scikit-learn · OmarManzoor · Sep 25, 2023 · Sep 15, 2023 · Sep 16, 2023 · Sep 16, 2023
diff --git a/doc/metadata_routing.rst b/doc/metadata_routing.rst
@@ -10,9 +10,11 @@ Metadata Routing
 
 .. note::
   The Metadata Routing API is experimental, and is not implemented yet for many
-  estimators. It may change without the usual deprecation cycle. By default
-  this feature is not enabled. You can enable this feature  by setting the
-  ``enable_metadata_routing`` flag to ``True``:
+  estimators. Please refer to the :ref:`list of supported and unsupported
+  models <metadata_routing_models>` for more information. It may change without
+  the usual deprecation cycle. By default this feature is not enabled. You can
+  enable this feature  by setting the ``enable_metadata_routing`` flag to
+  ``True``::
 
     >>> import sklearn
     >>> sklearn.set_config(enable_metadata_routing=True)
@@ -230,3 +232,72 @@ The issue can be fixed by explicitly setting the request value::
     >>> lr = LogisticRegression().set_fit_request(
     ...     sample_weight=True
     ... ).set_score_request(sample_weight=False)
+
+At the end we disable the configuration flag for metadata routing::
+
+    >>> sklearn.set_config(enable_metadata_routing=False)
+
+.. _metadata_routing_models:
+
+Metadata Routing Support Status
+*******************************
+All consumers (i.e. simple estimators which only consume metadata and don't
+route them) support metadata routing, meaning they can be used inside
+meta-estimators which support metadata routing. However, development of support
+for metadata routing for meta-estimators is in progress, and here is a list of
+meta-estimators and tools which support and don't yet support metadata routing.
+
+
+Meta-estimators and functions supporting metadata routing:
+
+- :class:`sklearn.calibration.CalibratedClassifierCV`
+- :class:`sklearn.compose.ColumnTransformer`
+- :class:`sklearn.linear_model.LogisticRegressionCV`
+- :class:`sklearn.model_selection.GridSearchCV`
+- :class:`sklearn.model_selection.HalvingGridSearchCV`
+- :class:`sklearn.model_selection.HalvingRandomSearchCV`
+- :class:`sklearn.model_selection.RandomizedSearchCV`
+- :func:`sklearn.model_selection.cross_validate`
+- :func:`sklearn.model_selection.cross_val_score`
+- :func:`sklearn.model_selection.cross_val_predict`
+- :class:`sklearn.multioutput.ClassifierChain`
+- :class:`sklearn.multioutput.MultiOutputClassifier`
+- :class:`sklearn.multioutput.MultiOutputRegressor`
+- :class:`sklearn.multioutput.RegressorChain`
+- :class:`sklearn.pipeline.Pipeline`
+
+Meta-estimators and tools not supporting metadata routing yet:
+
+- :class:`sklearn.compose.TransformedTargetRegressor`
+- :class:`sklearn.covariance.GraphicalLassoCV`
+- :class:`sklearn.ensemble.AdaBoostClassifier`
+- :class:`sklearn.ensemble.AdaBoostRegressor`
+- :class:`sklearn.ensemble.BaggingClassifier`
+- :class:`sklearn.ensemble.BaggingRegressor`
+- :class:`sklearn.ensemble.StackingClassifier`
+- :class:`sklearn.ensemble.StackingRegressor`
+- :class:`sklearn.ensemble.VotingClassifier`
+- :class:`sklearn.ensemble.VotingRegressor`
+- :class:`sklearn.feature_selection.RFE`
+- :class:`sklearn.feature_selection.RFECV`
+- :class:`sklearn.feature_selection.SelectFromModel`
+- :class:`sklearn.feature_selection.SequentialFeatureSelector`
+- :class:`sklearn.impute.IterativeImputer`
+- :class:`sklearn.linear_model.ElasticNetCV`
+- :class:`sklearn.linear_model.LarsCV`
+- :class:`sklearn.linear_model.LassoCV`
+- :class:`sklearn.linear_model.LassoLarsCV`
+- :class:`sklearn.linear_model.MultiTaskElasticNetCV`
+- :class:`sklearn.linear_model.MultiTaskLassoCV`
+- :class:`sklearn.linear_model.OrthogonalMatchingPursuitCV`
+- :class:`sklearn.linear_model.RANSACRegressor`
+- :class:`sklearn.linear_model.RidgeClassifierCV`
+- :class:`sklearn.linear_model.RidgeCV`
+- :class:`sklearn.model_selection.learning_curve`
+- :class:`sklearn.model_selection.permutation_test_score`
+- :class:`sklearn.model_selection.validation_curve`
+- :class:`sklearn.multiclass.OneVsOneClassifier`
+- :class:`sklearn.multiclass.OneVsRestClassifier`
+- :class:`sklearn.multiclass.OutputCodeClassifier`
+- :class:`sklearn.pipeline.FeatureUnion`
+- :class:`sklearn.semi_supervised.SelfTrainingClassifier`
diff --git a/doc/whats_new/v1.4.rst b/doc/whats_new/v1.4.rst
@@ -58,17 +58,22 @@ more details.
   estimator's ``fit``, the CV splitter, and the scorer. :pr:`27058` by `Adrin
   Jalali`_.
 
-- |Enhancement| :class:`~compose.ColumnTransformer` now supports metadata routing
+- |Feature| :class:`~compose.ColumnTransformer` now supports metadata routing
   according to :ref:`metadata routing user guide <metadata_routing>`. :pr:`27005`
   by `Adrin Jalali`_.
 
-- |Enhancement| :class:`linear_model.LogisticRegressionCV` now supports
+- |Feature| :class:`linear_model.LogisticRegressionCV` now supports
   metadata routing. :meth:`linear_model.LogisticRegressionCV.fit` now
   accepts ``**params`` which are passed to the underlying splitter and
   scorer. :meth:`linear_model.LogisticRegressionCV.score` now accepts
   ``**score_params`` which are passed to the underlying scorer.
   :pr:`26525` by :user:`Omar Salman <OmarManzoor>`.
 
+- |Fix| All meta-estimators for which metadata routing is not yet implemented
+  now raise a `NotImplementedError`  on `get_metadata_routing` and on `fit` if
+  metadata routing is enabled and any metadata is passed to them. :pr:`27389`
+  by `Adrin Jalali`_.
+
 Support for SciPy sparse arrays
 -------------------------------
 

diff --git a/sklearn/compose/_target.py b/sklearn/compose/_target.py
@@ -12,12 +12,18 @@
 from ..utils import _safe_indexing, check_array
 from ..utils._param_validation import HasMethods
 from ..utils._tags import _safe_tags
+from ..utils.metadata_routing import (
+    _raise_for_unsupported_routing,
+    _RoutingNotSupportedMixin,
+)
 from ..utils.validation import check_is_fitted
 
 __all__ = ["TransformedTargetRegressor"]
 
 
-class TransformedTargetRegressor(RegressorMixin, BaseEstimator):
+class TransformedTargetRegressor(
+    _RoutingNotSupportedMixin, RegressorMixin, BaseEstimator
+):
     """Meta-estimator to regress on a transformed target.
 
     Useful for applying a non-linear transformation to the target `y` in
@@ -222,6 +228,7 @@ def fit(self, X, y, **fit_params):
         self : object
             Fitted estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", **fit_params)
         if y is None:
             raise ValueError(
                 f"This {self.__class__.__name__} estimator "

diff --git a/sklearn/covariance/_graph_lasso.py b/sklearn/covariance/_graph_lasso.py
@@ -22,6 +22,7 @@
 from ..linear_model import lars_path_gram
 from ..model_selection import check_cv, cross_val_score
 from ..utils._param_validation import Interval, StrOptions, validate_params
+from ..utils.metadata_routing import _RoutingNotSupportedMixin
 from ..utils.parallel import Parallel, delayed
 from ..utils.validation import (
     _is_arraylike_not_scalar,
@@ -705,7 +706,7 @@ def graphical_lasso_path(
     return covariances_, precisions_
 
 
-class GraphicalLassoCV(BaseGraphicalLasso):
+class GraphicalLassoCV(_RoutingNotSupportedMixin, BaseGraphicalLasso):
     """Sparse inverse covariance w/ cross-validated choice of the l1 penalty.
 
     See glossary entry for :term:`cross-validation estimator`.

diff --git a/sklearn/ensemble/_bagging.py b/sklearn/ensemble/_bagging.py
@@ -19,6 +19,10 @@
 from ..utils import check_random_state, column_or_1d, indices_to_mask
 from ..utils._param_validation import HasMethods, Interval, RealNotInt, StrOptions
 from ..utils._tags import _safe_tags
+from ..utils.metadata_routing import (
+    _raise_for_unsupported_routing,
+    _RoutingNotSupportedMixin,
+)
 from ..utils.metaestimators import available_if
 from ..utils.multiclass import check_classification_targets
 from ..utils.parallel import Parallel, delayed
@@ -326,6 +330,7 @@ def fit(self, X, y, sample_weight=None):
         self : object
             Fitted estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", sample_weight=sample_weight)
         # Convert data (X is required to be 2d and indexable)
         X, y = self._validate_data(
             X,
@@ -542,7 +547,7 @@ def estimators_samples_(self):
         return [sample_indices for _, sample_indices in self._get_estimators_indices()]
 
 
-class BaggingClassifier(ClassifierMixin, BaseBagging):
+class BaggingClassifier(_RoutingNotSupportedMixin, ClassifierMixin, BaseBagging):
     """A Bagging classifier.
 
     A Bagging classifier is an ensemble meta-estimator that fits base
@@ -990,7 +995,7 @@ def _more_tags(self):
         return {"allow_nan": _safe_tags(estimator, "allow_nan")}
 
 
-class BaggingRegressor(RegressorMixin, BaseBagging):
+class BaggingRegressor(_RoutingNotSupportedMixin, RegressorMixin, BaseBagging):
     """A Bagging regressor.
 
     A Bagging regressor is an ensemble meta-estimator that fits base

diff --git a/sklearn/ensemble/_stacking.py b/sklearn/ensemble/_stacking.py
@@ -26,6 +26,10 @@
 from ..utils import Bunch
 from ..utils._estimator_html_repr import _VisualBlock
 from ..utils._param_validation import HasMethods, StrOptions
+from ..utils.metadata_routing import (
+    _raise_for_unsupported_routing,
+    _RoutingNotSupportedMixin,
+)
 from ..utils.metaestimators import available_if
 from ..utils.multiclass import check_classification_targets, type_of_target
 from ..utils.parallel import Parallel, delayed
@@ -380,7 +384,7 @@ def _sk_visual_block_with_final_estimator(self, final_estimator):
         return _VisualBlock("serial", (parallel, final_block), dash_wrapped=False)
 
 
-class StackingClassifier(ClassifierMixin, _BaseStacking):
+class StackingClassifier(_RoutingNotSupportedMixin, ClassifierMixin, _BaseStacking):
     """Stack of estimators with a final classifier.
 
     Stacked generalization consists in stacking the output of individual
@@ -641,6 +645,7 @@ def fit(self, X, y, sample_weight=None):
         self : object
             Returns a fitted instance of estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", sample_weight=sample_weight)
         check_classification_targets(y)
         if type_of_target(y) == "multilabel-indicator":
             self._label_encoder = [LabelEncoder().fit(yk) for yk in y.T]
@@ -761,7 +766,7 @@ def _sk_visual_block_(self):
         return super()._sk_visual_block_with_final_estimator(final_estimator)
 
 
-class StackingRegressor(RegressorMixin, _BaseStacking):
+class StackingRegressor(_RoutingNotSupportedMixin, RegressorMixin, _BaseStacking):
     """Stack of estimators with a final regressor.
 
     Stacked generalization consists in stacking the output of individual
@@ -952,6 +957,7 @@ def fit(self, X, y, sample_weight=None):
         self : object
             Returns a fitted instance.
         """
+        _raise_for_unsupported_routing(self, "fit", sample_weight=sample_weight)
         y = column_or_1d(y, warn=True)
         return super().fit(X, y, sample_weight)
 

diff --git a/sklearn/ensemble/_voting.py b/sklearn/ensemble/_voting.py
@@ -30,6 +30,10 @@
 from ..utils import Bunch
 from ..utils._estimator_html_repr import _VisualBlock
 from ..utils._param_validation import StrOptions
+from ..utils.metadata_routing import (
+    _raise_for_unsupported_routing,
+    _RoutingNotSupportedMixin,
+)
 from ..utils.metaestimators import available_if
 from ..utils.multiclass import check_classification_targets
 from ..utils.parallel import Parallel, delayed
@@ -152,7 +156,7 @@ def _more_tags(self):
         return {"preserves_dtype": []}
 
 
-class VotingClassifier(ClassifierMixin, _BaseVoting):
+class VotingClassifier(_RoutingNotSupportedMixin, ClassifierMixin, _BaseVoting):
     """Soft Voting/Majority Rule classifier for unfitted estimators.
 
     Read more in the :ref:`User Guide <voting_classifier>`.
@@ -336,6 +340,7 @@ def fit(self, X, y, sample_weight=None):
         self : object
             Returns the instance itself.
         """
+        _raise_for_unsupported_routing(self, "fit", sample_weight=sample_weight)
         check_classification_targets(y)
         if isinstance(y, np.ndarray) and len(y.shape) > 1 and y.shape[1] > 1:
             raise NotImplementedError(
@@ -478,7 +483,7 @@ def get_feature_names_out(self, input_features=None):
         return np.asarray(names_out, dtype=object)
 
 
-class VotingRegressor(RegressorMixin, _BaseVoting):
+class VotingRegressor(_RoutingNotSupportedMixin, RegressorMixin, _BaseVoting):
     """Prediction voting regressor for unfitted estimators.
 
     A voting regressor is an ensemble meta-estimator that fits several base
@@ -601,6 +606,7 @@ def fit(self, X, y, sample_weight=None):
         self : object
             Fitted estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", sample_weight=sample_weight)
         y = column_or_1d(y, warn=True)
         return super().fit(X, y, sample_weight)
 

diff --git a/sklearn/ensemble/_weight_boosting.py b/sklearn/ensemble/_weight_boosting.py
@@ -42,6 +42,10 @@
 from ..utils import _safe_indexing, check_random_state
 from ..utils._param_validation import HasMethods, Interval, StrOptions
 from ..utils.extmath import softmax, stable_cumsum
+from ..utils.metadata_routing import (
+    _raise_for_unsupported_routing,
+    _RoutingNotSupportedMixin,
+)
 from ..utils.validation import (
     _check_sample_weight,
     _num_samples,
@@ -132,6 +136,7 @@ def fit(self, X, y, sample_weight=None):
         self : object
             Fitted estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", sample_weight=sample_weight)
         X, y = self._validate_data(
             X,
             y,
@@ -338,7 +343,9 @@ def _samme_proba(estimator, n_classes, X):
     )
 
 
-class AdaBoostClassifier(ClassifierMixin, BaseWeightBoosting):
+class AdaBoostClassifier(
+    _RoutingNotSupportedMixin, ClassifierMixin, BaseWeightBoosting
+):
     """An AdaBoost classifier.
 
     An AdaBoost [1]_ classifier is a meta-estimator that begins by fitting a
@@ -980,7 +987,7 @@ def predict_log_proba(self, X):
         return np.log(self.predict_proba(X))
 
 
-class AdaBoostRegressor(RegressorMixin, BaseWeightBoosting):
+class AdaBoostRegressor(_RoutingNotSupportedMixin, RegressorMixin, BaseWeightBoosting):
     """An AdaBoost regressor.
 
     An AdaBoost [1] regressor is a meta-estimator that begins by fitting a

diff --git a/sklearn/feature_selection/_from_model.py b/sklearn/feature_selection/_from_model.py
@@ -10,6 +10,10 @@
 from ..exceptions import NotFittedError
 from ..utils._param_validation import HasMethods, Interval, Options
 from ..utils._tags import _safe_tags
+from ..utils.metadata_routing import (
+    _raise_for_unsupported_routing,
+    _RoutingNotSupportedMixin,
+)
 from ..utils.metaestimators import available_if
 from ..utils.validation import _num_features, check_is_fitted, check_scalar
 from ._base import SelectorMixin, _get_feature_importances
@@ -78,7 +82,9 @@ def _estimator_has(attr):
     )
 
 
-class SelectFromModel(MetaEstimatorMixin, SelectorMixin, BaseEstimator):
+class SelectFromModel(
+    _RoutingNotSupportedMixin, MetaEstimatorMixin, SelectorMixin, BaseEstimator
+):
     """Meta-transformer for selecting features based on importance weights.
 
     .. versionadded:: 0.17
@@ -342,6 +348,7 @@ def fit(self, X, y=None, **fit_params):
         self : object
             Fitted estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", **fit_params)
         self._check_max_features(X)
 
         if self.prefit:

diff --git a/sklearn/feature_selection/_rfe.py b/sklearn/feature_selection/_rfe.py
@@ -17,6 +17,10 @@
 from ..model_selection._validation import _score
 from ..utils._param_validation import HasMethods, Interval, RealNotInt
 from ..utils._tags import _safe_tags
+from ..utils.metadata_routing import (
+    _raise_for_unsupported_routing,
+    _RoutingNotSupportedMixin,
+)
 from ..utils.metaestimators import _safe_split, available_if
 from ..utils.parallel import Parallel, delayed
 from ..utils.validation import check_is_fitted
@@ -56,7 +60,7 @@ def _estimator_has(attr):
     )
 
 
-class RFE(SelectorMixin, MetaEstimatorMixin, BaseEstimator):
+class RFE(_RoutingNotSupportedMixin, SelectorMixin, MetaEstimatorMixin, BaseEstimator):
     """Feature ranking with recursive feature elimination.
 
     Given an external estimator that assigns weights to features (e.g., the
@@ -251,6 +255,7 @@ def fit(self, X, y, **fit_params):
         self : object
             Fitted estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", **fit_params)
         return self._fit(X, y, **fit_params)
 
     def _fit(self, X, y, step_score=None, **fit_params):
@@ -680,6 +685,7 @@ def fit(self, X, y, groups=None):
         self : object
             Fitted estimator.
         """
+        _raise_for_unsupported_routing(self, "fit", groups=groups)
         tags = self._get_tags()
         X, y = self._validate_data(
             X,