[
https://issues.apache.org/jira/browse/CRUNCH-286?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13810081#comment-13810081
]
Gabriel Reid commented on CRUNCH-286:
-------------------------------------
I'm good with the impl that Stefan put up, combined with (most) of the tests
that you added in the last patch. This means not running the map-phase combiner
function in a MemPipeline, which I think is the fine. The main thing I wanted
to check is that there's nobody who has anything against adding a new method to
PGroupedTable.
If there aren't any objections to that approach, I'll rework your last patch a
bit to remove the expectations on map-phase CombineFns in MemPipeline.
> 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)
