[
https://issues.apache.org/jira/browse/SPARK-13463?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15171280#comment-15171280
]
Xiao Li commented on SPARK-13463:
---------------------------------
When using the current external DataSet APIs, I am unable to find any use case
that can make {(child.outputSet -- objectOperator.references)} non empty. That
means, all the elements in child.outputSet are used in Dataset-specific
operators, when these operators are used as parent operators.
In addition, when dataset-specific operators are used as child operators, we
are unable to prune their outputSet by changing the class parameters.
Thus, I do not know how to support column pruning for dataset logical plans.
Please give me some hints. Thanks! : )
> Support Column pruning for Dataset logical plan
> -----------------------------------------------
>
> Key: SPARK-13463
> URL: https://issues.apache.org/jira/browse/SPARK-13463
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Reporter: Davies Liu
>
> Column pruning may not work with some logical plan for Dataset, we need check
> that and make sure they works.
--
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
