Clarify equals method docstring

[gsganden]

Make explicit that equals method requires that columns have the same dtypes but not that indices have the same types (e.g. pd.DataFrame({1:[0], 0:[1]}).equals(pd.DataFrame({1.0:[0], 0.0:[1]})) returns True while pd.DataFrame({1:[0], 0:[1]}, dtype='float32').equals(pd.DataFrame({1:[0], 0:[1]}, dtype='float64')) returns False)

• closes #xxxx
• tests added / passed
• passes git diff upstream/master -u -- "*.py" | flake8 --diff
• whatsnew entry

[codecov]

Codecov Report

Merging #19213 into master will increase coverage by 0.02%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #19213      +/-   ##
==========================================
+ Coverage   91.53%   91.55%   +0.02%     
==========================================
  Files         147      147              
  Lines       48797    48797              
==========================================
+ Hits        44664    44675      +11     
+ Misses       4133     4122      -11

Flag	Coverage Δ
#multiple	89.92% <ø> (+0.02%)	⬆️
#single	41.6% <ø> (ø)	⬆️

Impacted Files	Coverage Δ
pandas/core/generic.py	95.9% <ø> (ø)	⬆️
pandas/core/dtypes/cast.py	88.35% <0%> (-0.31%)	⬇️
pandas/core/internals.py	94.5% <0%> (ø)	⬆️
pandas/core/dtypes/common.py	94.58% <0%> (+0.04%)	⬆️
pandas/plotting/_converter.py	66.95% <0%> (+1.73%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 8347ff8...3bc9bfd. Read the comment docs.

[gsganden]

Sorry, I meant "column labels" rather than "indices."

[jreback]

We check the indices for equality. But we happen to have this case.

In [7]: a
Out[7]: Int64Index([1, 0], dtype='int64')

In [8]: b
Out[8]: Float64Index([1.0, 0.0], dtype='float64')

In [9]: a.equals(b)
Out[9]: True

[jreback]

We normally DO consider that case equal, but I think we discussed in some other issues that from a user point of view these are not equal.

cc @toobaz @TomAugspurger @jorisvandenbossche

[toobaz]

Without having thought much about this, I would have said that the indexes case (can be discussed, but mostly) is fine, while pd.DataFrame({1:[0], 0:[1]}, dtype='float32').equals(pd.DataFrame({1:[0], 0:[1]}, dtype='float64')) should have returned True...

but assuming I'm wrong (i.e. current behavior is OK), then I just suggest replacing the same dtypes at corresponding locations with the same dtype(s) in the patch.

[datapythonista]

Closing this PR as discontinued. It's not in a state to be merged, and the user deleted the fork/branch it's based on.

Opened #22462 as I think it's better to start from the beginning on this docstring.