LDA u_mass coherence calculation triggers numpy deprecation warning

[masscognition]

Using Python 2.7.11, latest gensim , latest numpy.

Line 41 of segmentation.py:

for w_star in top_words[:w_prime_index]:

triggers this warning from numpy:

DeprecationWarning: converting an array with ndim > 0 to an index will result in an error in the future

fix is to cast w_prime_index to int

Also via Devashish:

"the comment right next to

w_prime_index = np.where(top_words == w_prime)[0] should be # To get index of w_prime in top_words (apologies for my mistake)."

[devashishd12]

Thanks for raising this issue! Would you mind raising a PR fix for this? You can follow a workflow similar to that given here: http://scikit-learn.org/stable/developers/contributing.html#contributing-code. Would be happy to help if you encounter any problems.

[devashishd12]

Sorry @masscognition I addressed this issue in #793 since I happened to be working on the same thing. Thanks for raising the issue though!

[masscognition]

No problem :-)

On Mon, Jul 18, 2016 at 2:35 PM, Devashish Deshpande <
notifications@github.com> wrote:

Sorry @masscognition https://github.com/masscognition I addressed this
issue in #793 #793
since I happened to be working on the same thing. Thanks for raising the
issue though!

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#765 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ANdXmh09dNhOftQV_VZFXA7fDHjDPNgsks5qW8eEgaJpZM4JBlYf
.