jpountz commented on PR #15141:
URL: https://github.com/apache/lucene/pull/15141#issuecomment-3241614921
luceneutil on wikibigall reports no performance change:
```
TaskQPS baseline StdDevQPS
my_modified_version StdDev Pct diff p-value
FilteredPrefix3 152.01 (7.9%) 150.80
(8.6%) -0.8% ( -16% - 17%) 0.759
CombinedTerm 38.27 (9.2%) 38.02
(9.5%) -0.7% ( -17% - 19%) 0.823
CountPhrase 4.12 (8.3%) 4.10
(8.3%) -0.5% ( -15% - 17%) 0.836
TermDTSort 374.73 (9.5%) 372.89
(8.5%) -0.5% ( -16% - 19%) 0.863
FilteredOrHighMed 149.04 (6.2%) 148.34
(6.7%) -0.5% ( -12% - 13%) 0.817
FilteredOrStopWords 44.18 (6.7%) 43.98
(7.3%) -0.4% ( -13% - 14%) 0.840
CombinedAndHighHigh 22.81 (9.0%) 22.71
(9.0%) -0.4% ( -16% - 19%) 0.884
CountOrHighMed 347.69 (8.4%) 346.26
(7.9%) -0.4% ( -15% - 17%) 0.873
CombinedAndHighMed 87.13 (9.1%) 86.80
(9.3%) -0.4% ( -17% - 19%) 0.897
AndMedOrHighHigh 85.27 (8.5%) 84.99
(8.3%) -0.3% ( -15% - 18%) 0.901
FilteredIntNRQ 293.30 (8.5%) 292.44
(7.9%) -0.3% ( -15% - 17%) 0.910
AndHighOrMedMed 49.85 (7.8%) 49.73
(8.0%) -0.2% ( -14% - 16%) 0.923
OrHighMed 251.85 (8.5%) 251.31
(8.6%) -0.2% ( -16% - 18%) 0.937
TermDayOfYearSort 277.95 (8.3%) 277.38
(8.3%) -0.2% ( -15% - 17%) 0.938
OrHighHigh 76.81 (8.6%) 76.66
(8.4%) -0.2% ( -15% - 18%) 0.940
Or3Terms 228.99 (8.2%) 228.56
(8.5%) -0.2% ( -15% - 18%) 0.943
FilteredOr2Terms2StopWords 143.11 (6.9%) 142.91
(7.3%) -0.1% ( -13% - 15%) 0.951
FilteredOrHighHigh 65.01 (6.6%) 64.92
(7.1%) -0.1% ( -13% - 14%) 0.950
CombinedOrHighHigh 22.63 (9.3%) 22.60
(9.2%) -0.1% ( -17% - 20%) 0.964
FilteredOr3Terms 162.48 (6.5%) 162.28
(7.2%) -0.1% ( -12% - 14%) 0.954
CountFilteredOrHighHigh 134.53 (5.6%) 134.39
(5.0%) -0.1% ( -10% - 11%) 0.948
CombinedOrHighMed 85.84 (9.7%) 85.75
(9.2%) -0.1% ( -17% - 20%) 0.971
CountOrMany 28.16 (8.5%) 28.14
(8.4%) -0.1% ( -15% - 18%) 0.973
CountFilteredOrMany 26.46 (7.6%) 26.43
(7.8%) -0.1% ( -14% - 16%) 0.971
FilteredPhrase 31.29 (8.2%) 31.26
(8.8%) -0.1% ( -15% - 18%) 0.976
CountFilteredPhrase 24.70 (8.3%) 24.69
(8.5%) -0.1% ( -15% - 18%) 0.978
FilteredAndStopWords 65.42 (7.0%) 65.38
(6.8%) -0.1% ( -12% - 14%) 0.973
CountAndHighMed 300.78 (7.9%) 300.57
(7.7%) -0.1% ( -14% - 16%) 0.978
CountFilteredOrHighMed 146.42 (4.8%) 146.34
(4.5%) -0.1% ( -8% - 9%) 0.970
FilteredAndHighHigh 79.08 (7.0%) 79.05
(6.9%) -0.0% ( -13% - 14%) 0.988
TermMonthSort 3201.37 (9.5%) 3201.30
(9.0%) -0.0% ( -16% - 20%) 0.999
FilteredOrMany 16.12 (8.0%) 16.13
(8.5%) 0.0% ( -15% - 17%) 0.994
FilteredAnd3Terms 187.94 (7.0%) 188.06
(7.6%) 0.1% ( -13% - 15%) 0.977
FilteredAnd2Terms2StopWords 213.74 (7.9%) 213.91
(7.4%) 0.1% ( -14% - 16%) 0.974
Or2Terms2StopWords 201.20 (8.3%) 201.37
(8.4%) 0.1% ( -15% - 18%) 0.975
OrStopWords 48.11 (8.4%) 48.19
(8.4%) 0.2% ( -15% - 18%) 0.949
FilteredAndHighMed 155.18 (7.0%) 155.44
(7.0%) 0.2% ( -12% - 15%) 0.938
CountAndHighHigh 345.98 (8.8%) 346.59
(8.5%) 0.2% ( -15% - 19%) 0.949
CountOrHighHigh 329.55 (8.5%) 330.29
(8.4%) 0.2% ( -15% - 18%) 0.934
And2Terms2StopWords 201.41 (8.2%) 202.05
(8.2%) 0.3% ( -14% - 18%) 0.903
AndHighMed 200.55 (8.0%) 201.41
(8.2%) 0.4% ( -14% - 18%) 0.867
And3Terms 238.50 (8.7%) 239.77
(8.0%) 0.5% ( -14% - 18%) 0.841
AndStopWords 46.80 (8.1%) 47.06
(8.6%) 0.6% ( -14% - 18%) 0.833
FilteredTerm 155.70 (6.7%) 156.64
(7.4%) 0.6% ( -12% - 15%) 0.786
OrMany 22.57 (8.8%) 22.71
(9.1%) 0.6% ( -15% - 20%) 0.824
AndHighHigh 68.58 (8.3%) 69.04
(8.2%) 0.7% ( -14% - 18%) 0.798
Term 627.50 (9.7%) 632.55
(9.9%) 0.8% ( -17% - 22%) 0.795
CountTerm 8741.59 (10.9%) 8846.24
(10.0%) 1.2% ( -17% - 24%) 0.717
TermTitleSort 82.54 (9.3%) 83.80
(12.7%) 1.5% ( -18% - 26%) 0.664
OrHighRare 272.00 (11.8%) 277.14
(12.5%) 1.9% ( -20% - 29%) 0.623
```
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
