[pypy-issue] [issue1674] broken mask indexing / selection

[remi]

New submission from remi <remi2402 at free.fr>:

>>> import numpy as np
>>> data = np.arange(15).reshape(3, 5) * 1.0
>>> data
array([[  0.,   1.,   2.,   3.,   4.],
       [  5.,   6.,   7.,   8.,   9.],
       [ 10.,  11.,  12.,  13.,  14.]])
>>> m = data[:, 0] % 2 == 0
>>> m
array([ True, False,  True], dtype=bool)

on CPython 2.7 with numpy 1.6.2 (debian stable/wheezy)
>>> data[m, 0]
array([  0.,  10.])

on PyPy / Python 2.7.3 (25070f2267c4, Jan 13 2014, 14:17:42)
>>>> data[m, 0]
array([ 5.,  0.,  5.])

----------
assignedto: r.lobb
messages: 6485
nosy: pypy-issue, r.lobb, remi
priority: bug
status: unread
title: broken mask indexing / selection

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1674>
________________________________________