Merge pull request #274 from tidymodels/fix-bug-clean_levels

fix bug in step_clean_levels()
tidymodels · Oct 30, 2024 · 8997aef · 8997aef
2 parents 60999a6 + f38c4de
commit 8997aef
Show file tree

Hide file tree

Showing 3 changed files with 12 additions and 4 deletions.
diff --git a/NEWS.md b/NEWS.md
@@ -4,7 +4,9 @@
 
 * Calling `?tidy.step_*()` now sends you to the documentation for `step_*()` where the outcome is documented. (#261)
 
-* `step_textfeatures()` has been made faster and more robust. #265
+* `step_textfeatures()` has been made faster and more robust. (#265)
+
+* Fixed bug in `step_clean_levels()` where it would produce NAs for character columns. (#274)
 
 # textrecipes 1.0.6
 

diff --git a/R/clean_levels.R b/R/clean_levels.R
@@ -131,9 +131,14 @@ bake.step_clean_levels <- function(object, new_data, ...) {
   }
 
   for (col_name in col_names) {
-    new_data[[col_name]] <- dplyr::recode_factor(
-      new_data[[col_name]], !!!object$clean[[col_name]]
-    )
+    if (is.factor(new_data[[col_name]])) {
+      new_data[[col_name]] <- dplyr::recode_factor(
+        new_data[[col_name]], !!!object$clean[[col_name]]
+      )      
+    } else {
+      new_data[[col_name]] <- janitor::make_clean_names(new_data[[col_name]])
+
+    }
   }
 
   new_data

diff --git a/tests/testthat/test-clean_levels.R b/tests/testthat/test-clean_levels.R
@@ -22,6 +22,7 @@ test_that("character input", {
   cleaned_te <- bake(cleaned, new_data = smith_te)
 
   expect_equal(sum(grepl(" ", cleaned_tr$name)), 0)
+  expect_equal(sum(is.na(cleaned_tr$name)), 0)
   expect_equal(sum(levels(cleaned_tr$name) %in% smith_tr$name), 0)
 
   tidy_exp_tr <- tibble(