Cast error for document-term matrix with binary logistic slda

[kwajiehao]

hi @burbma, @bearnshaw , thanks for the package, it's been super useful so far. I ran into some issues when trying using blslda.fit(). I used the sklearn package to generate the document-term matrix, but when I tried to pass it into the blslda.fit() function I got the error : "Cannot cast array data from dtype('float64') to dtype('int64') according to the rule 'safe'".

The full traceback is:

TypeError Traceback (most recent call last)
~/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py in _wrapfunc(obj, method, *args, **kwds)
51 result = wrap(result)
---> 52 return result
53

TypeError: Cannot cast array data from dtype('float64') to dtype('int64') according to the rule 'safe'

During handling of the above exception, another exception occurred:

TypeError Traceback (most recent call last)
in ()

~/anaconda3/lib/python3.6/site-packages/slda/topic_models.py in fit(self, X, y)
286 self.n_docs, self.n_terms = X.shape
287 self.n_tokens = X.sum()
--> 288 doc_lookup, term_lookup = self._create_lookups(X)
289 # iterate
290 self.theta, self.phi, self.eta, self.loglikelihoods = gibbs_sampler_blslda(

~/anaconda3/lib/python3.6/site-packages/slda/topic_models.py in _create_lookups(self, X)
37 else:
38 x = X[docs, terms]
---> 39 doc_lookup = np.ascontiguousarray(np.repeat(docs, x), dtype=np.intc)
40 term_lookup = np.ascontiguousarray(np.repeat(terms, x), dtype=np.intc)
41 return doc_lookup, term_lookup

~/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py in repeat(a, repeats, axis)
421 mode : {'raise', 'wrap', 'clip'}, optional
422 Specifies how out-of-bounds indices will behave.
--> 423
424 * 'raise' -- raise an error (default)
425 * 'wrap' -- wrap around

~/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py in _wrapfunc(obj, method, *args, **kwds)
60 # such a method in its class.
61
---> 62 # A TypeError occurs if the object does have such a method
63 # in its class, but its signature is not identical to that
64 # of NumPy's. This situation has occurred in the case of

~/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py in _wrapit(obj, method, *args, **kwds)
40
41 # functions that are now methods
---> 42 def _wrapit(obj, method, *args, **kwds):
43 try:
44 wrap = obj.array_wrap

TypeError: Cannot cast array data from dtype('float64') to dtype('int64') according to the rule 'safe'

Thank you in advance!