[
https://issues.apache.org/jira/browse/MAHOUT-1156?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13600141#comment-13600141
]
Dan Filimon commented on MAHOUT-1156:
-------------------------------------
After talking offline to Ted, I made the following changes:
- moved the classes from o.a.m.clustering.streaming.search to
o.a.m.math.nearestneighbor; the class need to remain in core/ because they
depend on the distance measures which are in core/ (and those cannot be moved
to math/ because of Hadoop dependencies)
- moved lumpyRandomData() to LumpyData
- made SearchQualityTest a value-parameterized test that compares different
searchers' best result to the brute searcher's result and compares runtimes
> Adding nearest neighbor Searchers
> ---------------------------------
>
> Key: MAHOUT-1156
> URL: https://issues.apache.org/jira/browse/MAHOUT-1156
> Project: Mahout
> Issue Type: New Feature
> Components: Clustering
> Affects Versions: 0.8
> Reporter: Dan Filimon
> Attachments: MAHOUT_1156.patch, MAHOUT_1156_tests.patch
>
>
> Adding the Searcher, UpdatableSearcher abstract classes defining what a
> nearest-neighbor searcher does.
> The following implementation are available:
> - BruteSearch
> - ProjectionSearch
> - FastProjectionSearch
> - LocalitySensityHashSearch
> This is part of https://issues.apache.org/jira/browse/MAHOUT-1154
> There is a SearchQualityTest that tests the results for overlap with
> BruteSearch results.
> However, the results are highly variable between runs and I haven't added any
> assertions yet.
> I'll try fixing a test seed (RandomUtils.useTestSeed()) and see how that
> helps (but I have yet to try that version of the test).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
