feat: enable GroupedTransformer.set_output(..) (#697)

* feat: GroupedTransformer allow set_output

* skip polars for pre 1.4.0 version

sklego/meta/_grouped_utils.py

@@ -25,7 +25,7 @@ def parse_X_y(X, y, groups, check_X=True, **kwargs) -> nw.DataFrame:
         X = nw.from_native(pd.DataFrame(X))
 
     # Check groups and feaures values
-    if groups is not None:
+    if groups:
         _validate_groups_values(X, groups)
 
         if check_X:

sklego/meta/grouped_transformer.py

@@ -1,4 +1,4 @@
-from typing import Union
+from typing import List, Union
 
 import narwhals.stable.v1 as nw
 import numpy as np
@@ -108,14 +108,19 @@ def fit(self, X, y=None):
         """
         self.__check_transformer()
         self.fallback_ = None
-        self.groups_ = as_list(self.groups) if self.groups is not None else None
+        self.groups_ = as_list(self.groups) if self.groups is not None else []
 
         X = nw.from_native(X, strict=False, eager_only=True)
-        if not isinstance(X, nw.DataFrame) and self.groups_ is not None:
+
+        if isinstance(X, nw.DataFrame):
+            self.feature_names_out_ = [c for c in X.columns if c not in self.groups_]
+
+        else:
             # Accounts for negative indices if X is an array
             self.groups_ = [
                 X.shape[1] + group if isinstance(group, int) and group < 0 else group for group in self.groups_
             ]
+            self.feature_names_out_ = [f"x{i}" for i in range(X.shape[1] - len(self.groups_))]
 
         frame = parse_X_y(X, y, self.groups_, check_X=self.check_X, **self._check_kwargs)
 
@@ -203,3 +208,7 @@ def transform(self, X):
 
     def _more_tags(self):
         return {"allow_nan": True}
+
+    def get_feature_names_out(self) -> List[str]:
+        "Alias for the `feature_names_out_` attribute defined during fit."
+        return self.feature_names_out_

tests/test_meta/test_grouped_transformer.py

@@ -4,6 +4,7 @@
 import pandas as pd
 import polars as pl
 import pytest
+import sklearn
 from sklearn import clone
 from sklearn.linear_model import LinearRegression
 from sklearn.pipeline import Pipeline
@@ -25,6 +26,7 @@ def test_sklearn_compatible_estimator(estimator, check):
         "check_dtype_object",  # custom message
         "check_estimators_empty_data_messages",  # custom message
         "check_estimators_pickle",  # Fails if input contains nan
+        "check_fit1d",
     }:
         pytest.skip()
 
@@ -389,3 +391,16 @@ def test_transform_with_y(transformer):
     )
 
     assert np.allclose(X_naive.to_numpy(), X_grouped)
+
+
+@pytest.mark.parametrize(("frame_func", "transform_output"), [(pd.DataFrame, "pandas"), (pl.DataFrame, "polars")])
+def test_set_output(penguins_df, frame_func, transform_output):
+    if transform_output == "polars" and sklearn.__version__ < "1.4.0":
+        pytest.skip()
+
+    X = frame_func(penguins_df.drop(columns=["sex"]))
+    y = penguins_df["sex"]
+
+    meta = GroupedTransformer(StandardScaler(), groups="island").set_output(transform=transform_output)
+    transformed = meta.fit_transform(X, y)
+    assert isinstance(transformed, frame_func)