[
https://issues.apache.org/jira/browse/SPARK-9399?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Josh Rosen updated SPARK-9399:
------------------------------
Description:
There are a few minor optimizations in PythonRDD which may avoid garbage
creation or Scala overheads:
- Replace a foreach() loop with a while() loop.
- Returns nulls instead of Options to avoid calling `Option.apply()` and
`Option.isDefined` once per read() call.
- Call .size() instead of .length(), thereby avoiding an implicit Java -> Scala
collections conversion.
was:
There are a few minor optimizations in PythonRDD which may avoid garbage
creation or Scala overheads:
- Replace a foreach() loop with a while() loop.
- Returns nulls instead of Options to avoid allocating an Option once per
read() call.
- Call .size() instead of .length(), thereby avoiding an implicit Java -> Scala
collections conversion.
> Assorted micro-optimizations in PythonRDD
> -----------------------------------------
>
> Key: SPARK-9399
> URL: https://issues.apache.org/jira/browse/SPARK-9399
> Project: Spark
> Issue Type: Improvement
> Components: PySpark
> Reporter: Josh Rosen
> Assignee: Josh Rosen
> Priority: Minor
>
> There are a few minor optimizations in PythonRDD which may avoid garbage
> creation or Scala overheads:
> - Replace a foreach() loop with a while() loop.
> - Returns nulls instead of Options to avoid calling `Option.apply()` and
> `Option.isDefined` once per read() call.
> - Call .size() instead of .length(), thereby avoiding an implicit Java ->
> Scala collections conversion.
--
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
