search

lda-project / lda

Topic modeling with latent Dirichlet allocation using Gibbs sampling
https://lda.readthedocs.io/
Mozilla Public License 2.0
1.24k stars 390 forks source link

add verbose and tolerance feature #95

Closed jmcamac closed 4 years ago

jmcamac commented 4 years ago

The verbose feature would allow a user to specify whether or not they want the log likelihood messages. Also, the tolerance feature would allow the user to end Gibbs sampling early - not execute all iterations - once a desired tolerance level has been reached. Here's a branch with those changes: https://github.com/jmcamac/lda/tree/develop/lda

ariddell commented 4 years ago

I'm more than happy to consider these changes. (They do seem like two distinct changes.)

When lda was written, there was not a standard method in the community for reporting progress in a (very) long-running calculation. If one has emerged (e.g., for scikit-learn or something similar), I'll happily adopt it.

You can disable the logging output using the logging module. If documentation doesn't describe how to do this, that's a bug.

ariddell commented 4 years ago

The logging setup is a mess. Sorry!

No developer resources to fix this currently.

WilliamBonvini commented 1 year ago

Hello there, how has this been solved?