maclandrol/FisherExact

_fisher_sim fact variable computation uses an obsolete numpy dtype

Closed this issue · 2 comments

Line 248

fact = np.zeros(wkslimit + 1, dtype=np.float, order='F')

Should use a valid numpy float type

fact = np.zeros(wkslimit + 1, dtype=np.float32, order='F')

@LouisLalonde can you make a PR ?

close by #14