line 208, in learn
mk = np.sqrt(nk**-1 - n**-1)
line 86, in learn
alphaplus = np.where(ynew==1, nplus**-1, 0)
The values for nk, n, nplus are all derived from y and therefore have the np.int dtype as well.
Is anyone else having this same issue? Is there a way that I can resolve this?
Thanks,
Devin
alphaplus = np.where(ynew==1, nplus**-1, 0)
you can change it to :
alphaplus = np.where(ynew==1, 1/nplus, 0)
mine: python3.5 numpy 0.18