vcov argument not supported for plm models

Fixes #911

DESCRIPTION

@@ -1,7 +1,7 @@
 Type: Package
 Package: parameters
 Title: Processing of Model Parameters
-Version: 0.21.2.7
+Version: 0.21.2.8
 Authors@R:
     c(person(given = "Daniel",
              family = "Lüdecke",

NEWS.md

@@ -20,6 +20,8 @@
 
 * Minor fixes for `print_html()` method for `model_parameters()`.
 
+* Robust standard errors (argument `vcov`) now works for `plm` models.
+
 # parameters 0.21.2
 
 ## Changes

R/methods_plm.R

@@ -15,12 +15,41 @@ degrees_of_freedom.plm <- function(model, method = "wald", ...) {
 
 
 #' @export
-standard_error.plm <- function(model, ...) {
-  se <- stats::coef(summary(model))
+standard_error.plm <- function(model, vcov = NULL, vcov_args = NULL, verbose = TRUE, ...) {
+  dots <- list(...)
+  se <- NULL
+  se_standard <- stats::coef(summary(model))
+
+  # vcov: matrix
+  if (is.matrix(vcov)) {
+    se <- sqrt(diag(vcov))
+  }
+
+  # vcov: function which returns a matrix
+  if (is.function(vcov)) {
+    args <- c(list(model), vcov_args, dots)
+    se <- .safe(sqrt(diag(do.call("vcov", args))))
+  }
+
+  # vcov: character (with backward compatibility for `robust = TRUE`)
+  if (is.character(vcov) || isTRUE(dots[["robust"]])) {
+    .vcov <- insight::get_varcov(
+      model,
+      vcov = vcov,
+      vcov_args = vcov_args,
+      verbose = verbose,
+      ...
+    )
+    se <- sqrt(diag(.vcov))
+  }
+
+  if (is.null(se)) {
+    se <- as.vector(se_standard[, 2])
+  }
 
   .data_frame(
-    Parameter = .remove_backticks_from_string(rownames(se)),
-    SE = as.vector(se[, 2])
+    Parameter = .remove_backticks_from_string(rownames(se_standard)),
+    SE = se
   )
 }
 

tests/testthat/_snaps/plm.md

@@ -0,0 +1,34 @@
+# vcov standard errors
+
+    Code
+      print(model_parameters(ran))
+    Output
+      # Fixed Effects
+      
+      Parameter   | Coefficient |     SE |            95% CI |     z |  df |      p
+      -----------------------------------------------------------------------------
+      (Intercept) |      817.10 | 188.26 | [445.84, 1188.36] |  4.34 | 197 | < .001
+      capital     |       -0.58 |   0.15 | [ -0.88,   -0.29] | -3.95 | 197 | < .001
+      inv         |        2.92 |   0.30 | [  2.33,    3.51] |  9.80 | 197 | < .001
+    Message
+      
+      Uncertainty intervals (equal-tailed) and p-values (two-tailed) computed
+        using a Wald z-distribution approximation.
+
+---
+
+    Code
+      print(model_parameters(ran, vcov = "HC1"))
+    Output
+      # Fixed Effects
+      
+      Parameter   | Coefficient |     SE |            95% CI |     z |  df |     p
+      ----------------------------------------------------------------------------
+      (Intercept) |      817.10 | 274.75 | [275.28, 1358.92] |  2.97 | 197 | 0.003
+      capital     |       -0.58 |   0.43 | [ -1.43,    0.26] | -1.37 | 197 | 0.173
+      inv         |        2.92 |   0.89 | [  1.16,    4.67] |  3.28 | 197 | 0.001
+    Message
+      
+      Uncertainty intervals (equal-tailed) and p-values (two-tailed) computed
+        using a Wald z-distribution approximation.
+

tests/testthat/test-plm.R

@@ -33,6 +33,7 @@ m3 <- suppressWarnings(plm::plm(
   index = c("ID", "Year")
 ))
 
+
 test_that("ci", {
   expect_equal(
     ci(m1)$CI_low,
@@ -47,6 +48,7 @@ test_that("ci", {
   expect_equal(ci(m3)$CI_low, -2.60478, tolerance = 1e-3)
 })
 
+
 test_that("se", {
   expect_equal(
     standard_error(m1)$SE,
@@ -61,6 +63,7 @@ test_that("se", {
   expect_equal(standard_error(m3)$SE, 0.5166726, tolerance = 1e-3)
 })
 
+
 test_that("p_value", {
   expect_equal(
     p_value(m1)$p,
@@ -75,6 +78,7 @@ test_that("p_value", {
   expect_equal(p_value(m3)$p, 0.53696, tolerance = 1e-3)
 })
 
+
 test_that("model_parameters", {
   expect_equal(
     model_parameters(m1)$Coefficient,
@@ -88,3 +92,22 @@ test_that("model_parameters", {
   )
   expect_equal(model_parameters(m3)$Coefficient, -0.381721, tolerance = 1e-3)
 })
+
+
+test_that("vcov standard errors", {
+  skip_if_not_installed("sandwich")
+  data("Grunfeld", package = "plm")
+  ran <- suppressWarnings(
+    plm::plm(value ~ capital + inv, data = Grunfeld, model = "random", effect = "twoways")
+  )
+  out1 <- standard_error(ran)
+  out2 <- standard_error(ran, vcov = "HC1")
+  validate1 <- coef(summary(ran))[, 2]
+  validate2 <- sqrt(diag(sandwich::vcovHC(ran, type = "HC1")))
+
+  expect_equal(out1$SE, validate1, tolerance = 1e-3, ignore_attr = TRUE)
+  expect_equal(out2$SE, validate2, tolerance = 1e-3, ignore_attr = TRUE)
+
+  expect_snapshot(print(model_parameters(ran)))
+  expect_snapshot(print(model_parameters(ran, vcov = "HC1")))
+})