无法运行Gensim LDA

Question

I'm trying to run the distributed LDA example as described here:

https://radimrehurek.com/gensim/dist_lda.html

I created the set of documents by following the tutorial here:

https://radimrehurek.com/gensim/dist_lsi.html

by "inflat[ing] the corpus to 1M documents, by repeating its documents over&over" as it suggests

I'm using python 3.3 and numpy 1.9.2 I keep getting the following error:

Exception in thread oneway-call:
Traceback (most recent call last):
  File "/usr/lib64/python3.3/threading.py", line 901, in _bootstrap_inner
    self.run()
  File "/usr/lib/python3.3/site-packages/Pyro4/core.py", line 1484, in run
    super(_OnewayCallThread, self).run()
  File "/usr/lib64/python3.3/threading.py", line 858, in run
    self._target(*self._args, **self._kwargs)
  File "/usr/lib64/python3.3/site-packages/gensim/models/lda_worker.py", line 71, in requestjob
    self.processjob(job)
  File "/usr/lib64/python3.3/site-packages/gensim/utils.py", line 98, in _synchronizer
    result = func(self, *args, **kwargs)
  File "/usr/lib64/python3.3/site-packages/gensim/models/lda_worker.py", line 80, in processjob
    self.model.do_estep(job)
  File "/usr/lib64/python3.3/site-packages/gensim/models/ldamodel.py", line 480, in do_estep
    gamma, sstats = self.inference(chunk, collect_sstats=True)
  File "/usr/lib64/python3.3/site-packages/gensim/models/ldamodel.py", line 423, in inference
    if doc and not isinstance(doc[0][0], six.integer_types):
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

I ran the distributed lsi example and it ran fine, but for some reason I couldn't seem to get the lda to work.

I tried changing line 423 in /usr/lib64/python3.3/site-packages/gensim/models/ldamodel.py to:

if doc is not None and not isinstance(doc[0][0], six.integer_types):

The error went away, but I got a warning that

FutureWarning: comparison to `None` will result in an elementwise object comparison in the future.

Could someone explain what I've done wrong? Is my change to this file correct? Or should I be running LDA differently?

Answer 1

This a bug in gensim which is raised here .

EDIT - This issue is now resolved in this pull request .