Github user manishamde commented on the pull request:
https://github.com/apache/spark/pull/2595#issuecomment-58553154
@jkbradley Agree on the RF and boosting parameters.
Should we keep the treeAggregate option as an option in the tree code
instead of reviving it later? Boosting is coming up soon and lot of folks use
shallow trees for exploratory purposes before building deeper ones. The loss in
performance is greater than 20% in some cases. I agree with the conclusion to
do a distributed aggregate by default but we should probably consider having a
heuristic to use the treeAggregate when suitable.
---
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]
