comply to PEP8 about gensim

[piskvorky]

Scheduled for 0.8.0.

I'll probably add some wrappers to print deprecated warning when using the old names, together with hint how to fix it. To be completely removed in the next version (prolly 0.8.1).

[piskvorky]

Changed all function/variable names, commit 361003c3772cba677a80a57e0a7c3e648f42f6e0 .

There is no deprecated warnings, the changes weren't that many plus they're very straightforward (numTopics=>num_topics, addDocuments=>add_documents etc.).

[piskvorky]

reopening (forgot to change and test documentation as well :F )

[piskvorky]

Documentation updated in commit 88f2a3b .

All of gensim is now proper PEP8!

[mcobzarenco]

This is not the case anymore, (e.g. pyflake shows 63 warnings for models/wrappers/fasttext.py). Happy to PEP8 the whole repo and create a pr if that's cool.

[piskvorky]

@mcobzarenco what warnings are these? For example, we care more about readability than a strict 80-char-line limit.

But I also noticed some lapses (vertical indent in word2vec). @menshikh-iv what's the status on code style consistency in gensim?

[menshikh-iv]

@piskvorky

(math) ivan@P50:~/release/gensim$ git log | head -n5
commit e92b45d3f83ad670da9cf6ae20ae86a2a1c8558c
Author: jodevak <[email protected]>
Date:   Thu Oct 19 09:37:41 2017 +0300

    Add build_vocab_from_freq to Word2Vec, speedup scan_vocab (#1599)
(math) ivan@P50:~/release/gensim$ find gensim/ -name "*.py" | xargs flake8 --ignore=E501,E731,E12,W503,E402 --show-source
gensim/test/test_lsimodel.py:95:5: E303 too many blank lines (2)
    def testCorpusTransform(self):
    ^

[mcobzarenco]

Right, fair enough, all of the warnings were of the types you ignore above. Do you think we could we add those ignored rules to setup.cfg so they're documented and also ignored by default? I created a PR here #1636