FIX warn if an estimator does have a concrete __sklearn_tags__ implementation (#30516)

glemaitre · adrinjalali · thomasjpfan · jeremiedbb · commit 42831e522f9f · 2025-01-09T16:12:35.000+01:00
Co-authored-by: Adrin Jalali &lt;adrin.jalali@gmail.com&gt;
Co-authored-by: Thomas J. Fan &lt;thomasjpfan@gmail.com&gt;
Co-authored-by: Jérémie du Boisberranger &lt;jeremie@probabl.ai&gt;
diff --git a/doc/whats_new/upcoming_changes/sklearn.utils/30516.fix.rst b/doc/whats_new/upcoming_changes/sklearn.utils/30516.fix.rst
@@ -0,0 +1,4 @@
+- Raise a `DeprecationWarning` when there is no concrete implementation of `__sklearn_tags__`
+  in the MRO of the estimator. We request to inherit from `BaseEstimator` that
+  implements `__sklearn_tags__`.
+  By :user:`Guillaume Lemaitre <glemaitre>`
diff --git a/sklearn/utils/_tags.py b/sklearn/utils/_tags.py
@@ -393,7 +393,32 @@ def get_tags(estimator) -> Tags:
     tag_provider = _find_tags_provider(estimator)
 
     if tag_provider == "__sklearn_tags__":
-        tags = estimator.__sklearn_tags__()
+        # TODO(1.7): turn the warning into an error
+        try:
+            tags = estimator.__sklearn_tags__()
+        except AttributeError as exc:
+            if str(exc) == "'super' object has no attribute '__sklearn_tags__'":
+                # workaround the regression reported in
+                # https://github.com/scikit-learn/scikit-learn/issues/30479
+                # `__sklearn_tags__` is implemented by calling
+                # `super().__sklearn_tags__()` but there is no `__sklearn_tags__`
+                # method in the base class.
+                warnings.warn(
+                    f"The following error was raised: {str(exc)}. It seems that "
+                    "there are no classes that implement `__sklearn_tags__` "
+                    "in the MRO and/or all classes in the MRO call "
+                    "`super().__sklearn_tags__()`. Make sure to inherit from "
+                    "`BaseEstimator` which implements `__sklearn_tags__` (or "
+                    "alternatively define `__sklearn_tags__` but we don't recommend "
+                    "this approach). Note that `BaseEstimator` needs to be on the "
+                    "right side of other Mixins in the inheritance order. The "
+                    "default are now used instead since retrieving tags failed. "
+                    "This warning will be replaced by an error in 1.7.",
+                    category=DeprecationWarning,
+                )
+                tags = default_tags(estimator)
+            else:
+                raise
     else:
         # TODO(1.7): Remove this branch of the code
         # Let's go through the MRO and patch each class implementing _more_tags
diff --git a/sklearn/utils/tests/test_tags.py b/sklearn/utils/tests/test_tags.py
@@ -1,12 +1,15 @@
 from dataclasses import dataclass, fields
 
+import numpy as np
 import pytest
 
 from sklearn.base import (
     BaseEstimator,
+    ClassifierMixin,
     RegressorMixin,
     TransformerMixin,
 )
+from sklearn.pipeline import Pipeline
 from sklearn.utils import (
     ClassifierTags,
     InputTags,
@@ -637,3 +640,48 @@ def __sklearn_tags__(self):
     }
     assert old_tags == expected_tags
     assert _to_new_tags(_to_old_tags(new_tags), estimator=estimator) == new_tags
+
+
+# TODO(1.7): Remove this test
+def test_tags_no_sklearn_tags_concrete_implementation():
+    """Non-regression test for:
+    https://github.com/scikit-learn/scikit-learn/issues/30479
+
+    There is no class implementing `__sklearn_tags__` without calling
+    `super().__sklearn_tags__()`. Thus, we raise a warning and request to inherit from
+    `BaseEstimator` that implements `__sklearn_tags__`.
+    """
+
+    class MyEstimator(ClassifierMixin):
+        def __init__(self, *, param=1):
+            self.param = param
+
+        def fit(self, X, y=None):
+            self.is_fitted_ = True
+            return self
+
+        def predict(self, X):
+            return np.full(shape=X.shape[0], fill_value=self.param)
+
+    X = np.array([[1, 2], [2, 3], [3, 4]])
+    y = np.array([1, 0, 1])
+
+    my_pipeline = Pipeline([("estimator", MyEstimator(param=1))])
+    with pytest.warns(DeprecationWarning, match="The following error was raised"):
+        my_pipeline.fit(X, y).predict(X)
+
+    # check that we still raise an error if it is not a AttributeError or related to
+    # __sklearn_tags__
+    class MyEstimator2(MyEstimator, BaseEstimator):
+        def __init__(self, *, param=1, error_type=AttributeError):
+            self.param = param
+            self.error_type = error_type
+
+        def __sklearn_tags__(self):
+            super().__sklearn_tags__()
+            raise self.error_type("test")
+
+    for error_type in (AttributeError, TypeError, ValueError):
+        estimator = MyEstimator2(param=1, error_type=error_type)
+        with pytest.raises(error_type):
+            get_tags(estimator)
