ufunc cannot be cast with rule `same_kind`

Updgrading to 1.10, I noticed the following new behavior:

In [1]: import numpy as np

In [2]: a = np.array([1,2,3])

In [3]: b = 2.0

In [4]: a /= b
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-4-b9f4626d3ca8> in <module>()
----> 1 a /= b

TypeError: ufunc 'true_divide' output (typecode 'd') could not be coerced to provided output parameter (typecode 'l') according to the casting rule ''same_kind''

In [5]: a = a/b

In [6]: a
Out[6]: array([ 0.5,  1. ,  1.5])

See also https://github.com/nipy/dipy/issues/730. Is this an intended new behavior? If so, what does it portend (what should I stop doing?)?

[charris]

Probably means you didn’t see the deprecation warnings since forever 😉 If you want to cast the float result to integer, you can call true_divide with out=a and casting='unsafe' arguments.

[siryne]

I got this error while executing this line : X = preprocess_input(X, mode=‘tf’)

I fixed this error by changing tools versions : Numpy – 1.16.1 Pandas – 0.23.0 Matplotlib – 2.2.2 Keras – 2.2.4

hope this would help.