[jira] [Created] (ARROW-12431) [Python] pa.array mask inverted when type is binary and value to be converted in numpy array

["Daniel Nugent (Jira)" <ji...@apache.org>]

Daniel Nugent created ARROW-12431:
-------------------------------------

             Summary: [Python] pa.array mask inverted when type is binary and value to be converted in numpy array
                 Key: ARROW-12431
                 URL: https://issues.apache.org/jira/browse/ARROW-12431
             Project: Apache Arrow
          Issue Type: Bug
            Reporter: Daniel Nugent


{code:python}
Python 3.9.2 | packaged by conda-forge | (default, Feb 21 2021, 05:02:46)                                   
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> import pyarrow as pa
>>>
>>> pa.array(np.array([b'\x00']),type=pa.binary(1), mask = np.array([False]))
<pyarrow.lib.FixedSizeBinaryArray object at 0x7fa080ca3640>
[
  null
]
>>> pa.array(np.array([b'\x00']),type=pa.binary(1), mask = np.array([True]))
<pyarrow.lib.FixedSizeBinaryArray object at 0x7fa080ca3700>
[
  00
]
>>> pa.array([b'\x00'],type=pa.binary(1), mask = np.array([False]))
<pyarrow.lib.FixedSizeBinaryArray object at 0x7fa083cc9520>
[
  00
]
>>> pa.__version__
'3.0.0'
>>> np.__version__
'1.20.1'
{code}

Happens both with FixedSizeBinary and variable sized binary (I was working with FixedSizeBinary). Does not happen for integers (presumably other types, didn't exhaustively check)?



--
This message was sent by Atlassian Jira
(v8.3.4#803005)