Github user jkbradley commented on the pull request:
https://github.com/apache/spark/pull/9513#issuecomment-155225172
Updated per all comments, including changing optimizer to be a String
Param. Ready for final review.
The only remaining issue to debate is whether to make LDAModel abstract. I
agree with @feynmanliang that it's cleaner in terms of implementing a lazy val.
The main cons I see are:
* Will users be confused about getting an abstract type when they look at
the docs? Probably not, but worth mentioning.
* I do not want us to require beginners users to cast to LocalLDAModel or
DistributedLDAModel. That means we would need to make sure not to add
functionality to LocalLDAModel without requiring it in the abstraction LDAModel.
If we decide to make this change, we can do it in a follow-up PR. Feel
free to discuss below.
---
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]
