Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Some cupy.random distribution does not support `size=None`

See original GitHub issue

size=None is different from size=().

>>> np.random.uniform([2], [[3]])
array([[2.63150331]])
>>> np.random.uniform([2], [[3]], size=())
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "mtrand.pyx", line 1317, in mtrand.RandomState.uniform
  File "mtrand.pyx", line 267, in mtrand.cont2_array
ValueError: size is not compatible with inputs
>>> 
>>> cupy.random.uniform([2], [[3]])
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/kataoka/cupy/cupy/random/distributions.py", line 454, in uniform
    return rs.uniform(low, high, size=size, dtype=dtype)
  File "/home/kataoka/cupy/cupy/random/generator.py", line 484, in uniform
    return RandomState._scale_kernel(low, high, rand)
  File "cupy/core/_kernel.pyx", line 532, in cupy.core._kernel.ElementwiseKernel.__call__
    out_args = _get_out_args_with_params(
  File "cupy/core/_kernel.pyx", line 368, in cupy.core._kernel._get_out_args_with_params
    raise ValueError('Out shape is mismatched')
ValueError: Out shape is mismatched