BUG: KeyError is raised on .loc for MultiIndex if single keys exist but not the combination

[fwitte]

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


idx = pd.IndexSlice

index = pd.MultiIndex.from_tuples(
    [("a", "b", "c"), ("a", "a", "c"), ("a", "c", "b")],
    names=["col1", "col2", "col3"]
)

df = pd.DataFrame(index=index, data={"value": [1, 2, 3]})

# works and returns empty dataframe (for every key, that is not in the level "col2")
df.loc[idx[:, ["a"], ["somekey"]]]
# raises KeyError (for every key given for level "col2" that is available in that level, but not in the given combination of keys.
# If the above is itended (returning empty DataFrame), then this should return an empty DataFrame as well
df.loc[idx[:, ["a"], ["b"]]]

Issue Description

See reproducible example.

To fix that the lines

pandas/pandas/core/indexes/multi.py

Lines 3290 to 3291 in c35eca3

	if not np.any(indexer) and np.any(lvl_indexer):
	raise KeyError(seq)

should not raise an Exception but return

np.array([], dtype=np.intp)

like when an item cannot be found in the current index level:

pandas/pandas/core/indexes/multi.py

Lines 3269 to 3272 in c35eca3

	if lvl_indexer is None:
	# no matches we are done
	# test_loc_getitem_duplicates_multiindex_empty_indexer
	return np.array([], dtype=np.intp)

Expected Behavior

return an empty dataframe, if the combination of keys in the levels is not available.

>>> df.loc[idx[:, ["a"], ["b"]]]

Empty DataFrame
Columns: [value]
Index: []

Installed Versions

INSTALLED VERSIONS

commit : 8dab54d
python : 3.9.13.final.0
python-bits : 64
OS : Windows
OS-release : 10
Version : 10.0.19044
machine : AMD64
processor : Intel64 Family 6 Model 154 Stepping 3, GenuineIntel
byteorder : little
LC_ALL : None
LANG : de_DE.UTF-8
LOCALE : de_DE.cp1252

pandas : 1.5.2
numpy : 1.23.2
pytz : 2022.2.1
dateutil : 2.8.2
setuptools : 65.3.0
pip : 22.2.2
Cython : None
pytest : 7.1.2
hypothesis : None
sphinx : 5.1.1
blosc : None
feather : None
xlsxwriter : 3.0.3
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 3.1.2
IPython : 8.4.0
pandas_datareader: None
bs4 : 4.11.1
bottleneck : None
brotli : None
fastparquet : None
fsspec : None
gcsfs : None
matplotlib : 3.5.3
numba : None
numexpr : 2.8.3
odfpy : None
openpyxl : 3.0.10
pandas_gbq : None
pyarrow : None
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : 1.9.1
snappy : None
sqlalchemy : None
tables : 3.7.0
tabulate : 0.8.10
xarray : 2022.6.0
xlrd : None
xlwt : None
zstandard : None
tzdata : None

[fwitte]

My bad, I confused the pandas installations in my environment and the clone repository. I guess #42351 dealt with it, although I'd have preferred the other way round :(.

Thanks and have a nice weekend!

[phofl]

Closing then