BUG: GroupByRolling duplicates MultiIndex levels, which prevents assignment on parent

[drmarshall]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

import pandas as pd

df = pd.DataFrame({
    'x': {('a', i): i ** 2 for i in range(5)}
})

print(df.x.groupby(level=0).rolling(2).sum().index)

'''
MultiIndex([('a', 'a', 0),
            ('a', 'a', 1),
            ('a', 'a', 2),
            ('a', 'a', 3),
            ('a', 'a', 4),
           )
'''

# raises: AssertionError: Length of new_levels (3) must be <= self.nlevels (2)

df['new_col'] = df.x.groupby(level=0).rolling(2).sum()

Issue Description

When performing a groupby.rolling using a level parameter, an additional level(s) are inserted into the MultiIndex. The result is an index incompatible with original frame (raises: AssertionError: Length of new_levels (3) must be <= self.nlevels (2))

Expected Behavior

Groupby(level=n).rolling should return compatible MultiIndex for assignment to derived DataFrame, to allow ergonomics such as:

df['new_col'] = df.x.groupby(level=0).rolling(2).sum()

This behavior should be roughly the same as grouping by column(s) string names:

print(df.reset_index().groupby('level_0').x.rolling(2).sum().index)

'''
MultiIndex([('a', 0),
            ('a', 1),
            ('a', 2),
            ('a', 3),
            ('a', 4)],
           names=['level_0', None])
'''

# this works!
df['new_col'] = df.reset_index().groupby('level_0').x.rolling(2).sum()

Totally possible there is a specific difference in grouping by named columns versus using level param that I do not understand; but I have always understood them to have been syntactic sugar for functionally equivalent operations. Happy to submit a documentation update PR if this is my misunderstanding. 🙏

Installed Versions

INSTALLED VERSIONS
------------------
commit                : d9cdd2ee5a58015ef6f4d15c7226110c9aab8140
python                : 3.11.8.final.0
python-bits           : 64
OS                    : Darwin
OS-release            : 23.1.0
Version               : Darwin Kernel Version 23.1.0: Mon Oct  9 21:27:24 PDT 2023; root:xnu-10002.41.9~6/RELEASE_ARM64_T6000
machine               : arm64
processor             : arm
byteorder             : little
LC_ALL                : None
LANG                  : None
LOCALE                : None.UTF-8

pandas                : 2.2.2
numpy                 : 1.26.3
pytz                  : 2023.3.post1
dateutil              : 2.8.2
setuptools            : 68.0.0
pip                   : 23.2.1
Cython                : None
pytest                : 7.4.0
hypothesis            : None
...
zstandard             : 0.19.0
tzdata                : 2023.3
qtpy                  : 2.4.1
pyqt5                 : None

[rhshadrach]

Thanks for the report! I think this is a duplicate of #51751, which generally speaking, is a discussion on how to treat the index on a rolling operation. Can you take a look at that issue and provide any feedback there?

Closing as a duplicate for now.