[
https://issues.apache.org/jira/browse/SPARK-21104?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16050267#comment-16050267
]
darion yaphet commented on SPARK-21104:
---------------------------------------
The feature sort by index is not necessary . So I add a sort step in record
parser .
> Support sort with index when parse LibSVM Record
> ------------------------------------------------
>
> Key: SPARK-21104
> URL: https://issues.apache.org/jira/browse/SPARK-21104
> Project: Spark
> Issue Type: Improvement
> Components: MLlib
> Affects Versions: 2.1.1
> Reporter: darion yaphet
> Priority: Minor
>
> When I'm loading LibSVM from HDFS , I found feature index should be in
> ascending order .
> We can sorted with *indices* when we parse the input line input a (index,
> value) tuple and avoid check if indices are in ascending order after that.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
