[
https://issues.apache.org/jira/browse/ARROW-6043?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16895517#comment-16895517
]
Antoine Pitrou edited comment on ARROW-6043 at 7/29/19 6:37 PM:
----------------------------------------------------------------
Actually, if you look on the C++ side (see `cpp/src/arrow/compare.h`), the
equality comparisons now take an optional `EqualOptions` where you can choose
to set `nans_equal` to true.
So I would suggest to also expose those parameters in Python. Perhaps like it
is done in CSV (see _csv.pyx).
was (Author: pitrou):
Actually, if look on the C++ side (see `cpp/src/arrow/compare.h`), the equality
comparisons now take an optional `EqualOptions` where you can choose to set
`nans_equal` to true.
So I would suggest to also expose those parameters in Python. Perhaps like it
is done in CSV (see _csv.pyx).
> [Python] Array equals returns incorrectly if NaNs are in arrays
> ---------------------------------------------------------------
>
> Key: ARROW-6043
> URL: https://issues.apache.org/jira/browse/ARROW-6043
> Project: Apache Arrow
> Issue Type: Bug
> Components: Python
> Affects Versions: 0.14.1
> Reporter: Keith Kraus
> Priority: Major
> Fix For: 1.0.0
>
>
> {code:python}
> import numpy as np
> import pyarrow as pa
> data = [0, 1, np.nan, None, 4]
> arr1 = pa.array(data)
> arr2 = pa.array(data)
> pa.Array.equals(arr1, arr2)
> {code}
> Unsure if this is expected behavior, but in Arrow 0.12.1 this returned `True`
> as compared to `False` in 0.14.1.
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
