Nonuniform samples from `np.random.multinomial` on Windows

D2L Book
1

Using mxnet 1.6.0b20190926 on Windows (as recommended in D2L book) I am experiencing highly nonuniform samples from np.random.multinomial

from mxnet import np, npx
fair_probs = [1.0 / 6] * 6
np.random.multinomial(10, fair_probs, size=1)
Out[5]: array([[ 0,  0,  0,  0,  0, 10]], dtype=int64)
np.random.multinomial(100, fair_probs, size=1)
Out[6]: array([[  0,   0,   0,   0,   0, 100]], dtype=int64)
np.random.multinomial(10000, fair_probs, size=1)
Out[7]: array([[    0,     0,     0,     0,     0, 10000]], dtype=int64)
np.random.multinomial(10000, fair_probs, size=3)
Out[8]: 
array([[    0,     0,     0,     0,     0, 10000],
       [ 1610,     0,     0,     0,     0,  8390],
       [ 1684,  1710,     0,     0,     0,  6606]], dtype=int64)

Corresponding jupyter notebook gives the same result.

Note that the mxnet.ndarray.random.multinomial version behaves as expected.

I’ve found slightly similar issue on GitHub: https://github.com/apache/incubator-mxnet/issues/15383. But it was closed on Jul 16.