You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Adrien Grand (JIRA)" <ji...@apache.org> on 2016/02/26 19:11:18 UTC
[jira] [Updated] (LUCENE-7051) Remove the "estimate match count"
optimization from point queries
[ https://issues.apache.org/jira/browse/LUCENE-7051?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Adrien Grand updated LUCENE-7051:
---------------------------------
Attachment: LUCENE-7051.patch
Patch. luceneutil results for IntNRQ look unchanged:
{noformat}
TaskQPS baseline StdDev QPS patch StdDev Pct diff
Fuzzy1 91.75 (13.9%) 86.05 (15.7%) -6.2% ( -31% - 27%)
LowTerm 3636.70 (6.9%) 3591.63 (8.0%) -1.2% ( -15% - 14%)
LowSloppyPhrase 63.24 (4.6%) 62.55 (5.0%) -1.1% ( -10% - 8%)
MedPhrase 750.88 (8.6%) 743.27 (7.0%) -1.0% ( -15% - 15%)
OrHighLow 285.91 (5.9%) 283.27 (5.1%) -0.9% ( -11% - 10%)
HighSpanNear 141.98 (3.8%) 140.82 (4.1%) -0.8% ( -8% - 7%)
MedSpanNear 324.58 (2.7%) 322.60 (3.3%) -0.6% ( -6% - 5%)
MedSloppyPhrase 304.96 (4.1%) 303.55 (4.2%) -0.5% ( -8% - 8%)
HighPhrase 59.46 (4.5%) 59.31 (4.8%) -0.2% ( -9% - 9%)
Prefix3 268.26 (4.5%) 267.70 (5.2%) -0.2% ( -9% - 9%)
AndHighMed 850.41 (3.9%) 850.37 (5.3%) -0.0% ( -8% - 9%)
Fuzzy2 50.95 (12.3%) 50.95 (17.9%) 0.0% ( -26% - 34%)
HighTerm 410.99 (5.8%) 411.95 (5.5%) 0.2% ( -10% - 12%)
Wildcard 403.48 (8.4%) 404.81 (6.8%) 0.3% ( -13% - 16%)
Respell 175.79 (8.8%) 176.57 (7.3%) 0.4% ( -14% - 18%)
OrHighMed 243.10 (9.0%) 244.40 (6.2%) 0.5% ( -13% - 17%)
HighSloppyPhrase 131.84 (4.5%) 132.58 (4.1%) 0.6% ( -7% - 9%)
AndHighLow 1776.00 (5.8%) 1791.85 (7.1%) 0.9% ( -11% - 14%)
MedTerm 1389.09 (5.8%) 1401.91 (7.4%) 0.9% ( -11% - 14%)
OrHighHigh 199.46 (7.1%) 202.13 (7.5%) 1.3% ( -12% - 17%)
LowPhrase 511.73 (4.1%) 519.77 (4.1%) 1.6% ( -6% - 10%)
IntNRQ 88.04 (10.1%) 89.50 (10.1%) 1.7% ( -16% - 24%)
AndHighHigh 508.76 (5.6%) 519.35 (3.5%) 2.1% ( -6% - 11%)
LowSpanNear 938.03 (9.6%) 958.64 (6.2%) 2.2% ( -12% - 19%)
{noformat}
> Remove the "estimate match count" optimization from point queries
> -----------------------------------------------------------------
>
> Key: LUCENE-7051
> URL: https://issues.apache.org/jira/browse/LUCENE-7051
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Adrien Grand
> Assignee: Adrien Grand
> Priority: Minor
> Attachments: LUCENE-7051.patch
>
>
> Point queries try to estimate the number of matches in the visitor so that the doc id set that they build does not have to do it by itself. However, this is incorrect in the multi-valued case and does not seem to buy much (if any) in terms of performance?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org