[jira] [Commented] (SPARK-17514) df.take(1) and df.limit(1).collect() perform differently in Python

["Apache Spark (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-17514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15485821#comment-15485821 ] 

Apache Spark commented on SPARK-17514:
--------------------------------------

User 'JoshRosen' has created a pull request for this issue:
https://github.com/apache/spark/pull/15068

> df.take(1) and df.limit(1).collect() perform differently in Python
> ------------------------------------------------------------------
>
>                 Key: SPARK-17514
>                 URL: https://issues.apache.org/jira/browse/SPARK-17514
>             Project: Spark
>          Issue Type: Bug
>          Components: PySpark, SQL
>            Reporter: Josh Rosen
>            Assignee: Josh Rosen
>
> In PySpark, {{df.take(1)}} ends up running a single-stage job which computes only one partition of {{df}}, while {{df.limit(1).collect()}} ends up computing all partitions of {{df}} and runs a two-stage job. This difference in performance is confusing, so I think that we should generalize the fix from SPARK-10731 so that {{Dataset.collect()}} can be implemented efficiently in Python.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org