You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "holdenk (JIRA)" <ji...@apache.org> on 2016/10/16 12:25:21 UTC
[jira] [Commented] (SPARK-7653) ML Pipeline and meta-algs should
take random seed param
[ https://issues.apache.org/jira/browse/SPARK-7653?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15579884#comment-15579884 ]
holdenk commented on SPARK-7653:
--------------------------------
I think the simplest workaround would be exposing HasSeed to the public. What do you think?
> ML Pipeline and meta-algs should take random seed param
> -------------------------------------------------------
>
> Key: SPARK-7653
> URL: https://issues.apache.org/jira/browse/SPARK-7653
> Project: Spark
> Issue Type: Improvement
> Components: ML
> Reporter: Joseph K. Bradley
>
> ML Pipelines and other meta-algorithms should implement HasSeed. If the seed is set, then the meta-alg will use that seed to generate new seeds to pass to every component PipelineStage.
> * Note: This may require a little discussion about whether HasSeed should be a public API.
> This will make it easier for users to have reproducible results for entire pipelines (rather than setting the seed for each stage manually).
--
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