Summarise works

R/Fic_calculations.R

@@ -231,17 +231,102 @@ ras.Fu_feces_mass_balance_10.2.5 <- function(df_calc,
   return(df_calc)
 }
 
-ras.Fu_feces_summary <- function(df,
+#' Fu feces mean & SD
+#'
+#' Add mean & SD columns, grouped by `grouping_col`, for all
+#' the calculated numeric columns.
+#'
+#' @param df Data frame
+#' @param grouping_col Column name to group by
+#' @param sample_type Column name for sample type
+#' @param compound Column name for compound
+#'
+#' @return Data frame with mean & SD columns
+#' @export
+#'
+#' @examples
+#'   # No example
+ras.Fu_feces_meanSD <- function(df,
                                  grouping_col = "Sample_ID",
-                                 .summarize = TRUE,
-                                 .SD = TRUE) {
-  if (.SD) {
+                                 sample_type = "Sample Type",
+                                 compound = "Analyte Peak Name") {
+  df_sum <- df %>%
+    dplyr::select(
+      -tidyselect::any_of(c(
+        "Date",
+        "Initials",
+        "Sample_origin",
+        "Dosing",
+        "Timepoint",
+        "Replicate",
+        {{sample_type}},
+        {{compound}}
+        )
+      )
+    )
 
-  }
+  num_columns <- df_sum %>%
+    dplyr::select(tidyselect::where(is.numeric)) %>%
+    colnames()
 
-  if (.summarize) {
+  for (col in num_columns) {
+    col_mean <- paste0(col,"_mean")
+    col_sd <- paste0(col,"_sd")
 
+    df <- df %>%
+      dplyr::mutate(
+        {{col_mean}} := mean(.data[[col]], na.rm = T),
+        {{col_sd}} := sd(.data[[col]], na.rm = T),
+        .after = {{col}},
+        .by = {{grouping_col}}
+      )
   }
-
   return(df)
 }
+
+ras.Fu_feces_summarize <- function(df,
+                                   grouping_col = "Sample_ID",
+                                   sample_type = "Sample Type",
+                                   compound = "Analyte Peak Name") {
+  df_sub <- df %>%
+    dplyr::select(
+      -tidyselect::any_of(c(
+        "Date",
+        "Initials",
+        "Sample_origin",
+        "Dosing",
+        "Timepoint",
+        "Replicate",
+        {{sample_type}},
+        {{compound}}
+      )
+      )
+    )
+
+  num_columns <- df_sub %>%
+    dplyr::select(tidyselect::where(is.numeric)) %>%
+    colnames()
+
+  df_sum <- df %>%
+    dplyr::select({{grouping_col}}) %>%
+    unique()
+
+  for (col in num_columns) {
+    col_mean <- paste0(col,"_mean")
+    col_sd <- paste0(col,"_sd")
+
+    df_temp <- df %>%
+      dplyr::summarise(
+        {{col_mean}} := mean(.data[[col]], na.rm = T),
+        {{col_sd}} := sd(.data[[col]], na.rm = T),
+        .by = {{grouping_col}}
+      )
+
+    df_sum <- dplyr::right_join(
+      x = df_sum,
+      y = df_temp,
+      by = {{grouping_col}}
+    )
+  }
+  return(df_sum)
+}

R/Fu_feces_workflow.R

@@ -16,8 +16,8 @@
 #' `NULL` => "Analyte Peak Name"
 #' @param .checkValues Removes "<" from the values column to make sure it can
 #' be coerced to numeric.
-#' @param .summarize Summarize values into averages
-#' @param .SD Create column with Standard Deviation
+#' @param .summarize Summarize the data into unique rows in
+#' grouping column. Drops the original value columns.
 #'
 #' @return Dataframe with all variables, used & calculated.
 #' Also writes data frame to .csv (EU) & .xlsx files.
@@ -42,8 +42,7 @@ ras.Fu_feces_workflow <- function(
     MassBalance_2.5 = 1.75,
     .compound = "Analyte Peak Name",
     .checkValues = TRUE,
-    .summarize = TRUE,
-    .SD = TRUE
+    .summarize = FALSE
 ) {
   # Set output directory
   output_dir <- tcltk::tk_choose.dir(caption = "Select output directory")
@@ -106,12 +105,21 @@ ras.Fu_feces_workflow <- function(
       mass_factor = MassBalance_2.5
     )
 
-  df_calc <- df_calc %>%
-    ras.Fu_feces_summary(
-      grouping_col = "Sample_ID",
-      .summarize = .summarize,
-      .SD = .SD
-    )
+  if (.summarize) {
+    df_calc <- df_calc %>%
+      ras.Fu_feces_summarize(
+        grouping_col = "Sample_ID",
+        sample_type = Sample_type,
+        compound = .compound
+      )
+  } else {
+    df_calc <- df_calc %>%
+      ras.Fu_feces_meanSD(
+        grouping_col = "Sample_ID",
+        sample_type = Sample_type,
+        compound = .compound
+      )
+  }
 
 
   # Write df_calc to file ----