[
https://issues.apache.org/jira/browse/SYSTEMML-909?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15492355#comment-15492355
]
Mike Dusenberry commented on SYSTEMML-909:
------------------------------------------
Also I would split the scripts up and just use one to convert from DataFrames
to SystemML matrices, but the issue in SYSTEMML-869 is present still.
Just to be clear, the API interfaces are great in terms of being able to create
scripts and pass around inputs/outputs. We just need to knock out a few of
these internal performance issues and scratch space bugs and we'll be set.
> `determineDataFrameDimensionsIfNeeded(...)` is a bottleneck.
> ------------------------------------------------------------
>
> Key: SYSTEMML-909
> URL: https://issues.apache.org/jira/browse/SYSTEMML-909
> Project: SystemML
> Issue Type: Improvement
> Reporter: Mike Dusenberry
>
> The {{[determineDataFrameDimensionsIfNeeded(...) |
> https://github.com/apache/incubator-systemml/blob/master/src/main/java/org/apache/sysml/api/mlcontext/MLContextConversionUtil.java#L585]}}
> function in {{MLContext}} is a major bottleneck, particularly due to the
> `javaRDD` call.
> The issue I'm seeing is that the javaRDD.count() function causes execution of
> the lazy DataFrames I pass in, which are created from another DataFrame via
> df.randomSplit([0.8, 0.2]), thus a shuffle occurs. I know that this is going
> to happen anyways in the internal conversion, but it wastes a lot of time by
> having to also do it in this step too. Assume that I have more data than I
> can efficiently cache (~7TB with the potential for much more), so I need to
> incur the shuffle step only once on the way into the engine.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
