Fixes #453. Categorical targets.

Author: PaulWestenthanner

CHANGELOG.md

@@ -1,3 +1,9 @@
+v.2.8.1
+=======
+
+* Fix: Support and test string targets and `pd.Categorical` targets.
+* Fix: Docs typo.
+ 
 v.2.8.0
 =======
 

category_encoders/utils.py

@@ -457,7 +457,7 @@ def fit(self, X: X_type, y: y_type | None = None, **kwargs):
         if self.__sklearn_tags__().target_tags.required:
             if not is_numeric_dtype(y):
                 self.lab_encoder_ = LabelEncoder()
-                y = self.lab_encoder_.fit_transform(y)
+                y = pd.Series(self.lab_encoder_.fit_transform(y), index=y.index)
             else:
                 self.lab_encoder_ = None
 
@@ -621,7 +621,7 @@ def transform(self, X: X_type, y: y_type | None = None, override_return_df: bool
         X, y = convert_inputs(X, y, deep=True)
         self._check_transform_inputs(X)
         if y is not None and self.lab_encoder_ is not None:
-            y = self.lab_encoder_.transform(y)
+            y = pd.Series(self.lab_encoder_.transform(y), index=y.index)
 
         if not list(self.cols):
             return X

tests/test_encoders.py

@@ -436,6 +436,36 @@ def test_types(self):
             encoder = getattr(encoders, encoder_name)()
             encoder.fit_transform(X, y)
 
+    def test_string_targets(self):
+        """Test encoders with targets of type pd.Categorical or string."""
+        X = pd.DataFrame({'feature': ['A', 'B', 'A', 'C']})
+        y_string = pd.Series(['yes', 'no', 'yes', 'no'])
+
+        for encoder_name in encoders.__all__:
+            with self.subTest(encoder_name=encoder_name):
+                enc = getattr(encoders, encoder_name)()
+
+                # Test with string target
+                enc.fit(X, y_string)
+                transformed = enc.transform(X)
+                th.verify_numeric(transformed)
+                self.assertEqual(len(transformed), 4)
+    def test_categorical_targets(self):
+        """Test encoders with targets of type pd.Categorical or string."""
+        X = pd.DataFrame({'feature': ['A', 'B', 'A', 'C']})
+        y_categorical = pd.Categorical([1, 0, 1, 0])
+
+        for encoder_name in encoders.__all__:
+            with self.subTest(encoder_name=encoder_name):
+                enc = getattr(encoders, encoder_name)()
+
+                # Test with pd.Categorical target
+                enc.fit(X, y_categorical)
+                transformed = enc.transform(X)
+                th.verify_numeric(transformed)
+                self.assertEqual(len(transformed), 4)
+
+
     def test_preserve_column_order(self):
         """Test that the encoder preserves the column order."""
         binary_cat_example = pd.DataFrame(