Drop label data

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: tidyREDCap
 Title: Helper Functions for Working with 'REDCap' Data
-Version: 1.2.0
+Version: 2.0.0
 Authors@R: 
     c(person(
         given = "Raymond",
@@ -72,6 +72,7 @@ Imports:
     dplyr,
     janitor,
     labelVector,
+    lifecycle,
     magrittr,
     purrr,
     REDCapR,

NAMESPACE

@@ -47,6 +47,7 @@ importFrom(janitor,adorn_pct_formatting)
 importFrom(janitor,tabyl)
 importFrom(labelVector,is_labelled)
 importFrom(labelVector,set_label)
+importFrom(lifecycle,deprecate_soft)
 importFrom(magrittr,"%>%")
 importFrom(purrr,map_chr)
 importFrom(purrr,map_df)

NEWS.md

@@ -1,3 +1,17 @@
+# tidyREDCap 2.0.0 (CRAN release)
+
+## Breaking changes
+
+* `drop_label()` now preserves variable class attributes (e.g., `"character"`, `"numeric"`) while removing only the `"labelled"` class and `label` attribute. Previous versions removed all attributes.
+
+## New features
+
+* `drop_label()` now can be used on data frames. When called without arguments on a data frame, it removes labels from all variables.
+
+## Fixes/Changes
+
+* `drop_labels()` is now deprecated in favor of `drop_label()`. Use `drop_label()` without arguments to remove labels from all variables.
+
 # tidyREDCap 1.2.0 (CRAN release)
 
 * Updated `drop_label()` now accepts multiple variables and uses `tidyselect` helpers.

R/drop_label.R

@@ -0,0 +1,84 @@
+
+#' Drop attributes/labels from variables or data frames
+#' 
+#' @description Some functions don't work with labelled variables. As a solution, 
+#' this function can be used to drop labels (and all other attributes) from 
+#' one or more variables within a data frame, from all variables if none are 
+#' specified, or from a vector directly.
+#'
+#' @param x A data frame or a vector/column.
+#' @param ... When `x` is a data frame, select variables using tidyselect 
+#'   helpers (e.g., `contains()`, `starts_with()`) or unquoted names. 
+#'   If empty, removes labels from ALL variables. Ignored when `x` is a vector.
+#'
+#' @return The modified data frame or vector with attributes removed.
+#'
+#' @examples
+#' \dontrun{
+#' # Dataset-level: Remove labels from ALL variables
+#' df |> drop_label()
+#'
+#' # Dataset-level: Remove labels from specific variables
+#' df |> drop_label(employment, starts_with("dem_"))
+#'
+#' # Variable-level: Use inside mutate
+#' df |> mutate(name_first = drop_label(name_first))
+#' 
+#' # Variable-level: Use with across()
+#' df |> mutate(across(c(age, income), drop_label))
+#' }
+#'
+#' @export
+drop_label <- function(x, ...) {
+  # 1. Dataset-level logic
+  if (is.data.frame(x)) {
+    old_class_char <- class(x)
+
+    # Check if any variables were selected
+    vars_idx <- tidyselect::eval_select(rlang::expr(c(...)), x)
+
+    # If no variables selected, process ALL columns
+    if (length(vars_idx) == 0) {
+      # seq_along will either capture the variable column number
+      vars_idx <- seq_along(x)
+    }
+
+    # Remove attributes from selected columns
+    for (col_idx in vars_idx) {
+      # col_idx evaluates to:
+      # 1. variable name(s) if passed with the `...` argument
+      # 2. variable column position number if the whole dataset was passed for dropping
+
+      # Get all current attributes
+      col_attrs <- attributes(x[[col_idx]])
+
+      # Remove the label attribute
+      col_attrs$label <- NULL
+
+      # Remove the "labelled" from class if it exists
+      if (!is.null(col_attrs$class)) {
+        col_attrs$class <- col_attrs$class[col_attrs$class != "labelled"]
+        # If the class now becomes empty, remove it to use R's default
+        if (length(col_attrs$class) == 0) col_attrs$class <- NULL
+      }
+
+      # Reapply the modified attributes
+      attributes(x[[col_idx]]) <- col_attrs
+    }
+
+    # Preserve original class structure (data.frame vs tibble)
+    if (("data.frame" %in% old_class_char) && !("tbl_df" %in% old_class_char)) {
+      class(x) <- "data.frame"
+    }
+
+    return(x)
+  }
+
+  # 2. Variable-level logic
+  if (is.character(substitute(x))) {
+    stop('It looks like you quoted your variable. The variable must be unquoted when used inside mutate().')
+  }
+
+  attributes(x) <- NULL
+  return(x)
+}

R/drop_labels.R

@@ -5,6 +5,7 @@
 #' @param df The data frame with column labels that you want to drop
 #' 
 #' @importFrom purrr map_df
+#' @importFrom lifecycle deprecate_soft
 #'
 #' @export
 #' 
@@ -17,6 +18,13 @@
 #'   skimr::skim()
 #' }
 drop_labels <- function(df) {
+  # Add deprecate message:
+  lifecycle::deprecate_soft(
+    when = "1.2.0",  # Version when this was deprecated
+    what = "drop_labels()",
+    with = "drop_label()"
+  )
+
   old_class_char <- class(df)
   if (!("data.frame" %in% old_class_char)) {
     stop("df must have class data.frame", call. = FALSE)

R/make_instrument_auto.R

@@ -171,51 +171,3 @@ fix_class_bug <- function(df) {
   return(df)
 }
 "fix_class_bug"
-
-
-#' Drop attributes/labels from variables or data frames
-#' 
-#' @description Some functions don't work with labelled variables. As a solution, 
-#' this function can be used to drop labels (and all other attributes) from 
-#' one or more variables within a data frame, or from a vector directly.
-#'
-#' @param x A data frame or a vector/column.
-#' @param ... When `x` is a data frame, select variables using tidyselect 
-#'   helpers (e.g., `contains()`, `starts_with()`) or unquoted names. 
-#'   Ignored when `x` is a vector.
-#'
-#' @return The modified data frame or vector with attributes removed.
-#'
-#' @examples
-#' \dontrun{
-#' # Dataset-level: Remove labels from specific variables
-#' df |> drop_label(employment, starts_with("dem_"))
-#'
-#' # Variable-level: Use inside mutate
-#' df |> mutate(name_first = drop_label(name_first))
-#' 
-#' # Variable-level: Use with across()
-#' df |> mutate(across(c(age, income), drop_label))
-#' }
-#'
-#' @export
-drop_label <- function(x, ...) {
-  # 1. Dataset-level logic
-  if (is.data.frame(x)) {
-    vars_idx <- tidyselect::eval_select(rlang::expr(c(...)), x)
-    if (length(vars_idx) == 0) return(x)
-
-    for (col_idx in vars_idx) {
-      attributes(x[[col_idx]]) <- NULL
-    }
-    return(x)
-  }
-
-  # 2. Variable-level logic
-  if (is.character(substitute(x))) {
-    stop('It looks like you quoted your variable. The variable must be unquoted when used inside mutate().')
-  }
-
-  attributes(x) <- NULL
-  return(x)
-}

README.md

@@ -25,7 +25,7 @@ tidyREDCap is an R package with functions for processing REDCap data.
 
 * After loading data into R using RStudio with the `import_instruments()` function, you can see both the variable name and the text that appears to users of REDCap.  All you need to do is click on the dataset's name in the **Environment** tab or use the `View()` function. The column headings will include both the Variable Name and the Field Label from REDCap. 
 
-* 💥 NEW in Version 1.1 💥  Functions coming from packages outside of `tidyREDCap` may not understand what to do with labeled variables.  So, `tidyREDCap` includes a new `drop_labels()` function that will allow you to strip the labels before using functions that want unlabeled data.
+* 💥 NEW in Version 2.0.0 💥  Functions coming from packages outside of `tidyREDCap` may not understand what to do with labeled variables.  So, `tidyREDCap` includes a new `drop_label()` function that will allow you to strip the labels before using functions that want unlabeled data.
 
 #### Working with <i>Choose One</i> Questions