720 Remove replace_emptys_with_na (#768)

Closes #720

Co-authored-by: github-actions <41898282+github-actions[bot]@users.noreply.github.com>

R/df_explicit_na.R

@@ -45,7 +45,6 @@
 #' # Encode missing values in a subset of columns.
 #' df_explicit_na(my_data, omit_columns = c("x", "y"))
 #'
-#' @seealso [replace_emptys_with_na()]
 df_explicit_na <- function(data,
                            omit_columns = NULL,
                            char_as_factor = TRUE,
@@ -86,7 +85,6 @@ df_explicit_na <- function(data,
     }
 
     if (is.factor(xi) || is.character(xi)) {
-
       # Handle empty strings and NA values.
       xi <- explicit_na(sas_na(xi), label = na_level)
 

R/logistic_regression.R

@@ -904,11 +904,10 @@ logistic_summary_by_flag <- function(flag_var) {
 #' @param drop_and_remove_str string to be dropped and removed
 #'
 #' @examples
-#' # Internal function - replace_emptys_with_na
 #' \dontrun{
 #' # flagging empty strings with "_"
-#' df <- replace_emptys_with_na(df, rep_str = "_")
-#' df2 <- replace_emptys_with_na(df2, rep_str = "_")
+#' df <- df_explicit_na(df, na_level = "_")
+#' df2 <- df_explicit_na(df2, na_level = "_")
 #'
 #' result1 <- basic_table() %>%
 #'   summarize_logistic(
@@ -931,7 +930,6 @@ logistic_summary_by_flag <- function(flag_var) {
 summarize_logistic <- function(lyt,
                                conf_level,
                                drop_and_remove_str = "") {
-
   # checks
   checkmate::assert_string(drop_and_remove_str)
 

R/utils_factor.R

@@ -285,42 +285,3 @@ fct_collapse_only <- function(.f, ..., .na_level = "<Missing>") {
   x <- forcats::fct_collapse(.f, ..., other_level = .na_level)
   do.call(forcats::fct_relevel, args = c(list(.f = x), as.list(new_lvls)))
 }
-
-#' Replace all empty string values in data frame
-#'
-#' @description
-#' Function used to fix for update in `Rtables issue` `tern#593` (`NA` alternative -> `" "`)
-#' This function is similar in scope as `df_explicit_na()`. In the future a merge
-#' is to be expected.
-#'
-#' @param df `data.frame` table to act upon
-#' @param rep_str replacement string for empty strings
-#'
-#' @details this functions relies onto `for` loop, `levels`, and `nchar`. It can
-#' easily be optimized for different cases. For the moment, this fits the purpose
-#' to fix issue `tern#593`
-#'
-#' @seealso [df_explicit_na()]
-#'
-#' @keywords internal
-replace_emptys_with_na <- function(df, rep_str = "NA") {
-
-  # checks
-  checkmate::assert_string(rep_str)
-
-  # col logical v
-  where_to_mod <- apply(df, 2, function(x) any(nchar(x) == 0))
-
-  # main loop on cols
-  for (cl_nm in which(where_to_mod)) {
-    tmp_cl <- df[, cl_nm]
-    if (is.null(levels(tmp_cl))) {
-      tmp_cl[sapply(tmp_cl, nchar) == 0] <- rep_str
-    } else {
-      levels(tmp_cl)[sapply(levels(tmp_cl), nchar) == 0] <- rep_str
-    }
-    df[, cl_nm] <- tmp_cl
-  }
-
-  return(df)
-}

tests/testthat/test-logistic_regression.R

@@ -1022,17 +1022,15 @@ testthat::test_that("summarize_logistic works as expected for interaction model
       interaction = "AGE"
     )
   )
-  df <- broom::tidy(mod1, conf_level = 0.99)
-
-  # fix for update in Rtables tern#593
-  df2 <- replace_emptys_with_na(df, rep_str = "_") # _ is the flag value
+  df <- broom::tidy(mod1, conf_level = 0.99) %>%
+    df_explicit_na(na_level = "_")
 
   result <- basic_table() %>%
     summarize_logistic(
       conf_level = 0.99,
       drop_and_remove_str = "_"
     ) %>%
-    build_table(df2)
+    build_table(df)
 
   result_matrix <- to_string_matrix(result)
   expected_matrix <- structure(
@@ -1072,11 +1070,8 @@ testthat::test_that("summarize_logistic works as expected for interaction model
       interaction = "SEX"
     )
   )
-
-  df <- broom::tidy(model, conf_level = 0.99)
-
-  # fix for update in Rtables tern#593
-  df <- replace_emptys_with_na(df, rep_str = "_")
+  df <- broom::tidy(model, conf_level = 0.99) %>%
+    df_explicit_na(na_level = "_")
 
   result <- basic_table() %>%
     summarize_logistic(
@@ -1119,10 +1114,8 @@ testthat::test_that("summarize_logistic works as expected for simple model witho
     adrs,
     variables = list(response = "Response", arm = "ARMCD", covariates = "AGE")
   )
-  df <- broom::tidy(mod1, conf_level = 0.99)
-
-  # fix for update in Rtables tern#593
-  df <- replace_emptys_with_na(df, rep_str = "_")
+  df <- broom::tidy(mod1, conf_level = 0.99) %>%
+    df_explicit_na(na_level = "_")
 
   result <- basic_table() %>%
     summarize_logistic(

tests/testthat/test-table_lgrt02.R

@@ -38,10 +38,8 @@ testthat::test_that("LGRT02 without interaction term is produced correctly", {
     variables = list(response = "Response", arm = "ARMCD", covariates = c("SEX", "RACE", "AGE"))
   )
   conf_level <- 0.95
-  df <- broom::tidy(model, conf_level = conf_level)
-
-  # fix for update in Rtables tern#593
-  df <- replace_emptys_with_na(df, rep_str = "_")
+  df <- broom::tidy(model, conf_level = conf_level) %>%
+    df_explicit_na(na_level = "_")
 
   result <- basic_table() %>%
     summarize_logistic(
@@ -89,10 +87,8 @@ testthat::test_that("LGRT02 with categorical interaction is produced correctly",
     )
   )
   conf_level <- 0.95
-  df <- broom::tidy(model, conf_level = conf_level)
-
-  # fix for update in Rtables tern#593
-  df <- replace_emptys_with_na(df, rep_str = "_")
+  df <- broom::tidy(model, conf_level = conf_level) %>%
+    df_explicit_na(na_level = "_")
 
   result <- basic_table() %>%
     summarize_logistic(
@@ -142,10 +138,8 @@ testthat::test_that("LGRT02 with continuous interaction is produced correctly",
     )
   )
   conf_level <- 0.95
-  df <- broom::tidy(model, conf_level = conf_level, at = c(18, 65))
-
-  # fix for update in Rtables tern#593
-  df <- replace_emptys_with_na(df, rep_str = "_")
+  df <- broom::tidy(model, conf_level = conf_level, at = c(18, 65)) %>%
+    df_explicit_na(na_level = "_")
 
   result <- basic_table() %>%
     summarize_logistic(
@@ -195,10 +189,8 @@ testthat::test_that("LGRT02 with setting values indicating an event and custom a
     response_definition = "1 - response"
   )
   conf_level <- 0.9
-  df <- broom::tidy(model, conf_level = conf_level)
-
-  # fix for update in Rtables tern#593
-  df <- replace_emptys_with_na(df, rep_str = "_")
+  df <- broom::tidy(model, conf_level = conf_level) %>%
+    df_explicit_na(na_level = "_")
 
   result <- basic_table() %>%
     summarize_logistic(