scikit-learn · jeremiedbb · Mar 24, 2023 · Apr 7, 2022 · Apr 7, 2022 · Apr 7, 2022
diff --git a/sklearn/calibration.py b/sklearn/calibration.py
@@ -25,16 +25,17 @@
     RegressorMixin,
     clone,
     MetaEstimatorMixin,
-    is_classifier,
 )
 from .preprocessing import label_binarize, LabelEncoder
 from .utils import (
     column_or_1d,
     indexable,
     check_matplotlib_support,
+    _safe_indexing,
 )
+from .utils._response import _get_response_values_binary
 
-from .utils.multiclass import check_classification_targets
+from .utils.multiclass import check_classification_targets, type_of_target
 from .utils.parallel import delayed, Parallel
 from .utils._param_validation import StrOptions, HasMethods, Hidden
 from .utils.validation import (
@@ -44,12 +45,10 @@
     check_consistent_length,
     check_is_fitted,
 )
-from .utils import _safe_indexing
 from .isotonic import IsotonicRegression
 from .svm import LinearSVC
 from .model_selection import check_cv, cross_val_predict
 from .metrics._base import _check_pos_label_consistency
-from .metrics._plot.base import _get_response
 
 
 class CalibratedClassifierCV(ClassifierMixin, MetaEstimatorMixin, BaseEstimator):
@@ -1264,11 +1263,9 @@ def from_estimator(
         method_name = f"{cls.__name__}.from_estimator"
         check_matplotlib_support(method_name)
 
-        if not is_classifier(estimator):
-            raise ValueError("'estimator' should be a fitted classifier.")
-
-        y_prob, pos_label = _get_response(
-            X, estimator, response_method="predict_proba", pos_label=pos_label
+        check_is_fitted(estimator)
+        y_prob, pos_label = _get_response_values_binary(
+            estimator, X, response_method="predict_proba", pos_label=pos_label
         )
 
         name = name if name is not None else estimator.__class__.__name__
@@ -1381,9 +1378,15 @@ def from_predictions(
         >>> disp = CalibrationDisplay.from_predictions(y_test, y_prob)
         >>> plt.show()
         """
-        method_name = f"{cls.__name__}.from_estimator"
+        method_name = f"{cls.__name__}.from_predictions"
         check_matplotlib_support(method_name)
 
+        target_type = type_of_target(y_true)
+        if target_type != "binary":
+            raise ValueError(
+                f"The target y is not binary. Got {target_type} type of target."
+            )
+
         prob_true, prob_pred = calibration_curve(
             y_true, y_prob, n_bins=n_bins, strategy=strategy, pos_label=pos_label
         )

diff --git a/sklearn/ensemble/_stacking.py b/sklearn/ensemble/_stacking.py
@@ -30,11 +30,14 @@
 from ..utils import Bunch
 from ..utils.multiclass import check_classification_targets, type_of_target
 from ..utils.metaestimators import available_if
-from ..utils.validation import check_is_fitted
-from ..utils.validation import column_or_1d
 from ..utils.parallel import delayed, Parallel
 from ..utils._param_validation import HasMethods, StrOptions
-from ..utils.validation import _check_feature_names_in
+from ..utils.validation import (
+    _check_feature_names_in,
+    _check_response_method,
+    check_is_fitted,
+    column_or_1d,
+)
 
 
 def _estimator_has(attr):
@@ -146,20 +149,15 @@ def _method_name(name, estimator, method):
         if estimator == "drop":
             return None
         if method == "auto":
-            if getattr(estimator, "predict_proba", None):
-                return "predict_proba"
-            elif getattr(estimator, "decision_function", None):
-                return "decision_function"
-            else:
-                return "predict"
-        else:
-            if not hasattr(estimator, method):
-                raise ValueError(
-                    "Underlying estimator {} does not implement the method {}.".format(
-                        name, method
-                    )
-                )
-            return method
+            method = ["predict_proba", "decision_function", "predict"]
+        try:
+            method_name = _check_response_method(estimator, method).__name__
+        except AttributeError as e:
+            raise ValueError(
+                f"Underlying estimator {name} does not implement the method {method}."
+            ) from e
+
+        return method_name
 
     def fit(self, X, y, sample_weight=None):
         """Fit the estimators.

diff --git a/sklearn/ensemble/tests/test_stacking.py b/sklearn/ensemble/tests/test_stacking.py
@@ -572,9 +572,13 @@ def test_stacking_prefit(Stacker, Estimator, stack_method, final_estimator, X, y
 
     # mock out fit and stack_method to be asserted later
     for _, estimator in estimators:
-        estimator.fit = Mock()
+        estimator.fit = Mock(name="fit")
         stack_func = getattr(estimator, stack_method)
-        setattr(estimator, stack_method, Mock(side_effect=stack_func))
+        predict_method_mocked = Mock(side_effect=stack_func)
+        # Mocking a method will not provide a `__name__` while Python methods
+        # do and we are using it in `_get_response_method`.
+        predict_method_mocked.__name__ = stack_method
+        setattr(estimator, stack_method, predict_method_mocked)
 
     stacker = Stacker(
         estimators=estimators, cv="prefit", final_estimator=final_estimator

diff --git a/sklearn/metrics/_plot/base.py b/sklearn/metrics/_plot/base.py
diff --git a/sklearn/metrics/_plot/det_curve.py b/sklearn/metrics/_plot/det_curve.py
@@ -1,11 +1,10 @@
 import scipy as sp
 
-from .base import _get_response
-
 from .. import det_curve
 from .._base import _check_pos_label_consistency
 
 from ...utils import check_matplotlib_support
+from ...utils._response import _get_response_values_binary
 
 
 class DetCurveDisplay:
@@ -168,9 +167,9 @@ def from_estimator(
 
         name = estimator.__class__.__name__ if name is None else name
 
-        y_pred, pos_label = _get_response(
-            X,
+        y_pred, pos_label = _get_response_values_binary(
             estimator,
+            X,
             response_method,
             pos_label=pos_label,
         )

diff --git a/sklearn/metrics/_plot/precision_recall_curve.py b/sklearn/metrics/_plot/precision_recall_curve.py
@@ -1,12 +1,10 @@
-from sklearn.base import is_classifier
-from .base import _get_response
-
 from .. import average_precision_score
 from .. import precision_recall_curve
 from .._base import _check_pos_label_consistency
 from .._classification import check_consistent_length
 
 from ...utils import check_matplotlib_support
+from ...utils._response import _get_response_values_binary
 
 
 class PrecisionRecallDisplay:
@@ -269,11 +267,10 @@ def from_estimator(
         """
         method_name = f"{cls.__name__}.from_estimator"
         check_matplotlib_support(method_name)
-        if not is_classifier(estimator):
-            raise ValueError(f"{method_name} only supports classifiers")
-        y_pred, pos_label = _get_response(
-            X,
+
+        y_pred, pos_label = _get_response_values_binary(
             estimator,
+            X,
             response_method,
             pos_label=pos_label,
         )

diff --git a/sklearn/metrics/_plot/roc_curve.py b/sklearn/metrics/_plot/roc_curve.py
@@ -1,10 +1,9 @@
-from .base import _get_response
-
 from .. import auc
 from .. import roc_curve
 from .._base import _check_pos_label_consistency
 
 from ...utils import check_matplotlib_support
+from ...utils._response import _get_response_values_binary
 
 
 class RocCurveDisplay:
@@ -231,9 +230,9 @@ def from_estimator(
 
         name = estimator.__class__.__name__ if name is None else name
 
-        y_pred, pos_label = _get_response(
-            X,
+        y_pred, pos_label = _get_response_values_binary(
             estimator,
+            X,
             response_method=response_method,
             pos_label=pos_label,
         )

diff --git a/sklearn/metrics/_plot/tests/test_base.py b/sklearn/metrics/_plot/tests/test_base.py