Replace sapply with vapply and add drop=F in [].

R/gs_score-methods.R

@@ -30,8 +30,11 @@ setMethod(
     )
 
     ## compute overall score
-    score <- sapply(names(features), \(i)
-    gs_score(data = data, features = features[[i]])) |>
+    score <- vapply(
+      names(features), \(i)
+      gs_score(data = data, features = features[[i]]),
+      rep(1, ncol(data))
+    ) |>
       data.frame()
     ## set colnames
     colnames(score) <- paste(names(features), suffix, sep = ".")

R/plot.R

@@ -91,7 +91,7 @@ sin_score_boxplot <- function(data, features = NULL,
                               ref.group, label,
                               method = "t.test") {
   if (is.null(features)) features <- rownames(data)
-  data[features, ] |>
+  data[features, , drop = FALSE] |>
     as.matrix() |>
     as.data.frame() |>
     dplyr::add_rownames("Gene") |>
@@ -159,14 +159,6 @@ score_barplot <- function(top_markers, column = ".dot", f_list, n = 30) {
     f_list <- list(Features = top_markers$Genes)
   }
 
-  # ## get top n markers
-  # top_markers <- top_markers(
-  #   data[features, ],
-  #   label = label,
-  #   n = n,
-  #   ...
-  # )
-
   ## extract top n markers
   top_markers <- dplyr::slice_max(top_markers, Scores, n = n)
 

R/scale_mgm.R

@@ -17,8 +17,13 @@ scale_mgm <- function(expr, label) {
   # colnames(sds) <- unique(label)
 
   ## compute group means
-  mgm <- sapply(unique(label), \(i)
-  sparseMatrixStats::rowMeans2(expr[, label == i], na.rm = TRUE)) |> # get mean of each group
+  mgm <- vapply(
+    unique(label), \(i)
+    sparseMatrixStats::rowMeans2(expr[, label == i, drop = FALSE],
+      na.rm = TRUE
+    ),
+    rep(1, nrow(expr))
+  ) |> # get mean of each group
     rowMeans(na.rm = TRUE) # get mean of group mean
 
   ## scale

R/score.R

@@ -26,7 +26,7 @@ gs_score_init <- function(score, features = NULL) {
   stopifnot("less than 2 features are in score rows!" = length(features) > 1)
 
   ## calculate mean score of features
-  m_score <- colMeans(score[features, ], na.rm = TRUE)
+  m_score <- colMeans(score[features, , drop = FALSE], na.rm = TRUE)
   return(m_score)
 }
 

R/top_markers.R

@@ -151,8 +151,11 @@ top_markers_glm <- function(data, label, n = 10,
   # betas <- apply(betas, 2, \(x) x - (sum(x) - x)/(length(x) - 1))
 
   ## compute logFC (1 vs max excluding self) for each group
-  betas <- sapply(seq_len(nrow(betas)), \(i)
-  betas[i, ] - sparseMatrixStats::colMaxs(matrix(betas[-i, ], ncol = ncol(betas)))) |>
+  betas <- vapply(
+    seq_len(nrow(betas)), \(i)
+    betas[i, ] - sparseMatrixStats::colMaxs(betas[-i, , drop = FALSE]),
+    rep(1, ncol(betas))
+  ) |>
     t()
   rownames(betas) <- levels(label)