[
https://issues.apache.org/jira/browse/SPARK-9592?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Yin Huai updated SPARK-9592:
----------------------------
Summary: Last implemented based on AggregateExpression1 are calculating the
values for entire DataFrame partition not on GroupedData partition. (was:
First and Last aggregates are calculating the values for entire DataFrame
partition not on GroupedData partition.)
> Last implemented based on AggregateExpression1 are calculating the values for
> entire DataFrame partition not on GroupedData partition.
> --------------------------------------------------------------------------------------------------------------------------------------
>
> Key: SPARK-9592
> URL: https://issues.apache.org/jira/browse/SPARK-9592
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 1.4.0
> Reporter: gaurav
> Priority: Minor
> Original Estimate: 4h
> Remaining Estimate: 4h
>
> In current implementation, First and Last aggregates were calculating the
> values for entire DataFrame partition and then the same value was returned
> for all GroupedData in the partition.
> sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/aggregates.scala
> Fixed the First and Last aggregates should compute first and last value per
> GroupedData instead of entire DataFrame.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
