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">189</td>
-                    <td class="coverage">15<em>x</em></td>
+                    <td class="num">100</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tryCatch(tbl &lt;- tibble::as_tibble(data), error = function(cnd) {</pre>
+                      <pre class="language-r">      } else if (ver_na[2]) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">190</td>
+                    <td class="num">101</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    stop("Could not coerce data to `tibble`. Try explicitly passing a",</pre>
+                      <pre class="language-r">        paste0("&gt;= ", ver[1])</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">191</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="num">102</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      "dataset to either the `data` or `newdata` argument.",</pre>
+                      <pre class="language-r">      } else if (ver_na[1]) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">192</td>
+                    <td class="num">103</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      call. = FALSE</pre>
+                      <pre class="language-r">        paste0("&lt;= ", ver[2])</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">193</td>
+                    <td class="num">104</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">      }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">194</td>
+                    <td class="num">105</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  })</pre>
+                      <pre class="language-r">    ))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">195</td>
-                    <td class="coverage">14<em>x</em></td>
+                  <tr class="never">
+                    <td class="num">106</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (tibble::has_rownames(data)) {</pre>
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">107</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">196</td>
-                    <td class="coverage">5<em>x</em></td>
+                    <td class="num">108</td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    tbl &lt;- tibble::add_column(tbl, .rownames = rownames(data), .before = TRUE)</pre>
+                      <pre class="language-r">  if (identical(pkg_ver &lt; ver[1], TRUE) || identical(pkg_ver &gt; ver[2], TRUE)) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">197</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">109</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">    warn_version(pkg, pkg_ver, ver)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">198</td>
-                    <td class="coverage">14<em>x</em></td>
+                    <td class="num">110</td>
+                    <td class="coverage">2<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  tbl</pre>
+                      <pre class="language-r">    return(invisible(FALSE))</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">199</td>
+                    <td class="num">111</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                </tbody>
-              </table>
-            </div>
-            <div id="R/residual.R" class="hidden">
-              <table class="table-condensed">
-                <tbody>
                   <tr class="never">
-                    <td class="num">1</td>
+                    <td class="num">112</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Residual Degrees of Freedom for One-Dimensional Contrast</pre>
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">113</td>
+                    <td class="coverage">4<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  invisible(TRUE)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">2</td>
+                    <td class="num">114</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">3</td>
+                    <td class="num">115</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_1d()] if method is</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">4</td>
+                    <td class="num">116</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' "Residual".</pre>
+                      <pre class="language-r">#' Format a Message to Emit When Tidymodels is Loaded</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">5</td>
+                    <td class="num">117</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">6</td>
+                    <td class="num">118</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_df_1d_sat</pre>
+                      <pre class="language-r">#' @return A character message to emit. Either a ansi-formatted cli output if</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">7</td>
+                    <td class="num">119</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inherit h_df_1d_sat return</pre>
+                      <pre class="language-r">#'   package 'cli' is available or a plain-text message otherwise.</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">8</td>
+                    <td class="num">120</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#'</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">121</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">9</td>
+                    <td class="num">122</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_1d_res &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">emit_tidymodels_register_msg &lt;- function() {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">10</td>
+                    <td class="num">123</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">  pkg &lt;- utils::packageName()</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">11</td>
+                    <td class="num">124</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">  ver &lt;- utils::packageVersion(pkg)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">12</td>
+                    <td class="num">125</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">13</td>
+                    <td class="num">126</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- component(object, "n_obs") - length(component(object, "beta_est"))</pre>
-                    </td>
-                  </tr>
-                  <tr class="never">
-                    <td class="num">14</td>
-                    <td class="coverage"></td>
-                    <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  if (isTRUE(getOption("tidymodels.quiet"))) {</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
-                    <td class="num">15</td>
-                    <td class="coverage">1<em>x</em></td>
+                  <tr class="missed">
+                    <td class="num">127</td>
+                    <td class="coverage">!</td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_test_1d(object, contrast, df)</pre>
+                      <pre class="language-r">    return()</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">16</td>
+                    <td class="num">128</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">}</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">17</td>
+                    <td class="num">129</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">18</td>
+                    <td class="num">130</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Calculation of Residual Degrees of Freedom for Multi-Dimensional Contrast</pre>
+                      <pre class="language-r">  # if tidymodels is attached, cli packages come as a dependency</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">19</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">131</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">  has_cli &lt;- requireNamespace("cli", quietly = TRUE)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">20</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">132</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @description Used in [df_md()] if method is "Residual".</pre>
+                      <pre class="language-r">  if (has_cli) {</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">21</td>
+                    <td class="num">133</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">    # unfortunately, cli does not expose many formatting tools for emitting</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">22</td>
+                    <td class="num">134</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams h_df_md_sat</pre>
+                      <pre class="language-r">    # messages (only via conditions to stderr) which can't be suppressed using</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">23</td>
+                    <td class="num">135</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inherit h_df_md_sat return</pre>
+                      <pre class="language-r">    # suppressPackageStartupMessages() so formatting must be done adhoc,</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">24</td>
+                    <td class="num">136</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @keywords internal</pre>
+                      <pre class="language-r">    # similar to how it's done in {tidymodels} R/attach.R</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">25</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">137</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">h_df_md_res &lt;- function(object, contrast) {</pre>
+                      <pre class="language-r">    paste0(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">26</td>
+                    <td class="num">138</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                      <pre class="language-r">      cli::rule(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">27</td>
+                    <td class="num">139</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
+                      <pre class="language-r">        left = cli::style_bold("Model Registration"),</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">140</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">        right = paste(pkg, ver)</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">28</td>
+                    <td class="num">141</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      ),</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">29</td>
+                    <td class="num">142</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  df &lt;- component(object, "n_obs") - length(component(object, "beta_est"))</pre>
+                      <pre class="language-r">      "\n",</pre>
                     </td>
                   </tr>
-                  <tr class="never">
-                    <td class="num">30</td>
-                    <td class="coverage"></td>
+                  <tr class="covered">
+                    <td class="num">143</td>
+                    <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">      cli::col_green(cli::symbol$tick), " ",</pre>
                     </td>
                   </tr>
                   <tr class="covered">
-                    <td class="num">31</td>
+                    <td class="num">144</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  h_test_md(object, contrast, df)</pre>
+                      <pre class="language-r">      cli::col_blue("mmrm"), "::", cli::col_green("mmrm()")</pre>
                     </td>
                   </tr>
                   <tr class="never">
-                    <td class="num">32</td>
+                    <td class="num">145</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    )</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">146</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  } else {</pre>
+                    </td>
+                  </tr>
+                  <tr class="missed">
+                    <td class="num">147</td>
+                    <td class="coverage">!</td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    paste0(pkg, "::mmrm() registered for use with tidymodels")</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">148</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">149</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
@@ -33324,14 +33136,14 @@ <h2>mmrm coverage - 97.05%</h2>
                 </tbody>
               </table>
             </div>
-            <div id="R/interop-parsnip.R" class="hidden">
+            <div id="R/utils-formula.R" class="hidden">
               <table class="table-condensed">
                 <tbody>
                   <tr class="never">
                     <td class="num">1</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' Register `mmrm` For Use With `tidymodels`</pre>
+                      <pre class="language-r">#' Extract Formula Terms used for Covariance Structure Definition</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -33345,14 +33157,14 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">3</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @inheritParams base::requireNamespace</pre>
+                      <pre class="language-r">#' @param f (`formula`)\cr a formula from which covariance terms should be</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">4</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @return A logical value indicating whether registration was successful.</pre>
+                      <pre class="language-r">#'   extracted.</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -33366,21 +33178,21 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">6</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#' @details We can use `parsnip::show_model_info("linear_reg")` to check the</pre>
+                      <pre class="language-r">#' @return A list of covariance structure expressions found in `f`.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">7</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'   registration with `parsnip` and thus the wider `tidymodels` ecosystem.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">8</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">#'</pre>
+                      <pre class="language-r">#' @importFrom stats terms</pre>
                     </td>
                   </tr>
                   <tr class="never">
@@ -33394,516 +33206,593 @@ <h2>mmrm coverage - 97.05%</h2>
                     <td class="num">10</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">parsnip_add_mmrm &lt;- function(quietly = FALSE) {</pre>
+                      <pre class="language-r">h_extract_covariance_terms &lt;- function(f) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">11</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">291<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  if (!requireNamespace("parsnip", quietly = quietly)) {</pre>
+                      <pre class="language-r">  specials &lt;- cov_types(c("abbr", "habbr"))</pre>
                     </td>
                   </tr>
-                  <tr class="missed">
+                  <tr class="covered">
                     <td class="num">12</td>
-                    <td class="coverage">!</td>
+                    <td class="coverage">291<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    return(FALSE)</pre>
+                      <pre class="language-r">  terms &lt;- stats::terms(formula_rhs(f), specials = specials)</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">13</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">291<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  }</pre>
+                      <pre class="language-r">  covariance_terms &lt;- Filter(length, attr(terms, "specials"))</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">14</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">291<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  variables &lt;- attr(terms, "variables")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">15</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">291<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  parsnip::set_model_engine(</pre>
+                      <pre class="language-r">  lapply(covariance_terms, function(i) variables[[i + 1]])</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">16</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    model = "linear_reg",</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">17</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    eng = "mmrm",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">18</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "regression"</pre>
+                      <pre class="language-r">#' Drop Formula Terms used for Covariance Structure Definition</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">19</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">20</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @param f (`formula`)\cr a formula from which covariance terms should be</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">21</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  parsnip::set_dependency(</pre>
+                      <pre class="language-r">#'   dropped.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">22</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    pkg = "mmrm",</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">23</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    model = "linear_reg",</pre>
+                      <pre class="language-r">#' @return The formula without accepted covariance terms.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">24</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    eng = "mmrm",</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">25</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "regression"</pre>
+                      <pre class="language-r">#' @details `terms` is used and it will preserve the environment attribute.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">26</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' This ensures the returned formula and the input formula have the same environment.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">27</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @importFrom stats terms drop.terms</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">28</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  parsnip::set_encoding(</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">29</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    model = "linear_reg",</pre>
+                      <pre class="language-r">h_drop_covariance_terms &lt;- function(f) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">30</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">274<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    eng = "mmrm",</pre>
+                      <pre class="language-r">  specials &lt;- cov_types(c("abbr", "habbr"))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">31</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "regression",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">32</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">274<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    options = list(</pre>
+                      <pre class="language-r">  terms &lt;- stats::terms(f, specials = specials)</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">33</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">274<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      predictor_indicators = "none",</pre>
+                      <pre class="language-r">  covariance_terms &lt;- Filter(Negate(is.null), attr(terms, "specials"))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">34</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      compute_intercept = FALSE,</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">35</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      remove_intercept = FALSE,</pre>
+                      <pre class="language-r">  # if no covariance terms were found, return original formula</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">36</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">274<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      allow_sparse_x = TRUE</pre>
+                      <pre class="language-r">  if (length(covariance_terms) == 0) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">37</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">6<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">    return(f)</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">38</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">39</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">268<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  if (length(f) != 3) {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">40</td>
                     <td class="coverage">1<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  parsnip::set_fit(</pre>
+                      <pre class="language-r">    update_str &lt;- "~ . -"</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">41</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    model = "linear_reg",</pre>
+                      <pre class="language-r">  } else {</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">42</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">267<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    eng = "mmrm",</pre>
+                      <pre class="language-r">    update_str &lt;- ". ~ . -"</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">43</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "regression",</pre>
+                      <pre class="language-r">  }</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">44</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">268<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    value = list(</pre>
+                      <pre class="language-r">  stats::update(</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">45</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">268<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      interface = "formula",</pre>
+                      <pre class="language-r">    f,</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">46</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">268<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      protect = c("formula", "data", "weights"),</pre>
+                      <pre class="language-r">    stats::as.formula(paste(update_str, deparse(attr(terms, "variables")[[covariance_terms[[1]] + 1]])))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">47</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      data = c(formula = "formula", data = "data", weights = "weights"),</pre>
+                      <pre class="language-r">  )</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">48</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      func = c(pkg = "mmrm", fun = "mmrm"),</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">49</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      defaults = list()</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">50</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' Add Individual Covariance Variables As Terms to Formula</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">51</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">52</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @param f (`formula`)\cr a formula to which covariance structure terms should</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">53</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  parsnip::set_pred(</pre>
+                      <pre class="language-r">#'   be added.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">54</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    model = "linear_reg",</pre>
+                      <pre class="language-r">#' @param covariance (`cov_struct`)\cr a covariance structure object from which</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">55</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    eng = "mmrm",</pre>
+                      <pre class="language-r">#'   additional variables should be sourced.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">56</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "regression",</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">57</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    type = "numeric",</pre>
+                      <pre class="language-r">#' @return A new formula with included covariance terms.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">58</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    value = parsnip::pred_value_template(</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">59</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # This is boilerplate.</pre>
+                      <pre class="language-r">#' @details [stats::update()] is used to append the covariance structure and the environment</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">60</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      func = c(fun = "predict"),</pre>
+                      <pre class="language-r">#' attribute will not be changed. This ensures the returned formula and the input formula</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">61</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      object = quote(object$fit),</pre>
+                      <pre class="language-r">#' have the same environment.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">62</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      newdata = quote(new_data)</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">63</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' @keywords internal</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">64</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">h_add_covariance_terms &lt;- function(f, covariance) {</pre>
                     </td>
                   </tr>
-                  <tr class="never">
+                  <tr class="covered">
                     <td class="num">65</td>
-                    <td class="coverage"></td>
+                    <td class="coverage">272<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">  cov_terms &lt;- with(covariance, c(subject, visits, group))</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">66</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">266<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  parsnip::set_pred(</pre>
+                      <pre class="language-r">  cov_terms &lt;- paste(cov_terms, collapse = " + ")</pre>
                     </td>
                   </tr>
                   <tr class="covered">
                     <td class="num">67</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage">266<em>x</em></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    model = "linear_reg",</pre>
+                      <pre class="language-r">  stats::update(f, stats::as.formula(paste(". ~ . + ", cov_terms)))</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">68</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    eng = "mmrm",</pre>
+                      <pre class="language-r">}</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">69</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    mode = "regression",</pre>
+                      <pre class="language-r"></pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">70</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    # This type allows to pass arguments via `opts` to `parsnip::predict.model_fit`.</pre>
+                      <pre class="language-r">#' Add Formula Terms with Character</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">71</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    type = "raw",</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">72</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    value = parsnip::pred_value_template(</pre>
+                      <pre class="language-r">#' Add formula terms from the original formula with character representation.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">73</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # This is boilerplate.</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">74</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      func = c(fun = "predict"),</pre>
+                      <pre class="language-r">#' @param f (`formula`)\cr a formula to be updated.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">75</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      object = quote(object$fit),</pre>
+                      <pre class="language-r">#' @param adds (`character`)\cr representation of elements to be added.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">76</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      newdata = quote(new_data)</pre>
+                      <pre class="language-r">#' @param drop_response (`flag`)\cr whether response should be dropped.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">77</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # We don't specify additional argument defaults here since otherwise</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">78</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">      # the user is not able to change them (they will be fixed).</pre>
+                      <pre class="language-r">#' @details Elements in `adds` will be added from the formula, while the environment</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">79</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">    )</pre>
+                      <pre class="language-r">#' of the formula is unchanged. If `adds` is `NULL` or `character(0)`, the formula is</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">80</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  )</pre>
+                      <pre class="language-r">#' unchanged.</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">81</td>
                     <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r"></pre>
+                      <pre class="language-r">#' @return A new formula with elements in `drops` removed.</pre>
                     </td>
                   </tr>
-                  <tr class="covered">
+                  <tr class="never">
                     <td class="num">82</td>
-                    <td class="coverage">1<em>x</em></td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
-                      <pre class="language-r">  TRUE</pre>
+                      <pre class="language-r">#'</pre>
                     </td>
                   </tr>
                   <tr class="never">
                     <td class="num">83</td>
                     <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @keywords internal</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">84</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">h_add_terms &lt;- function(f, adds, drop_response = FALSE) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">85</td>
+                    <td class="coverage">599<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_character(adds, null.ok = TRUE)</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">86</td>
+                    <td class="coverage">599<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  if (length(adds) &gt; 0L) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">87</td>
+                    <td class="coverage">321<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    add_terms &lt;- stats::as.formula(sprintf(". ~ . + %s", paste(adds, collapse = "+")))</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">88</td>
+                    <td class="coverage">321<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    f &lt;- stats::update(f, add_terms)</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">89</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">90</td>
+                    <td class="coverage">599<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  if (drop_response &amp;&amp; length(f) == 3L) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">91</td>
+                    <td class="coverage">35<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">    f[[2]] &lt;- NULL</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">92</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  }</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">93</td>
+                    <td class="coverage">599<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  f</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">94</td>
+                    <td class="coverage"></td>
                     <td class="col-sm-12">
                       <pre class="language-r">}</pre>
                     </td>
@@ -35259,6 +35148,236 @@ <h2>mmrm coverage - 97.05%</h2>
                 </tbody>
               </table>
             </div>
+            <div id="R/residual.R" class="hidden">
+              <table class="table-condensed">
+                <tbody>
+                  <tr class="never">
+                    <td class="num">1</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' Calculation of Residual Degrees of Freedom for One-Dimensional Contrast</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">2</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#'</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">3</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @description Used in [df_1d()] if method is</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">4</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' "Residual".</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">5</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#'</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">6</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @inheritParams h_df_1d_sat</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">7</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @inherit h_df_1d_sat return</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">8</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @keywords internal</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">9</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">h_df_1d_res &lt;- function(object, contrast) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">10</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">11</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_numeric(contrast, len = length(component(object, "beta_est")))</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">12</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">13</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  df &lt;- component(object, "n_obs") - length(component(object, "beta_est"))</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">14</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">15</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  h_test_1d(object, contrast, df)</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">16</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">}</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">17</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">18</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' Calculation of Residual Degrees of Freedom for Multi-Dimensional Contrast</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">19</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#'</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">20</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @description Used in [df_md()] if method is "Residual".</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">21</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#'</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">22</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @inheritParams h_df_md_sat</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">23</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @inherit h_df_md_sat return</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">24</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">#' @keywords internal</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">25</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">h_df_md_res &lt;- function(object, contrast) {</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">26</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_class(object, "mmrm")</pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">27</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  assert_matrix(contrast, mode = "numeric", any.missing = FALSE, ncols = length(component(object, "beta_est")))</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">28</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">29</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  df &lt;- component(object, "n_obs") - length(component(object, "beta_est"))</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">30</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r"></pre>
+                    </td>
+                  </tr>
+                  <tr class="covered">
+                    <td class="num">31</td>
+                    <td class="coverage">1<em>x</em></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">  h_test_md(object, contrast, df)</pre>
+                    </td>
+                  </tr>
+                  <tr class="never">
+                    <td class="num">32</td>
+                    <td class="coverage"></td>
+                    <td class="col-sm-12">
+                      <pre class="language-r">}</pre>
+                    </td>
+                  </tr>
+                </tbody>
+              </table>
+            </div>
             <div id="src/chol_cache.h" class="hidden">
               <table class="table-condensed">
                 <tbody>