[
https://issues.apache.org/jira/browse/LUCENE-6697?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14698417#comment-14698417
]
Steve Rowe commented on LUCENE-6697:
------------------------------------
Another failing test in the {{TestRangeTree}} suite, didn't try to repro, from
[http://jenkins.sarowe.net/job/Lucene-Solr-tests-5.x-Java7/1519/]:
{noformat}
[junit4] Suite: org.apache.lucene.rangetree.TestRangeTree
[junit4] 2> NOTE: reproduce with: ant test -Dtestcase=TestRangeTree
-Dtests.method=testMultiValuedSortedSet -Dtests.seed=824523825E862EFE
-Dtests.slow=true -Dtests.locale=en -Dtests.timezone=SystemV/EST5
-Dtests.asserts=true -Dtests.file.encoding=UTF-8
[junit4] ERROR 0.09s J0 | TestRangeTree.testMultiValuedSortedSet <<<
[junit4] > Throwable #1: java.lang.IllegalArgumentException:
maxValuesSortInHeap must be >= maxValuesInLeafNode; got 1552 vs
maxValuesInLeafNode=1624
[junit4] > at
__randomizedtesting.SeedInfo.seed([824523825E862EFE:F93C7006ECC25330]:0)
[junit4] > at
org.apache.lucene.rangetree.RangeTreeWriter.verifyParams(RangeTreeWriter.java:114)
[junit4] > at
org.apache.lucene.rangetree.RangeTreeDocValuesFormat.<init>(RangeTreeDocValuesFormat.java:98)
[junit4] > at
org.apache.lucene.rangetree.TestRangeTree.testMultiValuedSortedSet(TestRangeTree.java:215)
[junit4] > at java.lang.Thread.run(Thread.java:745)
{noformat}
> Use 1D KD tree for alternative to postings based numeric range filters
> ----------------------------------------------------------------------
>
> Key: LUCENE-6697
> URL: https://issues.apache.org/jira/browse/LUCENE-6697
> Project: Lucene - Core
> Issue Type: New Feature
> Reporter: Michael McCandless
> Assignee: Michael McCandless
> Fix For: 5.3, Trunk
>
> Attachments: LUCENE-6697.patch, LUCENE-6697.patch, LUCENE-6697.patch
>
>
> Today Lucene uses postings to index a numeric value at multiple
> precision levels for fast range searching. It's somewhat costly: each
> numeric value is indexed with multiple terms (4 terms by default)
> ... I think a dedicated 1D BKD tree should be more compact and perform
> better.
> It should also easily generalize beyond 64 bits to arbitrary byte[],
> e.g. for LUCENE-5596, but I haven't explored that here.
> A 1D BKD tree just sorts all values, and then indexes adjacent leaf
> blocks of size 512-1024 (by default) values per block, and their
> docIDs, into a fully balanced binary tree. Building the range filter
> is then just a recursive walk through this tree.
> It's the same structure we use for 2D lat/lon BKD tree, just with 1D
> instead. I implemented it as a DocValuesFormat that also writes the
> numeric tree on the side.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
