[
https://issues.apache.org/jira/browse/SPARK-7000?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14512828#comment-14512828
]
Joseph K. Bradley commented on SPARK-7000:
------------------------------------------
Decision after conferring with [~mengxr]: The structure will be:
* ml.Predictor (Predictor will be made public and moved to the ml package in
[SPARK-5995])
* ml.trees.* (Tree abstractions shared by ml.classification and ml.regression)
* ml.ensembles.* (Generic ensemble abstractions shared by ml.classification and
ml.regression)
Since we can't think of more shared abstractions currently, we'll aim for a
flatter directory structure. Closing this JIRA
> spark.ml Prediction abstractions should reside in ml.prediction subpackage
> --------------------------------------------------------------------------
>
> Key: SPARK-7000
> URL: https://issues.apache.org/jira/browse/SPARK-7000
> Project: Spark
> Issue Type: Improvement
> Components: ML
> Affects Versions: 1.3.1
> Reporter: Joseph K. Bradley
> Assignee: Joseph K. Bradley
> Priority: Minor
>
> spark.ml prediction abstractions are currently not gathered; they are in both
> ml.impl and ml.tree. Instead, they should be gathered into ml.prediction.
> This will become more important as more abstractions, such as ensembles, are
> added.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
