Github user jkbradley commented on the pull request:
https://github.com/apache/spark/pull/4419#issuecomment-88995458
@hhbyyh Thanks for the results! It looks like it's doing something
reasonable, though of course it's always hard to tell. How hard would it be
for you to compare your implementation with
[https://github.com/Blei-Lab/onlineldavb] (or some other Online LDA VB code) in
a deterministic way? If both could be run for the same number of iterations
and use the full dataset on each iteration, then they should produce identical
results. That would be a great verification if possible.
Also, let us know on the linked PR above if you have thoughts about the API
updates being made there. I just posted my thoughts at the bottom of that PR.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
