scikit-learn · jnothman · Apr 5, 2021 · Mar 31, 2021 · Apr 4, 2021
diff --git a/doc/whats_new/v1.0.rst b/doc/whats_new/v1.0.rst
@@ -249,6 +249,14 @@ Changelog
  Use ``var_`` instead.
  :pr:`18842` by :user:`Hong Shao Yang <hongshaoyang>`.

+:mod:`sklearn.pipeline`
+.......................
+
+- |API| The `predict_proba` and `predict_log_proba` methods of the
+  :class:`Pipeline` class now support passing prediction kwargs to
+  the final estimator.
+  :pr:`19790` by :user:`Christopher Flynn <crflynn>`.
+
 :mod:`sklearn.preprocessing`
 ............................


diff --git a/sklearn/pipeline.py b/sklearn/pipeline.py
@@ -456,7 +456,7 @@ def fit_predict(self, X, y=None, **fit_params):
        return y_pred

    @if_delegate_has_method(delegate='_final_estimator')
-    def predict_proba(self, X):
+    def predict_proba(self, X, **predict_proba_params):
        """Apply transforms, and predict_proba of the final estimator

        Parameters
@@ -465,14 +465,18 @@ def predict_proba(self, X):
            Data to predict on. Must fulfill input requirements of first step
            of the pipeline.

+        **predict_proba_params : dict of string -> object
+            Parameters to the ``predict_proba`` called at the end of all
+            transformations in the pipeline.
+
        Returns
        -------
        y_proba : array-like of shape (n_samples, n_classes)
        """
        Xt = X
        for _, name, transform in self._iter(with_final=False):
            Xt = transform.transform(Xt)
-        return self.steps[-1][-1].predict_proba(Xt)
+        return self.steps[-1][-1].predict_proba(Xt, **predict_proba_params)

    @if_delegate_has_method(delegate='_final_estimator')
    def decision_function(self, X):
@@ -513,7 +517,7 @@ def score_samples(self, X):
        return self.steps[-1][-1].score_samples(Xt)

    @if_delegate_has_method(delegate='_final_estimator')
-    def predict_log_proba(self, X):
+    def predict_log_proba(self, X, **predict_log_proba_params):
        """Apply transforms, and predict_log_proba of the final estimator

        Parameters
@@ -522,14 +526,20 @@ def predict_log_proba(self, X):
            Data to predict on. Must fulfill input requirements of first step
            of the pipeline.

+        **predict_log_proba_params : dict of string -> object
+            Parameters to the ``predict_log_proba`` called at the end of all
+            transformations in the pipeline.
+
        Returns
        -------
        y_score : array-like of shape (n_samples, n_classes)
        """
        Xt = X
        for _, name, transform in self._iter(with_final=False):
            Xt = transform.transform(Xt)
-        return self.steps[-1][-1].predict_log_proba(Xt)
+        return self.steps[-1][-1].predict_log_proba(
+            Xt, **predict_log_proba_params
+        )

    @property
    def transform(self):

diff --git a/sklearn/tests/test_pipeline.py b/sklearn/tests/test_pipeline.py
@@ -160,6 +160,14 @@ def predict(self, X, got_attribute=False):
        self.got_attribute = got_attribute
        return self

+    def predict_proba(self, X, got_attribute=False):
+        self.got_attribute = got_attribute
+        return self
+
+    def predict_log_proba(self, X, got_attribute=False):
+        self.got_attribute = got_attribute
+        return self
+

 def test_pipeline_init():
    # Test the various init parameters of the pipeline.
@@ -449,12 +457,16 @@ def test_fit_predict_with_intermediate_fit_params():
    assert 'should_succeed' not in pipe.named_steps['transf'].fit_params


-def test_predict_with_predict_params():
-    # tests that Pipeline passes predict_params to the final estimator
-    # when predict is invoked
+@pytest.mark.parametrize("method_name", [
+    "predict", "predict_proba", "predict_log_proba"
+])
+def test_predict_methods_with_predict_params(method_name):
+    # tests that Pipeline passes predict_* to the final estimator
+    # when predict_* is invoked
    pipe = Pipeline([('transf', Transf()), ('clf', DummyEstimatorParams())])
    pipe.fit(None, None)
-    pipe.predict(X=None, got_attribute=True)
+    method = getattr(pipe, method_name)
+    method(X=None, got_attribute=True)

    assert pipe.named_steps['clf'].got_attribute