[
https://issues.apache.org/jira/browse/MAHOUT-586?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Owen resolved MAHOUT-586.
------------------------------
Resolution: Won't Fix
Fix Version/s: 0.5
Assignee: Sean Owen
OK, it's possible to reopen if/when there's a submittable patch to consider.
It's fine to iterate on it, here if you like or on the mailing list.
Since the intent is to add functionality, my guess is that the shape of the
patch is more like adding some interface/implementation, and some refactoring
to make some logic shareable to the new parts. That is I think it would be
strongly preferable to keep the existing API, and extend and mimic it to add
new functions. Which is what the OrderBasedRecommenderEvaluator did well.
The sampling iterator shouldn't need to be repeatable, except for unit testing
purposes, and it is insofar as all RNGs are set to a known state on each test.
"Shouldn't" doesn't mean there isn't some practical reason it can't be, but
ideally it's not something anyone depends on.
> Redo RecommenderEvaluator for modularity
> ----------------------------------------
>
> Key: MAHOUT-586
> URL: https://issues.apache.org/jira/browse/MAHOUT-586
> Project: Mahout
> Issue Type: Improvement
> Components: Collaborative Filtering
> Reporter: Lance Norskog
> Assignee: Sean Owen
> Fix For: 0.5
>
> Attachments: MAHOUT-586.patch
>
>
> The RecommenderEvaluator implementation is hard-coded around one algorithm.
> This is a more flexible, modular rewrite.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
