jcapde / biterm Goto Github PK

IndexError Traceback (most recent call last)
in ()
86 pbd_names = [pbd(doc, True) for doc in btmp]
87
---> 88 Nz, Nwz, Z = gibbs_sampler_LDA(It=20, V=V, B=B, num_topics=num_topics, b=b, alpha=alpha, beta=beta)
89
90 topics = [[dictionary[ident] for ident in np.argsort(-Nwz[:,k])[0:10]] for k in range(num_topics)]

in gibbs_sampler_LDA(It, V, B, num_topics, b, alpha, beta)
28 Nzold = np.copy(Nz)
29 for ibi, bi in enumerate(b):
---> 30 Nwz[bi[0], Z[ibi]] -= 1
31 Nwz[bi[1], Z[ibi]] -= 1
32 Nz[Z[ibi]] -= 1

IndexError: only integers, slices (:), ellipsis (...), numpy.newaxis (None) and integer or boolean arrays are valid indices