[
https://issues.apache.org/jira/browse/BEAM-106?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17223105#comment-17223105
]
Martin Blais commented on BEAM-106:
-----------------------------------
This is a really important request/feature, as it enables the implementation of
a class of iterative algorithms. Many graph algorithm implementations that use
successive joins on very large graphs & clustering, for example, are enabled
(or at least substantially simplified and improved) with the presence of a
runtime conditional operation.
An impoverished conditional that merely skips evaluation forward to a future
step would probably handle many of the cases (handles the looping case).
> Native support for conditional iteration
> ----------------------------------------
>
> Key: BEAM-106
> URL: https://issues.apache.org/jira/browse/BEAM-106
> Project: Beam
> Issue Type: New Feature
> Components: sdk-ideas
> Reporter: Luke Cwik
> Priority: P3
>
> Ported from: https://github.com/GoogleCloudPlatform/DataflowJavaSDK/issues/50
> There are a variety of use cases which would benefit from native support for
> conditional iteration.
> For instance,
> http://stackoverflow.com/questions/31654421/conditional-iterations-in-google-cloud-dataflow/31659923?noredirect=1#comment51264604_31659923
> asks about being able to write a loop like the following:
> {code}
> PCollection data = ...
> while(needsMoreWork(data)) {
> data = doAStep(data)
> }
> {code}
> If there are specific use cases please let us know the details. In the future
> we will use this issue to post progress updates.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
