most_similar_cosmul bug fix

[dkim010]

remove unhashable type: 'numpy.ndarray' error

Signed-off-by: Dongwon Kim dongwon.kim@navercorp.com

[tmylk]

Thank you for the quick fix!

model.most_similar_cosmul('graph', topn=10) works but model.most_similar_cosmul(positive=[graph_vector], topn=11) still produces results TypeError: unhashable type: 'numpy.ndarray'.

Please add this unit test to test_word2vec.py

def test_cosmul():
        model = word2vec.Word2Vec(sentences, size=2, min_count=1, hs=1, negative=0)
        sims = model.most_similar_cosmul('graph', topn=10)
        # self.assertTrue(sims[0][0] == 'trees', sims)  # most similar

        # test querying for "most similar" by vector
        graph_vector = model.wv.syn0norm[model.wv.vocab['graph'].index]
        sims2 = model.most_similar_cosmul(positive=[graph_vector], topn=11)
        sims2 = [(w, sim) for w, sim in sims2 if w != 'graph']  # ignore 'graph' itself
        self.assertEqual(sims, sims2)

[dkim010]

Thank you for the quick reply!

I think that most_similar_cosmul() method allows only the string (or a string list) as the positive/negative parameters.
The method is related to the most_similar() method and both methods cannot pass the vectors of the words.
Is my idea wrong?

[tmylk]

most_similar works with vectors and most_similar_cosmul is expected to work too https://github.com/RaRe-Technologies/gensim/blob/develop/gensim/test/test_word2vec.py#L360

[dkim010]

Okay, I'll fix it.

[dkim010]

I fixed it. most_similar_cosmul works with vectors!
I refered most_similar.

[tmylk]

Thanks for the quick fix!