pandas.DataFrame.reindex_like throws if called from a derived class instance

[lguyot]

Code Sample, a copy-pastable example if possible

import pandas as pd
import pandas.testing as pdt

class MyDataFrame(pd.DataFrame):
    pass

mdf = MyDataFrame()
df = pd.DataFrame()
# In DataFrame.reindex_like, the derived class is compared to
# the base class. The following line will throw
# 'AssertionError: <class 'pandas.core.frame.DataFrame'>'
mdf.reindex_like(df)

# In pdt.assert_frame_equal,
# check_like=True triggers a call to DataFrame.reindex_like.
# The following line will throw the same
# 'AssertionError: <class 'pandas.core.frame.DataFrame'>'
pdt.assert_frame_equal(mdf, df, check_like=True)

Problem description

The method reindex_like throws an AssertionError when the caller is an instance of a derived class of pandas.DataFrame.

Expected Output

No exception thrown.

Output of pd.show_versions()

INSTALLED VERSIONS

commit : None
python : 3.6.9.final.0
python-bits : 64
OS : Linux
OS-release : 4.15.0-65-generic
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : fr_FR.UTF-8
LOCALE : fr_FR.UTF-8

pandas : 1.0.1
numpy : 1.18.1
pytz : 2019.3
dateutil : 2.8.1
pip : 9.0.1
setuptools : 39.0.1
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : 0.999999999
pymysql : None
psycopg2 : None
jinja2 : None
IPython : 7.12.0
pandas_datareader: None
bs4 : None
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : None
matplotlib : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pytables : None
pytest : None
pyxlsb : None
s3fs : None
scipy : 1.4.1
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
xlwt : None
xlsxwriter : None
numba : None

[jorisvandenbossche]

@topper-123 another one from the _ensure_type checking

[jbrockmendel]

@jorisvandenbossche sounds like there has been some discussion about ensure_type already. is someone on top of this?

[TomAugspurger]

sounds like there has been some discussion about ensure_type already

I think Joris was referring to #31785 / #31788.

Don't think anyone is working on this yet (unless @topper-123 has something locally?)

[ericchansen]

Isn't this as simple as using issubclass instead of isinstance?

pandas/pandas/core/base.py

Lines 90 to 98 in bdcb5da

	def _ensure_type(self: T, obj) -> T:
	"""
	Ensure that an object has same type as self.
	Used by type checkers.
	"""
	assert isinstance(obj, type(self)), type(obj)
	return obj

• Commit capturing proposed change ericchansen@f2c3eab
• Updated code https://github.com/ericchansen/pandas/blob/f2c3eabdfee735787a30eada5ffb808fe111d8c1/pandas/core/base.py#L90-L97
• Review of PR that introduced _ensure_type TYP: Add TypeVars to NDFrame #30613 (review)

[jorisvandenbossche]

@ericchansen do you want to make a PR with that change? (and a test)

I would personally also be fine with just getting rid of _ensure_type, the type checking should not have runtime consequences

[TomAugspurger]

the type checking should not have runtime consequences

Agreed.

[ericchansen]

@jorisvandenbossche I'll open the PR today once I'm done with the tests.

@TomAugspurger @jorisvandenbossche It looks as if the type checking has something to do with mypy (as per discussion here #30613 (review) and here #30565). I agree it shouldn't have runtime consequences, but it does look like there was some discussion about why they @topper-123 @jreback did it this way.