on master, pd.show_versions() emits a FutureWarning

[Dr-Irv]

Code Sample, a copy-pastable example if possible

import pandas as  pd
pd.show_versions()

Problem description

Gets this output:

C:\Code\pandas_dev\pandas36\pandas\core\index.py:29: FutureWarning: pandas.core.index is deprecated and will be removed in a future version.  The public classes are available in the top-level namespace.
  FutureWarning,

INSTALLED VERSIONS
------------------

followed by all the version stuff

Expected Output

No FutureWarning

Output of pd.show_versions()

C:\Code\pandas_dev\pandas36\pandas\core\index.py:29: FutureWarning: pandas.core.index is deprecated and will be removed in a future version. The public classes are available in the top-level namespace.
FutureWarning,

INSTALLED VERSIONS

commit : 7cce2c7
python : 3.6.7.final.0
python-bits : 64
OS : Windows
OS-release : 10
machine : AMD64
processor : Intel64 Family 6 Model 158 Stepping 13, GenuineIntel
byteorder : little
LC_ALL : None
LANG : None
LOCALE : None.None

pandas : 0.26.0.dev0+1739.g7cce2c7f4
numpy : 1.15.4
pytz : 2019.3
dateutil : 2.8.1
pip : 19.3.1
setuptools : 44.0.0.post20200102
Cython : 0.29.14
pytest : 5.3.2
hypothesis : 5.1.1
sphinx : None
blosc : None
feather : None
xlsxwriter : 1.2.7
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 2.10.3
IPython : 7.11.1
pandas_datareader: None
bs4 : None
bottleneck : 1.3.1
fastparquet : 0.3.2
gcsfs : None
lxml.etree : None
matplotlib : 3.0.2
numexpr : 2.7.1
odfpy : None
openpyxl : 3.0.1
pandas_gbq : None
pyarrow : 0.14.1
pytables : None
pytest : 5.3.2
s3fs : None
scipy : 1.3.1
sqlalchemy : None
tables : 3.6.1
tabulate : None
xarray : None
xlrd : 1.2.0
xlwt : 1.3.0
xlsxwriter : 1.2.7
numba : 0.46.0

[Dr-Irv]

This seems to be due to a dependency on fastparquet

Python 3.7.6 | packaged by conda-forge | (default, Dec 26 2019, 23:30:11) [MSC v.1916 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import fastparquet
c:\Code\pandas_dev\pandas\pandas\core\index.py:29: FutureWarning: pandas.core.index is deprecated and will be removed in a future version.  The public classes are available in the top-level namespace.
  FutureWarning,

@jorisvandenbossche I think you mentioned you were working on some parquet stuff, so maybe you are aware of this

[simonjayhawkins]

I've added the milestone, since I think this has only just appeared. (Can't be sure without checking)

[jorisvandenbossche]

That's up to fastparquet to fix, not pandas (we are missing a stacklevel to make it clearer where it is coming from, though).

Anybody wants to open an issue on the fastparquet tracker?

[Dr-Irv]

@jorisvandenbossche you opened the issue about a month ago: dask/fastparquet#470

Having said that, since we ask people to use pd.show_versions() when reporting issues, do we want to suppress any such warnings in that function to avoid people getting confused by the message?

[jorisvandenbossche]

@jorisvandenbossche you opened the issue about a month ago

Ah, yes ;-)

I don't have a strong opinion on filtering it in show_versions. On the one hand, it also signals problems (as you noticed now), on the other hand it gives warnings users can't do nothing about.

[Dr-Irv]

@jorisvandenbossche I would suggest the following:

• With pd.show_versions(), we don't emit warnings
• With pd.show_versions(show_warnings=True), we then emit the warnings

My sense is that the default should not emit the warnings so we don't get people wondering what is going on when they report issues.

Separately, I'm guessing we could do some kind of regular thing in CI to do pd.show_versions(show_warnings=True) and catch the warnings, and then we know which projects to raise issues in so they update their usage of pandas.

Also, given our discussion here, I've reopened this issue (you had closed it) so we can settle this point.

[TomAugspurger]

I think we should just handle this in fastparquet. I'll make a PR.

[Dr-Irv]

@TomAugspurger What if we start getting warnings from other libraries in the future as we deprecate more things in pandas? Shouldn't we do something in pd.show_versions()?