total counts as log10

bigomics · Nov 2, 2023 · 3b2ba5b · 3b2ba5b
1 parent f1f0dee
commit 3b2ba5b
Show file tree

Hide file tree

Showing 3 changed files with 19 additions and 33 deletions.
diff --git a/components/board.dataview/R/dataview_plot_totalcounts.R b/components/board.dataview/R/dataview_plot_totalcounts.R
@@ -28,25 +28,25 @@ dataview_plot_totalcounts_ui <- function(
 }
 
 dataview_plot_totalcounts_server <- function(id,
-                                             getCountsTable,
+                                             getCountStatistics,
                                              r.data_groupby = reactive(""),
                                              watermark = FALSE) {
   moduleServer(id, function(input, output, session) {
     plot_data <- shiny::reactive({
       data_groupby <- r.data_groupby()
 
-      tbl <- getCountsTable()
+      tbl <- getCountStatistics()
       req(tbl)
 
-      ylab <- "Total counts"
+      ylab <- "total counts (log10)"
       if (data_groupby != "<ungrouped>") {
-        ylab <- "Mean total counts"
+        ylab <- "average total counts (log10)"
       }
 
       res <- list(
         df = data.frame(
           sample = names(tbl$total.counts),
-          counts = tbl$total.counts
+          counts = log10(tbl$total.counts)
         ),
         ylab = ylab
       )
@@ -103,20 +103,23 @@ dataview_plot_totalcounts_server <- function(id,
             "<extra></extra>"
           )
         ) %>%
+        plotly_default() %>%
         plotly::layout(
           xaxis = list(title = FALSE),
           yaxis = list(title = res$ylab),
-          font = list(family = "Lato"),
-          margin = list(l = 10, r = 10, b = 10, t = 10)
-        ) %>%
-        plotly_default()
+          margin = list(l = 30, r = 0, t = 0, b = 0)
+        ) 
 
       fig
     }
 
     modal_plotly.RENDER <- function() {
       plotly.RENDER() %>%
-        plotly_modal_default()
+        plotly_modal_default() %>%
+        plotly::layout(
+          margin = list(l = 35, r = 0, t = 0, b = 0)
+        ) 
+
     }
 
     PlotModuleServer(

diff --git a/components/board.dataview/R/dataview_server.R b/components/board.dataview/R/dataview_server.R
@@ -279,7 +279,7 @@ DataViewBoard <- function(id, pgx) {
 
     getCountStatistics <- reactiveVal()
 
-    observeEvent(c(input$data_groupby, input$data_samplefilter),
+    observeEvent( c(input$data_groupby, input$data_samplefilter, input$data_type),
       {
         shiny::req(pgx$X, pgx$Y, pgx$samples)
         shiny::validate(shiny::need("counts" %in% names(pgx), "no 'counts' in object."))
@@ -288,35 +288,19 @@ DataViewBoard <- function(id, pgx) {
         samples <- colnames(pgx$X)
         samples <- playbase::selectSamplesFromSelectedLevels(pgx$Y, input$data_samplefilter)
         nsamples <- length(samples)
-        if ("counts" %in% names(pgx)) {
+        if (input$data_type == "counts") {
           counts <- pgx$counts[, samples, drop = FALSE]
         } else {
-          cat("WARNING:: no counts table. estimating from X\n")
           counts <- pmax(2**pgx$X - 1, 0)
-          k <- grep("lib.size", colnames(pgx$samples))[1]
-          if (length(k) > 0) {
-            libsize <- pgx$samples[colnames(counts), k]
-            libsize
-            counts <- t(t(counts) * libsize)
-          }
-        }
-        if (sum(is.na(counts)) > 0) {
-          cat("WARNING:: plot counts: counts has missing values!\n")
         }
 
         grpvar <- input$data_groupby
         gr <- pgx$Y[samples, grpvar]
         grps <- sort(unique(gr))
         if (input$data_groupby != "<ungrouped>" && length(grps) > 1) {
-          newx <- c()
-          for (g in grps) {
-            mx <- rowMeans(counts[, which(gr == g), drop = FALSE], na.rm = TRUE)
-            newx <- cbind(newx, mx)
-          }
-          if (NCOL(newx) == 1) newx <- matrix(newx, ncol = 1)
-          rownames(newx) <- rownames(counts)
-          colnames(newx) <- grps
-          counts <- newx
+          mx <- tapply( 1:ncol(counts), gr,
+            function(ii) rowMeans(counts[,ii,drop=FALSE],na.rm=TRUE))
+          counts <- do.call(cbind, mx)
         }
 
         ## if too many samples (like scRNA-seq do subsampling...)
@@ -378,7 +362,6 @@ DataViewBoard <- function(id, pgx) {
         colnames(log2counts) <- substring(colnames(log2counts), 1, 30)
         colnames(prop.counts) <- substring(colnames(prop.counts), 1, 30)
 
-
         res <- list(
           total.counts = total.counts,
           subtt = subtt,

diff --git a/components/board.dataview/R/dataview_ui.R b/components/board.dataview/R/dataview_ui.R
@@ -168,7 +168,7 @@ DataViewUI <- function(id) {
           ),
           dataview_plot_boxplot_ui(
             ns("counts_boxplot"),
-            title = "Median counts distribution",
+            title = "Counts boxplots",
             info.text = "The samples (or cells) can be grouped/ungrouped in the grouped setting under the main Options.",
             caption = "Distribution of total counts per sample/group. The center horizontal bar correspond to the median.",
             height = c("100%", TABLE_HEIGHT_MODAL),