[
https://issues.apache.org/jira/browse/SPARK-17848?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15561367#comment-15561367
]
Apache Spark commented on SPARK-17848:
--------------------------------------
User 'zhengruifeng' has created a pull request for this issue:
https://github.com/apache/spark/pull/15414
> Move LabelCol datatype cast into Predictor.fit
> ----------------------------------------------
>
> Key: SPARK-17848
> URL: https://issues.apache.org/jira/browse/SPARK-17848
> Project: Spark
> Issue Type: Improvement
> Components: ML
> Reporter: zhengruifeng
>
> `Classifier.getNumClasses` can not support Non-Double types, and
> classification algos relying on it do not support non-double labelCol (like
> `NavieBayes`)
> As suggested by [~sethah], it is not a reasonable way to do datatype cast
> everywhere. And we can make cast only happen in `Predictor`.
> [~yanboliang][~josephkb][~srowen]
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
