Merge pull request #18 from ipanepen/rle-bug

fix: makes fit_transform behavior consistent with fit and transform

src/sagemaker_sklearn_extension/preprocessing/encoders.py

@@ -278,8 +278,6 @@ def _is_sorted(self, iterable):
     def fit_transform(self, y):
         """Fit label encoder and return encoded labels.
 
-        ``fill_unseen_labels=True`` does nothing in ``fit_transform`` because there will be no unseen labels.
-
         Parameters
         ----------
         y : array-like of shape [n_samples]
@@ -290,12 +288,7 @@ def fit_transform(self, y):
         y_encoded : array-like of shape [n_samples]
                     Encoded label values.
         """
-        y = column_or_1d(y, warn=True)
-        sorted_labels = self._check_labels_and_sort()
-        self.classes_, y_encoded = (
-            _encode(y, uniques=sorted_labels, encode=True) if sorted_labels else _encode(y, encode=True)
-        )
-        return y_encoded
+        return self.fit(y).transform(y)
 
     def transform(self, y):
         """Transform labels to normalized encoding.

test/test_preprocessing_encoders.py

@@ -148,6 +148,13 @@ def test_robust_label_encoder_sorted_labels(labels):
     assert_array_equal(list(enc.classes_), labels)
     assert_array_equal(enc.transform([labels[2], labels[1], "173"]), [2, 1, 3])
 
+    # Test that fit_transform has the same behavior
+    enc = RobustLabelEncoder(labels=labels)
+    y_transformed = enc.fit_transform([labels[2], labels[1], "173"])
+
+    assert_array_equal(list(enc.classes_), labels)
+    assert_array_equal(y_transformed, [2, 1, 3])
+
 
 @pytest.mark.parametrize("labels", (["-12", "9", "3"], ["-12.", "9.", "3."]))
 def test_robust_label_encoder_unsorted_labels_warning(labels):
@@ -158,6 +165,29 @@ def test_robust_label_encoder_unsorted_labels_warning(labels):
     assert_array_equal(list(enc.classes_), sorted(labels))
     assert_array_equal(enc.transform([labels[1], labels[2], "173"]), [2, 1, 3])
 
+    # Test that fit_transform has the same behavior
+    enc = RobustLabelEncoder(labels=labels)
+    with pytest.warns(UserWarning):
+        y_transformed = enc.fit_transform([labels[1], labels[2], "173"])
+
+    assert_array_equal(list(enc.classes_), sorted(labels))
+    assert_array_equal(y_transformed, [2, 1, 3])
+
+
+def test_robust_label_encoder_fill_label_value():
+    y = np.array([1, 1, 0, 1, 1])
+    enc = RobustLabelEncoder(labels=[1], fill_label_value=0)
+    enc.fit(y)
+    y_transform = enc.transform(y)
+    assert_array_equal(y_transform, [0, 0, 1, 0, 0])
+    assert_array_equal(enc.inverse_transform(y_transform), y)
+
+    # Test that fit_transform has the same behavior
+    enc = RobustLabelEncoder(labels=[1], fill_label_value=0)
+    y_transform = enc.fit_transform(y)
+    assert_array_equal(y_transform, [0, 0, 1, 0, 0])
+    assert_array_equal(enc.inverse_transform(y_transform), y)
+
 
 @pytest.mark.parametrize(
     "y, y_expected",