BUG: allow None as final estimator in Pipeline (#554)

Author: orausch

doc/whats_new/v0.5.rst

@@ -41,3 +41,7 @@ Bug
 - Fix bug which converting to COO format sparse when stacking the matrices in
   :class:`imblearn.over_sampling.SMOTENC`. This bug was only old scipy version.
   :issue:`539` by :user:`Guillaume Lemaitre <glemaitre>`.
+
+- Fix bug in :class:`imblearn.pipeline.Pipeline` where None could be the final
+  estimator.
+  :issue:`554` by :user:`Oliver Rausch <orausch>`.

imblearn/pipeline.py

@@ -276,7 +276,6 @@ def fit_transform(self, X, y=None, **fit_params):
         else:
             return last_step.fit(Xt, yt, **fit_params).transform(Xt)
 
-    @if_delegate_has_method(delegate='_final_estimator')
     def fit_resample(self, X, y=None, **fit_params):
         """Fit the model and sample with the final estimator
 
@@ -525,6 +524,13 @@ def _inverse_transform(self, X):
                 Xt = transform.inverse_transform(Xt)
         return Xt
 
+    # need to overwrite sklearn's _final_estimator since sklearn supports
+    # 'passthrough', but imblearn does not.
+    @property
+    def _final_estimator(self):
+        estimator = self.steps[-1][1]
+        return estimator
+
     @if_delegate_has_method(delegate='_final_estimator')
     def score(self, X, y=None, sample_weight=None):
         """Apply transformers/samplers, and score with the final estimator

imblearn/tests/test_pipeline.py

@@ -1104,3 +1104,22 @@ def test_predict_with_predict_params():
     pipe.fit(None, None)
     pipe.predict(X=None, got_attribute=True)
     assert pipe.named_steps['clf'].got_attribute
+
+
+def test_resampler_last_stage_passthrough():
+
+    X, y = make_classification(
+        n_classes=2,
+        class_sep=2,
+        weights=[0.1, 0.9],
+        n_informative=3,
+        n_redundant=1,
+        flip_y=0,
+        n_features=20,
+        n_clusters_per_class=1,
+        n_samples=50000,
+        random_state=0)
+
+    rus = RandomUnderSampler(random_state=42)
+    pipe = make_pipeline(rus, None)
+    pipe.fit_resample(X, y)