ValueError with numpy 1.16

Hi,

Our PRE release bots, which using Numpy 1.16, failed many times with this error: ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() (see https://github.com/nipy/dipy/issues/1579)

Here a small example to reproduce our problem:

simple_mp.pyx file:

import numpy as np
import cython

def masking_test():
	simple_wm = np.array([[0, 0, 0, 0, 0, 0],
                          [0, 0, 1, 0, 0, 0],
                          [0, 1, 1, 1, 0, 0],
                          [0, 1, 1, 1, 0, 0],
                          [0, 0, 0, 0, 0, 0]])
	mask = np.zeros(simple_wm.shape)

	simple_wm[mask==0] = 1

Test in a Python terminal after compiling:

>>> import simple_mp as m
>>> m.masking_test()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "simple_mp.pyx", line 13, in simple_mp.masking_test
    simple_wm[mask==0] = 1
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

Python 3.6 cython 0.29a0 numpy 1.16.0.dev0+2804c03

How can I resolve this problem? Thanks

[skoudoro]

Yes, it works with Cython 0.28.3 and numpy 1.16.0.dev0+2804c03.

[gabrieldemarmiesse]

My apologies. You are totally right. I managed to reproduce the error with python 3.6, cython master and numpy 1.14.5. I’m going to investigate and see if an earlier version of cython or numpy does the trick.