[
https://issues.apache.org/jira/browse/CRUNCH-286?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13806010#comment-13806010
]
Stefan De Smit commented on CRUNCH-286:
---------------------------------------
I also implemented the memory table by just skipping the combine phase, as
there is indeed no guarantee this is combining anything anyway.
As for the MultiPhaseCombineFn: that would of course also work. If you worry
about the amount of methods on PGroupedTable, this could be a way. But if
that's not an issue, it makes the functionality more hidden. So, it depends a
bit on how exceptional you think this use case is.
> ability to specify a different function for combiner & reducer
> --------------------------------------------------------------
>
> Key: CRUNCH-286
> URL: https://issues.apache.org/jira/browse/CRUNCH-286
> Project: Crunch
> Issue Type: New Feature
> Components: Core
> Reporter: Stefan De Smit
> Assignee: Josh Wills
> Priority: Minor
> Attachments:
> 0001-add-combineValues-method-with-2-function-arguments.patch, 0002-.patch,
> CRUNCH-286-prototype.patch
>
>
> Extend PGroupedTable with an extra combineValues function that accepts 2
> functions: 1 for combiner phase, 1 for reducer phase.
> This way, different algorithm can be applied.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
