Add per_word_topics parameter to apply call

[SND96]

Solving the issue
per_word_topics for a corpus #908

[tmylk]

Thanks for the pr. Please add a line in changelog and a test

[SND96]

Can you please advise me on how to proceed to modify the tests.

[tmylk]

First please pass all the params through the apply call, not just per_word_topics

[tmylk]

The tests should be the same as for a single word, just modified to be for a corpus.

[SND96]

I passed all the params in the apply call.
return self._apply(corpus, minimum_probability = self.minimum_probability, minimum_phi_value = self.minimum_phi_value, per_word_topics = self.per_word_topics)

But when I ran this line
lda.per_word_topics(corpus, per_word_topics = True)

I got this error
AttributeError: 'LdaModel' object has no attribute 'per_word_topics'

[tmylk]

lda.per_word_topics calls a non-existing method per_word_topics

[SND96]

Sorry I ran this,
lda.get_document_topics(corpus, per_word_topics = True)

And this was the error I got

'LdaModel' object has no attribute 'minimum_phi_value'

[tmylk]

which line is it? please report the full error stack

[SND96]

I added the attributes minimum_phi_value and per_word_topics to the attributes of LdaModel class but then this was the error I got

In [1]: import logging
   ...: import unittest
   ...: import os
   ...: import os.path
   ...: import tempfile
   ...: import numbers
   ...: 
   ...: import six
   ...: import bumpy
   ...: import scipy.linalg
   ...: 
   ...: from gensim.corpora import mmcorpus, Dictionary
   ...: from gensim.models import ldamodel, ldamulticore
   ...: from gensim import matutils
   ...: 
/Users/sahithdambekodi/Desktop/RaRe/gensim/gensim/utils.py:1014: UserWarning: Pattern library is not installed, lemmatization won't be available.
  warnings.warn("Pattern library is not installed, lemmatization won't be available.")
Slow version of gensim.models.word2vec is being used
Slow version of gensim.models.doc2vec is being used

In [2]: texts = [['human', 'interface', 'computer'],
   ...:  ['survey', 'user', 'computer', 'system', 'response', 'time'],
   ...:  ['eps', 'user', 'interface', 'system'],
   ...:  ['system', 'human', 'system', 'eps'],
   ...:  ['user', 'response', 'time'],
   ...:  ['trees'],
   ...:  ['graph', 'trees'],
   ...:  ['graph', 'minors', 'trees'],
   ...:  ['graph', 'minors', 'survey']]
   ...: dictionary = Dictionary(texts)
   ...: corpus = [dictionary.doc2bow(text) for text in texts]
   ...: 

In [3]: from gensim import models

In [4]: lda = models.LdaModel(corpus)

In [5]: lda.get_document_topic(corpus)
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-5-e2a60ca3cb20> in <module>()
----> 1 lda.get_document_topic(corpus)

AttributeError: 'LdaModel' object has no attribute 'get_document_topic'

In [6]: lda.get_document_topics(corpus)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-6-74e9bbe4f911> in <module>()
----> 1 lda.get_document_topics(corpus)

/Users/sahithdambekodi/Desktop/RaRe/gensim/gensim/models/ldamodel.py in get_document_topics(self, bow, minimum_probability, minimum_phi_value, per_word_topics)
    923         is_corpus, corpus = utils.is_corpus(bow)
    924         if is_corpus:
--> 925             return self._apply(corpus, minimum_probability = self.minimum_probability, minimum_phi_value= self.minimum_phi_value, per_word_topics = self.per_word_topics)
    926 
    927         gamma, phis = self.inference([bow], collect_sstats=True)

TypeError: _apply() got an unexpected keyword argument 'minimum_probability'

[tmylk]

Almost done in #978