Github user jkbradley commented on the issue:
https://github.com/apache/spark/pull/16623
@felixcheung Just saw this was backported to 2.1.1. Since this is a
fairly significant behavioral change, I recommend we revert this backport. I
could imagine workloads working with EM but failing with online LDA, and we
should be careful about having such failures in patch versions (2.1.0->2.1.1).
Let's wait until the next major release (2.2.0) instead.
I've also seen other PRs adding public APIs to SparkR for 2.1.1. I
understand that SparkR is still fairly experimental, but I recommend we start
treating it like other parts of Spark in terms of API stability. We can talk
about exceptions when they are really necessary.
Can you please revert this patch? Thanks!
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
