You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Joseph K. Bradley (JIRA)" <ji...@apache.org> on 2016/04/26 22:06:12 UTC

[jira] [Commented] (SPARK-12301) Remove final from classes in spark.ml trees and ensembles where possible

    [ https://issues.apache.org/jira/browse/SPARK-12301?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15258826#comment-15258826 ] 

Joseph K. Bradley commented on SPARK-12301:
-------------------------------------------

Update: I'm going to remove the uses of final.  However, I am leaving the constructors as {{private[ml]}} since we may need to refactor how the stats are maintained in tree nodes.  Users can still extend trees within the o.a.s.ml package.

> Remove final from classes in spark.ml trees and ensembles where possible
> ------------------------------------------------------------------------
>
>                 Key: SPARK-12301
>                 URL: https://issues.apache.org/jira/browse/SPARK-12301
>             Project: Spark
>          Issue Type: Improvement
>          Components: ML
>            Reporter: Joseph K. Bradley
>            Assignee: Joseph K. Bradley
>
> There have been continuing requests (e.g., [SPARK-7131]) for allowing users to extend and modify MLlib models and algorithms.
> I want this to happen for the next release.  For GBT, this may need to wait on some refactoring (to move the implementation to spark.ml).  But it could be done for trees already.  This will be broken into subtasks.
> If you are a user who needs these changes, please comment here about what specifically needs to be modified for your use case.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org