jpountz commented on PR #12382:
URL: https://github.com/apache/lucene/pull/12382#issuecomment-1729775625

   Running with taskCountPerCat=5 and taskRepeatCount=50 gave very similar 
results:
   
   ```
                               TaskQPS baseline      StdDevQPS 
my_modified_version      StdDev                Pct diff p-value
                           HighTerm      472.68      (4.7%)      464.58      
(3.8%)   -1.7% (  -9% -    7%) 0.392
                            MedTerm      585.82      (3.6%)      580.49      
(3.7%)   -0.9% (  -7% -    6%) 0.598
              HighTermDayOfYearSort      286.30      (1.2%)      284.36      
(1.9%)   -0.7% (  -3% -    2%) 0.362
                         OrHighHigh       51.81      (6.7%)       51.54      
(4.5%)   -0.5% ( -10% -   11%) 0.842
                           Wildcard      225.50      (3.3%)      224.32      
(3.3%)   -0.5% (  -6% -    6%) 0.734
                          OrHighLow      611.45      (2.6%)      608.55      
(3.0%)   -0.5% (  -5% -    5%) 0.723
                          OrHighMed      118.00      (5.0%)      117.46      
(3.6%)   -0.5% (  -8% -    8%) 0.823
                            LowTerm     1000.83      (3.3%)      996.92      
(3.5%)   -0.4% (  -7% -    6%) 0.809
                           PKLookup      222.91      (1.0%)      222.27      
(3.2%)   -0.3% (  -4% -    4%) 0.800
                             IntNRQ      106.47     (20.2%)      106.19     
(19.5%)   -0.3% ( -33% -   49%) 0.977
                          LowPhrase       26.20      (1.9%)       26.20      
(1.5%)   -0.0% (  -3% -    3%) 0.990
                            Respell       73.45      (1.1%)       73.50      
(1.5%)    0.1% (  -2% -    2%) 0.914
                             Fuzzy2      119.32      (0.4%)      119.64      
(0.5%)    0.3% (   0% -    1%) 0.182
                  HighTermMonthSort     5878.05      (1.9%)     5897.37      
(2.7%)    0.3% (  -4% -    5%) 0.764
                    CountAndHighMed      121.92      (2.9%)      122.35      
(3.2%)    0.4% (  -5% -    6%) 0.809
                         HighPhrase       46.45      (3.4%)       46.62      
(2.1%)    0.4% (  -4% -    6%) 0.776
                             Fuzzy1      104.63      (0.7%)      105.18      
(0.6%)    0.5% (   0% -    1%) 0.105
                          MedPhrase       27.38      (4.0%)       27.55      
(2.2%)    0.6% (  -5% -    7%) 0.692
                            Prefix3      227.82      (3.7%)      229.47      
(2.7%)    0.7% (  -5% -    7%) 0.633
                          CountTerm    20661.05      (2.5%)    20826.02      
(2.3%)    0.8% (  -3% -    5%) 0.476
                   CountAndHighHigh       40.20      (2.6%)       40.78      
(3.4%)    1.4% (  -4% -    7%) 0.313
                     CountOrHighMed       86.48     (14.9%)       88.87     
(17.9%)    2.8% ( -26% -   41%) 0.721
                    CountOrHighHigh       55.70     (15.0%)       57.27     
(18.0%)    2.8% ( -26% -   42%) 0.719
                        CountPhrase        3.32      (7.6%)        3.45     
(10.2%)    3.9% ( -12% -   23%) 0.356
                         AndHighLow      567.28      (3.7%)      742.59      
(4.6%)   30.9% (  21% -   40%) 0.000
                         AndHighMed      115.23      (2.8%)      161.49      
(4.9%)   40.1% (  31% -   49%) 0.000
                        AndHighHigh       32.62      (4.8%)       47.72      
(4.8%)   46.3% (  34% -   58%) 0.000
   ```
   
   I did a `diff -r` between my baseline and candidate and saw the code changes 
from this PR as expected.


-- 
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-unsubscr...@lucene.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org

Reply via email to