Add align_dt to align pow and add ggplotly to vignette

DESCRIPTION

@@ -107,6 +107,7 @@ Suggests:
     knitr,
     labeling,
     plotly (>= 4.10.2),
+    pow,
     rmarkdown,
     scales,
     testthat (>= 3.0.0),
@@ -117,4 +118,3 @@ RoxygenNote: 7.2.3
 VignetteBuilder: knitr
 Config/testthat/edition: 3
 Roxygen: list(markdown = TRUE)
-

R/align_pow-GCIMSDataset.R

@@ -2,6 +2,10 @@
 #'
 #' Uses the POW algorithm to generate waprs and align the samples
 #'
+#' To use this function you need to install the "pow" package hosted on github, to do this run he next line:
+#' remotes::install_github("sipss/pow"),
+#' also you will need the "pracma" package hosted in CRAN
+#'
 #' @param object A [GCIMSDataset] object, modified in-place
 #' @param lambdas vector of lambdas to be tested for the POW algorithm
 #' @param p points to be taken for validation for the POW algorithm
@@ -26,7 +30,7 @@ methods::setMethod(
     iv <- seq(2, m - 1, by = p)
     v[iv] <- 0
     W <- Matrix::Diagonal(x = v)
-    val<-pow::compute_val_error(X,y,W,iv,lambdas)
+    val<-pow::compute_val_error(X, y, W, iv, lambdas)
     opar <- pow::optimize_params(n_samples, lambdas, val$ti_ix, val$e_ix)
     best_lambdas2 <- opar$best_params
 
@@ -35,9 +39,16 @@ methods::setMethod(
     w<-append(w,list(1:m),idx_y-1)
     names(w)<-object$sampleNames
 
+    tis_matrix <- getTIS(object)
+    dt <- dtime(object)
+    rip_position <- apply(tis_matrix, 1L, which.max)
+    rip_ref_idx <- round(stats::median(rip_position, na.rm = TRUE))
+    rip_ref_ms <- dt[rip_ref_idx]
+
     delayed_op <- DelayedOperation(
       name = "align_pow",
       fun = align_pow,
+      params = list(rip_ref_ms = rip_ref_ms),
       params_iter = list(w=w)
     )
     object$appendDelayedOp(delayed_op)

R/align_pow-GCIMSSample.R

@@ -10,7 +10,13 @@
 methods::setMethod(
   "align_pow",
   "GCIMSSample",
-  function(object,w){
+  function(object, rip_ref_ms, w){
+    object <- alignDt(object, rip_ref_ms = rip_ref_ms)
+    dt_range <- c(
+      object@proc_params$align$dt_min_ms,
+      object@proc_params$align$dt_max_ms
+    )
+    object <- subset(object, dt_range = dt_range)
     int <- GCIMS::intensity(object)
     inter <- t(apply(int, 1, pow::interpolation, w = w, return = FALSE))
     sel <- !is.na(inter[1,])

vignettes/.gitignore

@@ -1,7 +1,5 @@
 *.html
 *.R
-Rplots.pdf
 *.log
-GCIMSDataset_demo1_
 sample1.csv
 sample2.csv

vignettes/introduction-to-gcims.Rmd

@@ -192,14 +192,9 @@ to_plot <- dplyr::bind_rows(
   Smoothed = as.data.frame(one_ims_smoothed),
   .id = "Status"
 )
-ggplot(to_plot) +
-  geom_line(aes(x = drift_time_ms, y = intensity, colour = Status)) +
-  coord_cartesian(xlim = c(7, 10)) +
-  labs(x = "Drift time (ms)", y = "Intensity (a.u.)")
-ggplot(to_plot) +
+plotly::ggplotly(ggplot(to_plot) +
   geom_line(aes(x = drift_time_ms, y = intensity, colour = Status)) +
-  coord_cartesian(xlim = c(7, 10), ylim = c(0, 300)) +
-  labs(x = "Drift time (ms)", y = "Intensity (a.u.)")
+  labs(x = "Drift time (ms)", y = "Intensity (a.u.)"))
 ```
 
 ```{r}
@@ -213,13 +208,9 @@ to_plot <- dplyr::bind_rows(
   Smoothed = as.data.frame(one_chrom_smoothed),
   .id = "Status"
 )
-ggplot(to_plot) +
-  geom_line(aes(x = retention_time_s, y = intensity, colour = Status)) +
-  labs(x = "Retention time (s)", y = "Intensity (a.u.)")
-ggplot(to_plot) +
+plotly::ggplotly(ggplot(to_plot) +
   geom_line(aes(x = retention_time_s, y = intensity, colour = Status)) +
-  coord_cartesian(xlim = c(200, 250)) +
-  labs(x = "Retention time (s)", y = "Intensity (a.u.)")
+  labs(x = "Retention time (s)", y = "Intensity (a.u.)"))
 ```
 
 You can also apply it to a single sample:
@@ -392,15 +383,15 @@ ket1 <- findPeaks(
   rt_peakDetectionCWTParams = list(exclude0scaleAmpThresh = TRUE),
   dt_extension_factor = 0,
   rt_extension_factor = 0,
-  exclude_rip = FALSE,
+  exclude_rip = TRUE,
   iou_overlap_threshold = 0.2
 )
 ```
 
 ```{r}
 peak_list_ket1 <- peaks(ket1)
-plot(ket1) +
-  overlay_peaklist(peak_list_ket1, color_by = "PeakID")
+plotly::ggplotly(plot(ket1) +
+  overlay_peaklist(peak_list_ket1, color_by = "PeakID"))
 ```
 
 Then do it on the whole dataset:
@@ -429,20 +420,16 @@ ket2 <- dataset$getSample("Ketones2")
 ```
 
 ```{r}
-plot(ket2) +
-  overlay_peaklist(peaks(ket2)) +
-  lims(x = c(7.5,15), y = c(0, 500))
+plotly::ggplotly(plot(ket2) +
+  overlay_peaklist(peaks(ket2)) )
 ```
 
 Or plot all the peaks from all the dataset together, overlayed on a single sample:
 
 ```{r}
 plt <- plot(ket2) +
   overlay_peaklist(peaks(dataset), color_by = "SampleID")
-# We could explore the plot interactively: 
-# plot_interactive(plt)
-# Or use a static version to save memory and render it faster:
-plt
+plotly::ggplotly(plt)
 ```
 
 # Clustering
@@ -464,18 +451,16 @@ peak_list_clustered <- peak_clustering$peak_list_clustered
 ```
 
 
-```{r fig.height=10, fig.width=10}
+```{r }
 plt <- plot(ket2) +
   overlay_peaklist(peak_list_clustered, color_by = "SampleID")
-print(plt)
-print(plt + lims(x = c(6, 15), y = c(0, 400)))
+plotly::ggplotly(plt)
 ```
 
 ```{r}
 plt <- plot(ket2) +
   overlay_peaklist(dplyr::filter(peak_list_clustered, !is.na(cluster)), color_by = "cluster")
-print(plt)
-print(plt + lims(x = c(6, 15), y = c(0, 400)))
+plotly::ggplotly(plt)
 ```
 
 The resulting cluster sizes (median position of individual clusters) is not a good reference
@@ -484,8 +469,7 @@ for integration. We are working on this.
 ```{r}
 plt <- plot(ket2) +
   overlay_peaklist(peak_clustering$cluster_stats, color_by = "cluster")
-#plot_interactive(plt)
-plt + lims(x = c(6, 15), y = c(0, 400))
+plotly::ggplotly(plt)
 ```
 
 # Baseline correction
@@ -510,21 +494,12 @@ to_plot <- data.frame(
   Intensity = intensity(one_ims_spec),
   Baseline = baseline(one_ims_spec)
 )
-ggplot(to_plot) +
+plotly::ggplotly(ggplot(to_plot) +
   geom_line(aes(x = drift_time_ms, y = Intensity), color = "red") +
   geom_line(aes(x = drift_time_ms, y = Baseline), color = "blue") +
-  labs(x = "Drift time (ms)", y = "Intensity (a.u.)")
+  labs(x = "Drift time (ms)", y = "Intensity (a.u.)"))
 ```
 
-```{r }
-ggplot(to_plot) +
-  geom_line(aes(x = drift_time_ms, y = Intensity), color = "red") +
-  geom_line(aes(x = drift_time_ms, y = Baseline), color = "blue") +
-  coord_cartesian(xlim = c(7, 10), ylim = c(0, 500)) +
-  labs(x = "Drift time (ms)", y = "Intensity (a.u.)")
-```
-
-
 ```{r}
 one_chrom <- getChromatogram(ket1, dt_range = 10.4)
 ```
@@ -564,7 +539,7 @@ ket1_basline <- estimateBaseline(
 
 ```{r}
 cowplot::plot_grid(
-  plot(ket1_no_basline, rt_range = c(0, 500), dt_range = c(6, 16)),
+  plot(x = ket1_no_basline, rt_range = c(0, 500), dt_range = c(6, 16)),
   plot(x = ket1_basline, rt_range = c(0, 500), dt_range = c(6, 16)),
   ncol = 2
 )
@@ -586,7 +561,7 @@ dataset$realize()
 
 Some clusters:
 
-```{r fig.height=10, fig.width=10}
+```{r}
 plot(ket2) + 
   overlay_peaklist(
     peak_clustering$peak_list_clustered,
@@ -626,14 +601,16 @@ peak_list <- peaks(dataset)
 
 ```{r}
 peak_table <- peakTable(peak_list, aggregate_conflicting_peaks = max)
-tail(peak_table$peak_table_matrix)
+t(tail(t(peak_table$peak_table_matrix)))
 ```
 We showed the last 6 clusters where we have "NA" Values
 
 # Imputation
 ```{r}
-peak_table_imputed <- imputePeakTable(peak_table$peak_table_matrix, dataset, peak_clustering$cluster_stats)
-tail(peak_table_imputed)
+peak_table_imputed <- imputePeakTable(peak_table$peak_table_matrix,
+                                      dataset, 
+                                      peak_clustering$cluster_stats)
+t(tail(t(peak_table_imputed)))
 ```
 As we can see the "NA" values were imputed