[jira] [Comment Edited] (SPARK-8827) pyspark.DStream top method

["Saisai Shao (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-8827?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14614562#comment-14614562 ] 

Saisai Shao edited comment on SPARK-8827 at 7/6/15 5:22 AM:
------------------------------------------------------------

I think you could get top values with following pattern like:

{code}
dstream.foreachRDD(lambda r: r.top(...))
{code}

this will get the top values of each bath duration, you could take a try.


was (Author: jerryshao):
I think you could get top values with follow pattern like:

{code}
dstream.foreachRDD(lambda r: r.top(...))
{code}

this will get the top values of each bath duration, you could take a try.

> pyspark.DStream top method
> --------------------------
>
>                 Key: SPARK-8827
>                 URL: https://issues.apache.org/jira/browse/SPARK-8827
>             Project: Spark
>          Issue Type: Improvement
>          Components: PySpark, Streaming
>            Reporter: Emanuele Cesena
>            Priority: Minor
>
> Is there a reason for not having DStream.top?
> Any issues with the following definition?
> {code}
> def topPartition(partition):
>     return sorted(partition, key=lambda p: p[1], reverse=True)[:10]
> def top(counts):
>     return counts.transform(lambda rdd: rdd.mapPartitions(topPartition)
>         .sortBy(lambda p: p[1],ascending=False))
> {code}



--
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