[ https://issues.apache.org/jira/browse/LUCENE-4236?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13418107#comment-13418107 ]
Robert Muir commented on LUCENE-4236: ------------------------------------- Current benchmarks with the generalized conjunction scorer: {noformat} Task QPS trunkStdDev trunk QPS patchStdDev patch Pct diff Phrase 15.69 0.51 15.29 0.30 -7% - 2% AndHighMed 44.35 0.55 43.34 0.63 -4% - 0% Fuzzy1 83.95 2.84 83.05 3.35 -8% - 6% Respell 75.00 3.43 74.30 3.79 -10% - 9% SloppyPhrase 7.05 0.44 7.00 0.17 -8% - 8% TermBGroup1M1P 58.41 1.15 58.17 0.96 -3% - 3% TermGroup1M 30.72 0.14 30.73 0.32 -1% - 1% Fuzzy2 33.81 1.67 33.87 1.26 -8% - 9% TermBGroup1M 48.73 0.34 48.89 0.27 0% - 1% AndHighHigh 8.04 0.16 8.09 0.10 -2% - 3% PKLookup 296.11 2.98 298.19 3.29 -1% - 2% Term 109.66 3.70 110.69 2.76 -4% - 7% Wildcard 61.99 0.80 63.00 2.53 -3% - 7% Prefix3 70.64 1.63 72.07 3.21 -4% - 9% OrHighMed 21.48 1.29 21.94 1.05 -8% - 13% OrHighHigh 8.48 0.47 8.68 0.40 -7% - 13% SpanNear 7.70 0.37 7.96 0.41 -6% - 14% IntNRQ 8.89 0.52 9.40 0.96 -10% - 23% {noformat} Luceneutil doesnt yet benchmark more complicated BQs (e.g. nested ones, or minShouldMatch, or whatever). So we don't see any benefit in these benchmarks. > clean up booleanquery conjunction optimizations a bit > ----------------------------------------------------- > > Key: LUCENE-4236 > URL: https://issues.apache.org/jira/browse/LUCENE-4236 > Project: Lucene - Java > Issue Type: Improvement > Reporter: Robert Muir > Fix For: 4.0, 5.0 > > Attachments: LUCENE-4236.patch, LUCENE-4236.patch, LUCENE-4236.patch > > > After LUCENE-3505, I want to do a slight cleanup: > * compute the term conjunctions optimization in scorer(), so its applied even > if we have optional and prohibited clauses that dont exist in the segment > (e.g. return null) > * use the term conjunctions optimization when optional.size() == > minShouldMatch, as that means they are all mandatory, too. > * don't return booleanscorer1 when optional.size() == minShouldMatch, because > it means we have required clauses and in general BS2 should do a much better > job (e.g. use advance). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org