BUG: pandas.DataFrame.index.map() works differently if debugpy debugger is attached

[ember91]

• 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 master branch of pandas.

Reproducible Example

import pandas as pd


def map_func(dt):
    if dt.month <= 2:
        return dt.year + dt.month - 1
    return dt.year + dt.month


dict = {
    "Name": ["Tom", "Joseph"],
    "Date": ["2021-10-01", "2021-11-01"],
}
df = pd.DataFrame(dict)
df = df.set_index("Date")
df.index = pd.to_datetime(df.index)
print(f"Result: {df.index.map(map_func)}")

Issue Description

Running the attached program in debugpy debugger produces different results than if I run it plainly.

Running it plainly with /usr/bin/env /usr/bin/python3.9 test.py produces output as expected:
Result: Int64Index([2031, 2032], dtype='int64', name='Date')

Running it in debugpy with:
cd $USER/test_error ; /usr/bin/env /usr/bin/python3.9 $USER/.vscode/extensions/ms-python.python-2021.10.1317843341/pythonFiles/lib/python/debugpy/launcher 37319 -- $USER/test_error/test.py
produces a crash on line 5. For some reason dt.month is the full index:
DatetimeIndex(['2021-10-01', '2021-11-01'], dtype='datetime64[ns]', name='Date', freq=None)
and not only a single date.

Note that the debugpy command above is the auto generated command when using the VS code debugger (Run -> Start Debugging, or F5). Note that it also assumes that you are in the '$USER/test_error' directory.

VS code version is 1.61.0 and Python extension version is v2021.10.1317843341.

I'm posting it here, but I'm not sure if it's a pandas, debugger or python issue.

Expected Behavior

No crash when running in debugpy, or at least the same result

Installed Versions

INSTALLED VERSIONS

commit : 73c6825
python : 3.9.6.final.0
python-bits : 64
OS : Linux
OS-release : 5.11.0-37-generic
Version : #41~20.04.2-Ubuntu SMP Fri Sep 24 09:06:38 UTC 2021
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.3.3
numpy : 1.21.2
pytz : 2021.3
dateutil : 2.8.2
pip : 21.2.4
setuptools : 58.2.0
Cython : 0.29.24
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : 3.0.1
lxml.etree : 4.6.3
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 3.0.2
IPython : None
pandas_datareader: None
bs4 : 4.10.0
bottleneck : 1.3.2
fsspec : None
fastparquet : None
gcsfs : None
matplotlib : 3.4.3
numexpr : 2.7.3
odfpy : None
openpyxl : 3.0.9
pandas_gbq : None
pyarrow : None
pyxlsb : None
s3fs : None
scipy : 1.7.1
sqlalchemy : None
tables : 3.6.1
tabulate : None
xarray : None
xlrd : None
xlwt : None
numba : None
None

[mzeitlin11]

Thanks for reporting this @ember91! Can't reproduce this with the pycharm debugger or pdb - can others reproduce? @ember91 best way to distinguish between a pandas issue or debugpy issue would be if you can step through your code and see exactly where the behaviors start to diverge

[ember91]

So I ran it with pdb as well. Breaking at line 5 (break test.py:5 in pdb), both debuggers show that dt contains DatetimeIndex(['2021-10-01', '2021-11-01'], dtype='datetime64[ns]', name='Date', freq=None). Next step, debugpy will crash, since dt isn't a singular date, but pdb will jump out of the function and go to pandas/core/indexes/extension.py(80)fget().

So by looking in the source code in extension.py I see the comment:
Try to run function on index first, and then on elements of index

[mroeschke]

Since this issue is not immediately reproducible with more standard debuggers and not immediately obvious that this is a pandas issue, closing. If the debugpy folks can pin this issue to pandas can reopen.

[ember91]

See debugpy bug report at microsoft/debugpy#775.