Removed here from dependencies, using where() from tidyselect and not…

… dplyr

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: moiraine
 Title: Construction of Reproducible Pipelines for Testing and Comparing Multi-omics Integration Tools
-Version: 1.0.0
+Version: 1.0.0.9000
 Authors@R: 
     c(person("Olivia", "Angelin-Bonnet", 
     email = "[email protected]",
@@ -30,7 +30,6 @@ Imports:
     ggpubr,
     ggrepel,
     grid,
-    here,
     methods,
     mixOmics,
     MOFA2,

NAMESPACE

@@ -177,5 +177,4 @@ importClassesFrom(Biobase,SnpSet)
 importClassesFrom(Biobase,eSet)
 importClassesFrom(MultiDataSet,MultiDataSet)
 importFrom(dplyr,sym)
-importFrom(dplyr,where)
 importFrom(ggplot2,aes)

NEWS.md

@@ -1,3 +1,7 @@
 # moiraine (development version)
 
 - In `comparison_heatmap_corr()`, the user can now choose the legend position through the `legend_position` parameter.
+
+- Package here has been removed from dependencies (not needed).
+
+- `where()` function now imported from tidyselect instead of dplyr (as it required a newer version of dplyr).

R/data_import.R

@@ -180,7 +180,10 @@ import_fmetadata_gff <- function(file, feature_type, add_fields = NULL) {
       ## become vector
       tibble::as_tibble() |>
       dplyr::mutate(
-        dplyr::across(where(is.list), ~ sapply(.x, paste, collapse = " "))
+        dplyr::across(
+          tidyselect::where(is.list),
+          ~ sapply(.x, paste, collapse = " ")
+        )
       )
 
     .check_names(

R/diablo.R

@@ -595,7 +595,7 @@ diablo_table_optim_keepX <- function(tune_res) {
       )
   }) |>
     purrr::reduce(dplyr::bind_rows) |>
-    dplyr::mutate(Total = rowSums(dplyr::across(where(is.numeric))))
+    dplyr::mutate(Total = rowSums(dplyr::across(tidyselect::where(is.numeric))))
 
   return(res)
 }

R/import_from_packages.R

@@ -7,7 +7,7 @@ NULL
 #' @import methods
 NULL
 
-#' @importFrom dplyr sym where
+#' @importFrom dplyr sym
 NULL
 
 #' @importFrom ggplot2 aes

R/multidataset_utils.R

@@ -238,7 +238,7 @@ join_features_metadata <- function(df, mo_data) {
     dplyr::filter(feature_id %in% df$feature_id) |>
     dplyr::select(
       feature_id,
-      where(~ any(!is.na(.x)))
+      tidyselect::where(~ any(!is.na(.x)))
     )
 
   if (nrow(fmeta) == 0) {
@@ -285,7 +285,7 @@ join_samples_metadata <- function(df, mo_data, datasets = NULL) {
     dplyr::filter(id %in% df$id) |>
     dplyr::select(
       id,
-      where(~ any(!is.na(.x)))
+      tidyselect::where(~ any(!is.na(.x)))
     )
 
   if (nrow(smeta) == 0) {

R/output_plot.R

@@ -1122,7 +1122,7 @@ plot_features_weight_covariate <- function(method_output,
         ) |>
         dplyr::mutate(
           dplyr::across(
-            .cols = dplyr::where(\(x){!is.numeric(x)}),
+            .cols = tidyselect::where(\(x){!is.numeric(x)}),
             .fns = as.character
           )
         ),

R/so2pls.R

@@ -666,7 +666,7 @@ so2pls_print_cv_sparsity <- function(cv_res_optim) {
       names_from = joint_component,
       values_from = n_features
     ) |>
-    dplyr::mutate(Total = rowSums(dplyr::across(where(is.numeric))))
+    dplyr::mutate(Total = rowSums(dplyr::across(tidyselect::where(is.numeric))))
 }
 
 #' Plot sparsity cross-validation results for sO2PLS