[
https://issues.apache.org/jira/browse/SPARK-20632?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16000651#comment-16000651
]
Hyukjin Kwon commented on SPARK-20632:
--------------------------------------
Is this a duplicate of SPARK-12806? It looks both describe the similar
functionalities in different ways.
> Allow 'Column.getItem()' API to accept Vector columns
> -----------------------------------------------------
>
> Key: SPARK-20632
> URL: https://issues.apache.org/jira/browse/SPARK-20632
> Project: Spark
> Issue Type: Improvement
> Components: Spark Core
> Affects Versions: 1.6.3, 2.1.1
> Reporter: Kevin Ushey
> Priority: Minor
>
> The 'getItem()' API is quite handy for extracting values from Dataset columns
> of type 'ArrayType'. It would be quite useful if this could also accept
> 'Vector' columns, e.g. those generated by the various MLLib routines
> (probability columns).
> If I understand correctly, users are forced to define custom UDFs to handle
> this case, and the UDF type required for vector columns is not always obvious.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
