tustvold opened a new pull request, #2929:
URL: https://github.com/apache/arrow-rs/pull/2929
# Which issue does this PR close?
<!--
We generally require a GitHub issue to be filed for all bug fixes and
enhancements and this helps us generate change logs for our releases. You can
link an issue to this PR using the GitHub syntax. For example `Closes #123`
indicates that this PR will close issue #123.
-->
Part of #2781
# Rationale for this change
<!--
Why are you proposing this change? If this is already explained clearly in
the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand your
changes and offer better suggestions for fixes.
-->
Benchmarks good :smile:
# What changes are included in this PR?
<!--
There is no need to duplicate the description in the issue here but it is
sometimes worth providing a summary of the individual changes in this PR.
-->
Adds some benchmarks of the row format, and adds a disclaimer to the lexsort
kernels
```
lexsort_to_indices([i32, i32_opt]): 4096
time: [464.01 µs 464.15 µs 464.32 µs]
Found 3 outliers among 100 measurements (3.00%)
1 (1.00%) high mild
2 (2.00%) high severe
lexsort_rows([i32, i32_opt]): 4096
time: [429.55 µs 429.66 µs 429.78 µs]
Found 4 outliers among 100 measurements (4.00%)
2 (2.00%) high mild
2 (2.00%) high severe
lexsort_to_indices([i32, i32_opt]): 32768
time: [4.5412 ms 4.5443 ms 4.5486 ms]
Found 5 outliers among 100 measurements (5.00%)
2 (2.00%) high mild
3 (3.00%) high severe
lexsort_rows([i32, i32_opt]): 32768
time: [4.0447 ms 4.0460 ms 4.0474 ms]
Found 5 outliers among 100 measurements (5.00%)
3 (3.00%) high mild
2 (2.00%) high severe
lexsort_to_indices([i32, str_opt(16)]): 4096
time: [465.90 µs 466.07 µs 466.26 µs]
Found 6 outliers among 100 measurements (6.00%)
4 (4.00%) high mild
2 (2.00%) high severe
lexsort_rows([i32, str_opt(16)]): 4096
time: [500.10 µs 500.27 µs 500.49 µs]
Found 8 outliers among 100 measurements (8.00%)
2 (2.00%) high mild
6 (6.00%) high severe
lexsort_to_indices([i32, str_opt(16)]): 32768
time: [4.5679 ms 4.5693 ms 4.5707 ms]
Found 9 outliers among 100 measurements (9.00%)
8 (8.00%) high mild
1 (1.00%) high severe
lexsort_rows([i32, str_opt(16)]): 32768
time: [4.7611 ms 4.7641 ms 4.7671 ms]
Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) high mild
lexsort_to_indices([i32, str(16)]): 4096
time: [466.06 µs 466.21 µs 466.36 µs]
Found 2 outliers among 100 measurements (2.00%)
2 (2.00%) high severe
lexsort_rows([i32, str(16)]): 4096
time: [391.45 µs 391.60 µs 391.76 µs]
Found 5 outliers among 100 measurements (5.00%)
5 (5.00%) high severe
lexsort_to_indices([i32, str(16)]): 32768
time: [4.5577 ms 4.5590 ms 4.5604 ms]
Found 6 outliers among 100 measurements (6.00%)
1 (1.00%) high mild
5 (5.00%) high severe
lexsort_rows([i32, str(16)]): 32768
time: [3.9101 ms 3.9132 ms 3.9162 ms]
lexsort_to_indices([str_opt(16), str(16)]): 4096
time: [878.19 µs 878.43 µs 878.72 µs]
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
lexsort_rows([str_opt(16), str(16)]): 4096
time: [461.13 µs 461.59 µs 462.23 µs]
Found 23 outliers among 100 measurements (23.00%)
23 (23.00%) high severe
lexsort_to_indices([str_opt(16), str(16)]): 32768
time: [9.0754 ms 9.0786 ms 9.0823 ms]
Found 8 outliers among 100 measurements (8.00%)
5 (5.00%) high mild
3 (3.00%) high severe
lexsort_rows([str_opt(16), str(16)]): 32768
time: [4.5031 ms 4.5072 ms 4.5113 ms]
lexsort_to_indices([str_opt(16), str_opt(50), str(16)]): 4096
time: [863.26 µs 863.49 µs 863.74 µs]
Found 6 outliers among 100 measurements (6.00%)
4 (4.00%) high mild
2 (2.00%) high severe
lexsort_rows([str_opt(16), str_opt(50), str(16)]): 4096
time: [537.53 µs 537.76 µs 537.99 µs]
Found 4 outliers among 100 measurements (4.00%)
4 (4.00%) high severe
lexsort_to_indices([str_opt(16), str_opt(50), str(16)]): 32768
time: [9.0009 ms 9.0051 ms 9.0098 ms]
Found 10 outliers among 100 measurements (10.00%)
6 (6.00%) high mild
4 (4.00%) high severe
lexsort_rows([str_opt(16), str_opt(50), str(16)]): 32768
time: [5.3922 ms 5.4006 ms 5.4092 ms]
lexsort_to_indices([str_opt(16), str(16), str_opt(16), str_opt(16),
str_opt(16)]): 4096
time: [880.31 µs 880.52 µs 880.75 µs]
Found 4 outliers among 100 measurements (4.00%)
3 (3.00%) high mild
1 (1.00%) high severe
lexsort_rows([str_opt(16), str(16), str_opt(16), str_opt(16), str_opt(16)]):
4096
time: [686.41 µs 686.66 µs 686.94 µs]
Found 3 outliers among 100 measurements (3.00%)
1 (1.00%) high mild
2 (2.00%) high severe
lexsort_to_indices([str_opt(16), str(16), str_opt(16), str_opt(16),
str_opt(16)]): 32768
time: [9.1124 ms 9.1163 ms 9.1207 ms]
Found 10 outliers among 100 measurements (10.00%)
4 (4.00%) high mild
6 (6.00%) high severe
lexsort_rows([str_opt(16), str(16), str_opt(16), str_opt(16), str_opt(16)]):
32768
time: [6.8218 ms 6.8290 ms 6.8362 ms]
lexsort_to_indices([i32_opt, dict(100,str_opt(50))]): 4096
time: [523.76 µs 523.95 µs 524.16 µs]
Found 8 outliers among 100 measurements (8.00%)
6 (6.00%) high mild
2 (2.00%) high severe
lexsort_rows([i32_opt, dict(100,str_opt(50))]): 4096
time: [430.36 µs 430.61 µs 430.90 µs]
Found 7 outliers among 100 measurements (7.00%)
4 (4.00%) high mild
3 (3.00%) high severe
lexsort_to_indices([i32_opt, dict(100,str_opt(50))]): 32768
time: [4.8896 ms 4.8922 ms 4.8950 ms]
Found 15 outliers among 100 measurements (15.00%)
13 (13.00%) high mild
2 (2.00%) high severe
lexsort_rows([i32_opt, dict(100,str_opt(50))]): 32768
time: [3.7030 ms 3.7046 ms 3.7063 ms]
Found 3 outliers among 100 measurements (3.00%)
3 (3.00%) high mild
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50))]): 4096
time: [153.02 µs 153.07 µs 153.11 µs]
Found 3 outliers among 100 measurements (3.00%)
1 (1.00%) high mild
2 (2.00%) high severe
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50))]): 4096
time: [200.52 µs 200.62 µs 200.73 µs]
Found 7 outliers among 100 measurements (7.00%)
3 (3.00%) high mild
4 (4.00%) high severe
Benchmarking lexsort_to_indices([dict(100,str_opt(50)),
dict(100,str_opt(50))]): 32768: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 5.0s. You may wish to increase
target time to 6.3s, enable flat sampling, or reduce sample count to 60.
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50))]): 32768
time: [1.2349 ms 1.2361 ms 1.2373 ms]
Found 3 outliers among 100 measurements (3.00%)
2 (2.00%) low mild
1 (1.00%) high severe
Benchmarking lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50))]):
32768: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 5.0s. You may wish to increase
target time to 7.4s, enable flat sampling, or reduce sample count to 50.
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50))]): 32768
time: [1.4587 ms 1.4594 ms 1.4601 ms]
Found 2 outliers among 100 measurements (2.00%)
1 (1.00%) high mild
1 (1.00%) high severe
Benchmarking lexsort_to_indices([dict(100,str_opt(50)),
dict(100,str_opt(50)), dict(100,str_opt(50)), str(16)]): ...: Warming up for
3.0000 s
Warning: Unable to complete 100 samples in 5.0s. You may wish to increase
target time to 7.3s, enable flat sampling, or reduce sample count to 50.
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str(16)]): ...
time: [1.4455 ms 1.4461 ms 1.4468 ms]
Found 11 outliers among 100 measurements (11.00%)
5 (5.00%) high mild
6 (6.00%) high severe
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str(16)]): 4096
time: [531.39 µs 531.58 µs 531.77 µs]
Found 6 outliers among 100 measurements (6.00%)
4 (4.00%) high mild
2 (2.00%) high severe
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str(16)]): ... #2
time: [15.592 ms 15.598 ms 15.604 ms]
Found 4 outliers among 100 measurements (4.00%)
3 (3.00%) high mild
1 (1.00%) high severe
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str(16)]): 32768
time: [4.7450 ms 4.7488 ms 4.7526 ms]
Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) high mild
Benchmarking lexsort_to_indices([dict(100,str_opt(50)),
dict(100,str_opt(50)), dict(100,str_opt(50)), str_opt(50)...: Warming up for
3.0000 s
Warning: Unable to complete 100 samples in 5.0s. You may wish to increase
target time to 7.1s, enable flat sampling, or reduce sample count to 50.
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)...
time: [1.4102 ms 1.4107 ms 1.4113 ms]
Found 12 outliers among 100 measurements (12.00%)
5 (5.00%) high mild
7 (7.00%) high severe
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)]): 40...
time: [546.89 µs 547.06 µs 547.23 µs]
Found 7 outliers among 100 measurements (7.00%)
6 (6.00%) high mild
1 (1.00%) high severe
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)... #2
time: [15.753 ms 15.760 ms 15.768 ms]
Found 5 outliers among 100 measurements (5.00%)
5 (5.00%) high mild
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)]): 32...
time: [4.9877 ms 4.9912 ms 4.9947 ms]
Benchmarking lexsort_to_indices([dict(100,str_opt(50)),
dict(100,str_opt(50)), dict(100,str_opt(50)), str_opt(50)... #3: Warming up for
3.0000 s
Warning: Unable to complete 100 samples in 5.0s. You may wish to increase
target time to 7.1s, enable flat sampling, or reduce sample count to 50.
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)... #3
time: [1.4112 ms 1.4118 ms 1.4124 ms]
Found 8 outliers among 100 measurements (8.00%)
3 (3.00%) high mild
5 (5.00%) high severe
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)]): 40... #2
time: [547.35 µs 547.64 µs 547.99 µs]
Found 3 outliers among 100 measurements (3.00%)
3 (3.00%) high severe
lexsort_to_indices([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)... #4
time: [15.796 ms 15.804 ms 15.813 ms]
Found 5 outliers among 100 measurements (5.00%)
5 (5.00%) high mild
lexsort_rows([dict(100,str_opt(50)), dict(100,str_opt(50)),
dict(100,str_opt(50)), str_opt(50)]): 32... #2
time: [5.0166 ms 5.0226 ms 5.0287 ms]
```
So sorting using the row format is in the same ballpark or significantly
faster, with the performance benefit becoming more stark with more columns
# Are there any user-facing changes?
No
<!--
If there are user-facing changes then we may require documentation to be
updated before approving the PR.
-->
<!---
If there are any breaking changes to public APIs, please add the `breaking
change` label.
-->
--
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]
