Fix GH-29442 DataFrame.groupby doesn't preserve _metadata

pandas/core/groupby/ops.py

@@ -955,7 +955,9 @@ class SeriesSplitter(DataSplitter):
     def _chop(self, sdata: Series, slice_obj: slice) -> Series:
         # fastpath equivalent to `sdata.iloc[slice_obj]`
         mgr = sdata._mgr.get_slice(slice_obj)
-        return type(sdata)(mgr, name=sdata.name, fastpath=True)
+        return sdata._constructor(mgr, name=sdata.name, fastpath=True).__finalize__(
+            sdata, method="groupby"
+        )
 
 
 class FrameSplitter(DataSplitter):
@@ -971,7 +973,7 @@ def _chop(self, sdata: DataFrame, slice_obj: slice) -> DataFrame:
         # else:
         #     return sdata.iloc[:, slice_obj]
         mgr = sdata._mgr.get_slice(slice_obj, axis=1 - self.axis)
-        return type(sdata)(mgr)
+        return sdata._constructor(mgr).__finalize__(sdata, method="groupby")
 
 
 def get_splitter(

pandas/tests/groupby/test_custom_metadata.py

@@ -0,0 +1,25 @@
+from pandas import DataFrame
+
+
+class CustomDataFrame(DataFrame):
+    """
+    Extension of DataFrame as described in [guidelines]
+
+    [guidelines]: https://pandas.pydata.org/pandas-docs/stable/development/extending.html#override-constructor-properties  # noqa
+    """
+
+    _metadata = ["_custom_metadata"]
+
+    @property
+    def _constructor(self):
+        return CustomDataFrame
+
+
+def test_groupby_with_custom_metadata():
+    custom_df = CustomDataFrame(
+        [[11, 12, 0], [21, 22, 0], [31, 32, 1]], columns=["a", "b", "g"]
+    )
+    custom_df._custom_metadata = "Custom metadata"
+
+    for _, group_df in custom_df.groupby("g"):
+        assert group_df._custom_metadata == "Custom metadata"