BUG: Maximum call stack size exceeded when indexing

[memeplex]

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

a = pd.DataFrame(dict(a=[1,2,3], b=[4,5,6], c=[2, 3, 4])).set_index(["a", "b"])
b = pd.DataFrame(dict(a=[1,2,3], b=[4,5,6]))
a.loc[pd.MultiIndex.from_frame(b)]  # case 1
a.reindex(pd.MultiIndex.from_frame(b))  # case 2

Issue Description

The above example works for both cases 1 and 2. But using a larger dataframe (with proprietary data) the first one yields:

Error loading preloads:
Maximum call stack size exceeded

while the second one still works. Sorry for not providing a reproducible example but maybe you can hint me in the right direction in order to reproduce it. Any hunch?

Expected Behavior

Both cases are implemented more or less the same and consistently succeed (or fail).

Installed Versions

INSTALLED VERSIONS

commit : 4bfe3d0
python : 3.9.12.final.0
python-bits : 64
OS : Darwin
OS-release : 20.6.0
Version : Darwin Kernel Version 20.6.0: Mon Aug 30 06:12:21 PDT 2021; root:xnu-7195.141.6~3/RELEASE_X86_64
machine : x86_64
processor : i386
byteorder : little
LC_ALL : en_US.UTF-8
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.4.2
numpy : 1.22.3
pytz : 2022.1
dateutil : 2.8.2
pip : 22.0.4
setuptools : 60.10.0
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : 2.9.3
jinja2 : 3.1.2
IPython : 8.3.0
pandas_datareader: None
bs4 : None
bottleneck : None
brotli : None
fastparquet : None
fsspec : None
gcsfs : None
markupsafe : 2.1.1
matplotlib : 3.5.2
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : 1.8.0
snappy : None
sqlalchemy : 1.4.36
tables : None
tabulate : None
xarray : None
xlrd : None
xlwt : None
zstandard : None

[simonjayhawkins]

Thanks @memeplex for the report.

Sorry for not providing a reproducible example but maybe you can hint me in the right direction in order to reproduce it. Any hunch?

can you post the stacktrace? not much to go on otherwise.

[memeplex]

Ok, the problem here was in the error reporting of vscode-jupyter, not in pandas. I was able to reproduce it with this code:

df = pd.DataFrame(
    dict(
        a=np.arange(100),
        b=np.arange(100),
        c=np.arange(100),
        d=np.arange(100),
        e=np.arange(100),
        f=['a'] * 50 + ['b'] * 50,
    ),
    dtype="category",
)

s = df[df.f == 'a'].groupby(["a", "b", "c", "d", "e"]).apply(lambda df: 1)
s.loc[pd.MultiIndex.from_frame(df[["a", "b", "c", "d", "e"]])]

This is representative of what I was attempting, the issue here is that some keys won't be in the groupby result because it's restricted to df.f == 'a', here reindex succeeds where loc[] fails, as expected. But given that the real frame is huge, vscode fails to properly report the error and shows the message about stack overflow instead.