You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@lucene.apache.org by GitBox <gi...@apache.org> on 2022/03/31 20:54:50 UTC
[GitHub] [lucene] Yuti-G opened a new pull request #779: LUCENE-10488: Optimize Facets#getTopDims in IntTaxonomyFacets
Yuti-G opened a new pull request #779:
URL: https://github.com/apache/lucene/pull/779
# Description
This change overrides and optimizes the default implementation of getTopDims in IntTaxonomyFacets which is extended by FastTaxonomyFacetCounts and TaxonomyFacetSumIntAssociations.
# Solution
Override getTopDims and refactor the getTopChildren function in IntTaxonomyFacets to get dimCount (aggregated dim values) more efficiently by checking if dimCount has been populated in indexing time for a dim that is hierarchical or multiValued && requireDimCount, before aggregating dimCount by iterating its child ordinal.
# Tests
Added new testing for the overridden implementations of getTopDims in IntTaxonomyFacets
# Checklist
Please review the following and check all that apply:
- [X] I have reviewed the guidelines for [How to Contribute](https://wiki.apache.org/lucene/HowToContribute) and my code conforms to the standards described there to the best of my ability.
- [X] I have created a Jira issue and added the issue ID to my pull request title.
- [X] I have given Lucene maintainers [access](https://help.github.com/en/articles/allowing-changes-to-a-pull-request-branch-created-from-a-fork) to contribute to my PR branch. (optional but recommended)
- [X] I have developed this patch against the `main` branch.
- [X] I have run `./gradlew check`.
- [X] I have added tests for my changes.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] Yuti-G commented on pull request #779: LUCENE-10488: Optimize Facets#getTopDims in IntTaxonomyFacets
Posted by GitBox <gi...@apache.org>.
Yuti-G commented on pull request #779:
URL: https://github.com/apache/lucene/pull/779#issuecomment-1085135788
I ran the benchmark and no regression found, please see attached:
TaskQPS baseline StdDevQPS candidate StdDev Pct diff p-value
BrowseDayOfYearTaxoFacets 18.86 (31.4%) 17.54 (25.5%) -7.0% ( -48% - 72%) 0.439
BrowseDateTaxoFacets 18.70 (31.5%) 17.42 (25.6%) -6.8% ( -48% - 73%) 0.451
BrowseMonthTaxoFacets 21.86 (45.0%) 20.41 (45.0%) -6.7% ( -66% - 151%) 0.639
BrowseRandomLabelTaxoFacets 16.09 (30.1%) 15.48 (27.6%) -3.8% ( -47% - 77%) 0.676
AndHighHighDayTaxoFacets 5.13 (3.6%) 5.00 (3.6%) -2.6% ( -9% - 4%) 0.021
MedIntervalsOrdered 67.73 (8.7%) 66.19 (8.9%) -2.3% ( -18% - 16%) 0.410
PKLookup 223.68 (3.2%) 218.83 (4.8%) -2.2% ( -9% - 6%) 0.095
OrNotHighMed 1054.36 (3.5%) 1033.09 (3.5%) -2.0% ( -8% - 5%) 0.070
OrHighHigh 25.33 (4.4%) 24.85 (4.2%) -1.9% ( -10% - 7%) 0.170
AndHighMed 281.83 (4.0%) 276.96 (4.7%) -1.7% ( -9% - 7%) 0.209
LowIntervalsOrdered 32.41 (4.8%) 31.87 (5.3%) -1.7% ( -11% - 8%) 0.294
AndHighMedDayTaxoFacets 126.03 (2.2%) 123.91 (3.4%) -1.7% ( -7% - 3%) 0.063
OrHighMed 162.06 (3.6%) 159.35 (4.9%) -1.7% ( -9% - 7%) 0.214
HighIntervalsOrdered 4.97 (7.9%) 4.89 (8.5%) -1.6% ( -16% - 16%) 0.529
AndHighHigh 103.06 (4.0%) 101.41 (4.6%) -1.6% ( -9% - 7%) 0.236
BrowseMonthSSDVFacets 17.04 (21.2%) 16.79 (21.4%) -1.5% ( -36% - 52%) 0.829
OrNotHighHigh 1800.95 (4.5%) 1776.81 (5.0%) -1.3% ( -10% - 8%) 0.369
LowTerm 2702.01 (4.6%) 2666.76 (4.5%) -1.3% ( -9% - 8%) 0.365
OrHighLow 902.36 (3.8%) 891.54 (4.3%) -1.2% ( -8% - 7%) 0.350
LowSpanNear 6.45 (5.3%) 6.37 (5.6%) -1.2% ( -11% - 10%) 0.493
HighSloppyPhrase 7.29 (6.1%) 7.21 (3.1%) -1.1% ( -9% - 8%) 0.466
OrNotHighLow 1238.66 (3.6%) 1224.97 (5.2%) -1.1% ( -9% - 8%) 0.437
MedSloppyPhrase 94.71 (3.0%) 93.77 (2.1%) -1.0% ( -5% - 4%) 0.228
MedSpanNear 10.51 (5.4%) 10.41 (5.5%) -0.9% ( -11% - 10%) 0.585
AndHighLow 1402.90 (5.4%) 1389.97 (6.7%) -0.9% ( -12% - 11%) 0.634
BrowseDateSSDVFacets 2.76 (12.9%) 2.74 (12.8%) -0.8% ( -23% - 28%) 0.844
OrHighNotMed 1056.33 (4.7%) 1048.21 (4.2%) -0.8% ( -9% - 8%) 0.585
OrHighNotLow 1139.29 (4.8%) 1131.24 (5.1%) -0.7% ( -10% - 9%) 0.652
HighSpanNear 18.95 (5.9%) 18.84 (5.6%) -0.6% ( -11% - 11%) 0.736
LowSloppyPhrase 153.65 (4.7%) 152.74 (2.7%) -0.6% ( -7% - 7%) 0.625
BrowseRandomLabelSSDVFacets 10.55 (8.6%) 10.49 (7.7%) -0.5% ( -15% - 17%) 0.835
OrHighNotHigh 1017.62 (3.9%) 1013.76 (5.1%) -0.4% ( -9% - 9%) 0.793
HighTerm 1984.46 (6.5%) 1979.54 (5.6%) -0.2% ( -11% - 12%) 0.897
OrHighMedDayTaxoFacets 9.50 (5.2%) 9.50 (4.4%) -0.1% ( -9% - 10%) 0.953
LowPhrase 105.81 (2.5%) 105.75 (3.5%) -0.1% ( -5% - 6%) 0.954
Fuzzy1 112.42 (2.4%) 112.45 (2.8%) 0.0% ( -5% - 5%) 0.975
Fuzzy2 77.93 (2.3%) 77.99 (2.4%) 0.1% ( -4% - 4%) 0.924
MedPhrase 51.59 (2.5%) 51.64 (2.7%) 0.1% ( -4% - 5%) 0.913
Respell 78.97 (2.6%) 79.30 (2.7%) 0.4% ( -4% - 5%) 0.614
HighPhrase 107.35 (3.0%) 107.81 (3.7%) 0.4% ( -6% - 7%) 0.690
MedTermDayTaxoFacets 29.02 (5.6%) 29.16 (6.0%) 0.5% ( -10% - 12%) 0.785
HighTermDayOfYearSort 58.52 (10.5%) 58.82 (18.2%) 0.5% ( -25% - 32%) 0.915
TermDTSort 230.85 (14.2%) 232.33 (17.7%) 0.6% ( -27% - 37%) 0.899
BrowseDayOfYearSSDVFacets 15.11 (17.6%) 15.37 (19.8%) 1.7% ( -30% - 47%) 0.771
MedTerm 2132.76 (6.0%) 2171.28 (5.2%) 1.8% ( -8% - 13%) 0.312
Prefix3 133.70 (15.2%) 136.27 (13.9%) 1.9% ( -23% - 36%) 0.676
Wildcard 239.44 (13.3%) 245.27 (11.7%) 2.4% ( -19% - 31%) 0.539
HighTermTitleBDVSort 157.48 (11.3%) 161.66 (17.1%) 2.7% ( -23% - 35%) 0.563
HighTermMonthSort 47.85 (11.3%) 49.32 (17.9%) 3.1% ( -23% - 36%) 0.519
IntNRQ 98.08 (29.9%) 108.25 (28.8%) 10.4% ( -37% - 98%) 0.264
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org