Fix compressed multiindex for output of groupby.rolling

doc/source/whatsnew/v1.2.0.rst

@@ -314,6 +314,7 @@ Groupby/resample/rolling
 - Bug when subsetting columns on a :class:`~pandas.core.groupby.DataFrameGroupBy` (e.g. ``df.groupby('a')[['b']])``) would reset the attributes ``axis``, ``dropna``, ``group_keys``, ``level``, ``mutated``, ``sort``, and ``squeeze`` to their default values. (:issue:`9959`)
 - Bug in :meth:`DataFrameGroupby.tshift` failing to raise ``ValueError`` when a frequency cannot be inferred for the index of a group (:issue:`35937`)
 - Bug in :meth:`DataFrame.groupby` does not always maintain column index name for ``any``, ``all``, ``bfill``, ``ffill``, ``shift`` (:issue:`29764`)
+- Bug in :meth:`Series.groupby.rolling` compressed number of levels of :class:`MultiIndex` in input to one (:issue:`36018`)
 -
 
 Reshaping

pandas/core/window/rolling.py

@@ -2211,17 +2211,19 @@ def _apply(
         # Compose MultiIndex result from grouping levels then rolling level
         # Aggregate the MultiIndex data as tuples then the level names
         grouped_object_index = self.obj.index
-        grouped_index_name = [grouped_object_index.name]
+        grouped_index_name = [*grouped_object_index.names]
         groupby_keys = [grouping.name for grouping in self._groupby.grouper._groupings]
         result_index_names = groupby_keys + grouped_index_name
 
         result_index_data = []
         for key, values in self._groupby.grouper.indices.items():
             for value in values:
-                if not is_list_like(key):
-                    data = [key, grouped_object_index[value]]
+                data = [key] if not is_list_like(key) else [*key]
+                if is_list_like(grouped_object_index[value]):
+                    grouped_object_index_value = [*grouped_object_index[value]]
                 else:
-                    data = [*key, grouped_object_index[value]]
+                    grouped_object_index_value = [grouped_object_index[value]]
+                data.extend(grouped_object_index_value)
                 result_index_data.append(tuple(data))
 
         result_index = MultiIndex.from_tuples(

pandas/tests/window/test_grouper.py

@@ -372,3 +372,24 @@ def test_groupby_subset_rolling_subset_with_closed(self):
             name="column1",
         )
         tm.assert_series_equal(result, expected)
+
+    @pytest.mark.parametrize("func", ["max", "min"])
+    def test_groupby_rolling_index_changed(self, func):
+        # GH: #36018 nlevels of MultiIndex changed
+        ds = Series(
+            [1, 2, 2],
+            index=pd.MultiIndex.from_tuples(
+                [("a", "x"), ("a", "y"), ("c", "z")], names=["1", "2"]
+            ),
+            name="a",
+        )
+
+        result = getattr(ds.groupby(ds).rolling(2), func)()
+        expected = Series(
+            [np.nan, np.nan, 2.0],
+            index=pd.MultiIndex.from_tuples(
+                [(1, "a", "x"), (2, "a", "y"), (2, "c", "z")], names=["a", "1", "2"]
+            ),
+            name="a",
+        )
+        tm.assert_series_equal(result, expected)