You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@arrow.apache.org by "Dandandan (via GitHub)" <gi...@apache.org> on 2026/01/20 12:31:05 UTC
[PR] Speedup filter/BitIndexIterator [arrow-rs]
Dandandan opened a new pull request, #9229:
URL: https://github.com/apache/arrow-rs/pull/9229
# 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.
-->
- Closes #NNN.
# 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.
-->
# 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.
-->
# Are these changes tested?
<!--
We typically require tests for all PRs in order to:
1. Prevent the code from being accidentally broken by subsequent changes
2. Serve as another way to document the expected behavior of the code
If tests are not included in your PR, please explain why (for example, are they covered by existing tests)?
-->
# Are there any user-facing changes?
<!--
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 call them out.
-->
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to 1.5x) filter/BitIndexIterator/iter_set_bits_rev [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on code in PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#discussion_r2708228686
##########
arrow-buffer/src/util/bit_iterator.rs:
##########
@@ -305,12 +305,12 @@ impl<'a> BitIndexIterator<'a> {
impl Iterator for BitIndexIterator<'_> {
type Item = usize;
- #[inline]
+ #[inline(always)]
fn next(&mut self) -> Option<Self::Item> {
loop {
if self.current_chunk != 0 {
let bit_pos = self.current_chunk.trailing_zeros();
- self.current_chunk ^= 1 << bit_pos;
+ self.current_chunk &= self.current_chunk - 1;
Review Comment:
This generates much more efficient code
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to 1.5x) filter/BitIndexIterator/iter_set_bits_rev [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772830108
run benchmark filter_kernels
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3773309106
🤖 `./gh_compare_arrow.sh` [gh_compare_arrow.sh](https://github.com/alamb/datafusion-benchmarking/blob/main/scripts/gh_compare_arrow.sh) Running
Linux aal-dev 6.14.0-1018-gcp #19~24.04.1-Ubuntu SMP Wed Sep 24 23:23:09 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
Comparing improve_filter_speed4 (7c22b75fce3a3a7b42a3ea385e2b8df920b549fd) to abd11d74bdc179840085bd99d6a697c098e5c751 [diff](https://github.com/apache/arrow-rs/compare/abd11d74bdc179840085bd99d6a697c098e5c751..7c22b75fce3a3a7b42a3ea385e2b8df920b549fd)
BENCH_NAME=filter_kernels
BENCH_COMMAND=cargo bench --features=arrow,async,test_common,experimental --bench filter_kernels
BENCH_FILTER=
BENCH_BRANCH_NAME=improve_filter_speed4
Results will be posted here when complete
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter/BitIndexIterator/iter_set_bits_rev [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772735692
🤖 `./gh_compare_arrow.sh` [gh_compare_arrow.sh](https://github.com/alamb/datafusion-benchmarking/blob/main/scripts/gh_compare_arrow.sh) Running
Linux aal-dev 6.14.0-1018-gcp #19~24.04.1-Ubuntu SMP Wed Sep 24 23:23:09 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
Comparing improve_filter_speed4 (4c2c40638aba5ae4be6861b91f5cf276ff516406) to abd11d74bdc179840085bd99d6a697c098e5c751 [diff](https://github.com/apache/arrow-rs/compare/abd11d74bdc179840085bd99d6a697c098e5c751..4c2c40638aba5ae4be6861b91f5cf276ff516406)
BENCH_NAME=filter_kernels
BENCH_COMMAND=cargo bench --features=arrow,async,test_common,experimental --bench filter_kernels
BENCH_FILTER=
BENCH_BRANCH_NAME=improve_filter_speed4
Results will be posted here when complete
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter/BitIndexIterator/iter_set_bits_rev [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772664907
🤖 `./gh_compare_arrow.sh` [gh_compare_arrow.sh](https://github.com/alamb/datafusion-benchmarking/blob/main/scripts/gh_compare_arrow.sh) Running
Linux aal-dev 6.14.0-1018-gcp #19~24.04.1-Ubuntu SMP Wed Sep 24 23:23:09 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
Comparing improve_filter_speed4 (4c2c40638aba5ae4be6861b91f5cf276ff516406) to abd11d74bdc179840085bd99d6a697c098e5c751 [diff](https://github.com/apache/arrow-rs/compare/abd11d74bdc179840085bd99d6a697c098e5c751..4c2c40638aba5ae4be6861b91f5cf276ff516406)
BENCH_NAME=filter_kernels
BENCH_COMMAND=cargo bench --features=arrow,async,test_common,experimental --bench filter_kernels
BENCH_FILTER=
BENCH_BRANCH_NAME=improve_filter_speed4
Results will be posted here when complete
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3773433867
🤖: Benchmark completed
<details><summary>Details</summary>
<p>
```
group improve_filter_speed4 main
----- --------------------- ----
filter context decimal128 (kept 1/2) 1.00 46.3±7.72µs ? ?/sec 1.14 52.5±6.87µs ? ?/sec
filter context decimal128 high selectivity (kept 1023/1024) 1.00 50.0±1.21µs ? ?/sec 1.00 50.2±1.39µs ? ?/sec
filter context decimal128 low selectivity (kept 1/1024) 1.01 232.0±1.66ns ? ?/sec 1.00 229.8±2.48ns ? ?/sec
filter context f32 (kept 1/2) 1.00 77.5±0.40µs ? ?/sec 1.13 87.6±0.38µs ? ?/sec
filter context f32 high selectivity (kept 1023/1024) 1.00 10.0±0.51µs ? ?/sec 1.02 10.2±0.58µs ? ?/sec
filter context f32 low selectivity (kept 1/1024) 1.00 423.1±1.96ns ? ?/sec 1.03 436.0±3.93ns ? ?/sec
filter context fsb with value length 20 (kept 1/2) 1.00 60.7±0.56µs ? ?/sec 1.17 70.9±1.60µs ? ?/sec
filter context fsb with value length 20 high selectivity (kept 1023/1024) 1.00 60.9±1.83µs ? ?/sec 1.16 70.6±0.57µs ? ?/sec
filter context fsb with value length 20 low selectivity (kept 1/1024) 1.00 60.6±0.46µs ? ?/sec 1.17 71.0±2.21µs ? ?/sec
filter context fsb with value length 5 (kept 1/2) 1.00 60.7±0.78µs ? ?/sec 1.17 70.9±1.44µs ? ?/sec
filter context fsb with value length 5 high selectivity (kept 1023/1024) 1.00 60.7±1.25µs ? ?/sec 1.16 70.8±1.27µs ? ?/sec
filter context fsb with value length 5 low selectivity (kept 1/1024) 1.00 60.5±0.48µs ? ?/sec 1.17 70.7±0.69µs ? ?/sec
filter context fsb with value length 50 (kept 1/2) 1.00 60.6±0.84µs ? ?/sec 1.17 70.8±1.08µs ? ?/sec
filter context fsb with value length 50 high selectivity (kept 1023/1024) 1.00 60.7±0.86µs ? ?/sec 1.17 70.8±1.03µs ? ?/sec
filter context fsb with value length 50 low selectivity (kept 1/1024) 1.00 60.7±0.88µs ? ?/sec 1.16 70.6±0.28µs ? ?/sec
filter context i32 (kept 1/2) 1.02 16.6±0.10µs ? ?/sec 1.00 16.3±0.08µs ? ?/sec
filter context i32 high selectivity (kept 1023/1024) 1.00 6.7±0.39µs ? ?/sec 1.00 6.7±0.49µs ? ?/sec
filter context i32 low selectivity (kept 1/1024) 1.00 216.0±1.55ns ? ?/sec 1.02 219.9±0.74ns ? ?/sec
filter context i32 w NULLs (kept 1/2) 1.00 77.5±0.36µs ? ?/sec 1.13 87.8±0.65µs ? ?/sec
filter context i32 w NULLs high selectivity (kept 1023/1024) 1.00 10.4±0.53µs ? ?/sec 1.00 10.4±0.41µs ? ?/sec
filter context i32 w NULLs low selectivity (kept 1/1024) 1.00 423.6±11.27ns ? ?/sec 1.04 438.9±2.68ns ? ?/sec
filter context mixed string view (kept 1/2) 1.00 105.1±4.80µs ? ?/sec 1.13 119.0±7.28µs ? ?/sec
filter context mixed string view high selectivity (kept 1023/1024) 1.01 54.8±1.01µs ? ?/sec 1.00 54.2±1.46µs ? ?/sec
filter context mixed string view low selectivity (kept 1/1024) 1.00 622.2±2.63ns ? ?/sec 1.00 623.7±7.83ns ? ?/sec
filter context short string view (kept 1/2) 1.00 111.3±7.28µs ? ?/sec 1.06 118.0±6.53µs ? ?/sec
filter context short string view high selectivity (kept 1023/1024) 1.02 55.2±1.49µs ? ?/sec 1.00 54.3±0.96µs ? ?/sec
filter context short string view low selectivity (kept 1/1024) 1.00 480.9±1.71ns ? ?/sec 1.03 497.1±16.38ns ? ?/sec
filter context string (kept 1/2) 1.00 568.8±14.17µs ? ?/sec 1.05 595.1±19.65µs ? ?/sec
filter context string dictionary (kept 1/2) 1.00 17.0±0.20µs ? ?/sec 1.00 17.0±0.18µs ? ?/sec
filter context string dictionary high selectivity (kept 1023/1024) 1.05 7.5±0.45µs ? ?/sec 1.00 7.1±0.40µs ? ?/sec
filter context string dictionary low selectivity (kept 1/1024) 1.00 635.9±53.78ns ? ?/sec 1.01 640.5±24.45ns ? ?/sec
filter context string dictionary w NULLs (kept 1/2) 1.00 78.3±1.61µs ? ?/sec 1.13 88.7±1.67µs ? ?/sec
filter context string dictionary w NULLs high selectivity (kept 1023/1024) 1.00 11.2±0.55µs ? ?/sec 1.01 11.3±0.49µs ? ?/sec
filter context string dictionary w NULLs low selectivity (kept 1/1024) 1.00 853.3±5.98ns ? ?/sec 1.01 865.0±4.10ns ? ?/sec
filter context string high selectivity (kept 1023/1024) 1.00 617.9±18.20µs ? ?/sec 1.03 638.9±16.29µs ? ?/sec
filter context string low selectivity (kept 1/1024) 1.03 912.8±11.02ns ? ?/sec 1.00 888.3±4.48ns ? ?/sec
filter context u8 (kept 1/2) 1.00 15.0±0.12µs ? ?/sec 1.00 15.0±0.19µs ? ?/sec
filter context u8 high selectivity (kept 1023/1024) 1.00 1898.6±23.65ns ? ?/sec 1.06 2.0±0.02µs ? ?/sec
filter context u8 low selectivity (kept 1/1024) 1.02 216.6±2.01ns ? ?/sec 1.00 212.5±1.11ns ? ?/sec
filter context u8 w NULLs (kept 1/2) 1.00 76.4±3.16µs ? ?/sec 1.13 86.0±0.70µs ? ?/sec
filter context u8 w NULLs high selectivity (kept 1023/1024) 1.00 5.1±0.16µs ? ?/sec 1.00 5.1±0.06µs ? ?/sec
filter context u8 w NULLs low selectivity (kept 1/1024) 1.00 418.7±5.03ns ? ?/sec 1.04 433.9±3.16ns ? ?/sec
filter decimal128 (kept 1/2) 1.17 57.9±3.20µs ? ?/sec 1.00 49.4±3.24µs ? ?/sec
filter decimal128 high selectivity (kept 1023/1024) 1.04 53.5±0.92µs ? ?/sec 1.00 51.7±1.40µs ? ?/sec
filter decimal128 low selectivity (kept 1/1024) 1.02 2.9±0.02µs ? ?/sec 1.00 2.9±0.05µs ? ?/sec
filter f32 (kept 1/2) 1.00 149.9±1.44µs ? ?/sec 1.10 164.6±0.45µs ? ?/sec
filter fsb with value length 20 (kept 1/2) 1.10 138.1±2.43µs ? ?/sec 1.00 125.4±0.64µs ? ?/sec
filter fsb with value length 20 high selectivity (kept 1023/1024) 1.00 70.4±2.23µs ? ?/sec 1.02 71.6±2.05µs ? ?/sec
filter fsb with value length 20 low selectivity (kept 1/1024) 1.00 3.4±0.02µs ? ?/sec 1.00 3.5±0.04µs ? ?/sec
filter fsb with value length 5 (kept 1/2) 1.19 143.7±4.89µs ? ?/sec 1.00 120.9±0.46µs ? ?/sec
filter fsb with value length 5 high selectivity (kept 1023/1024) 1.00 11.0±0.63µs ? ?/sec 1.09 12.0±0.67µs ? ?/sec
filter fsb with value length 5 low selectivity (kept 1/1024) 1.02 3.4±0.05µs ? ?/sec 1.00 3.4±0.01µs ? ?/sec
filter fsb with value length 50 (kept 1/2) 1.00 159.5±11.75µs ? ?/sec 1.02 163.4±13.47µs ? ?/sec
filter fsb with value length 50 high selectivity (kept 1023/1024) 1.00 202.4±6.98µs ? ?/sec 1.03 207.5±6.93µs ? ?/sec
filter fsb with value length 50 low selectivity (kept 1/1024) 1.02 3.4±0.03µs ? ?/sec 1.00 3.4±0.10µs ? ?/sec
filter i32 (kept 1/2) 1.25 54.0±0.67µs ? ?/sec 1.00 43.4±1.30µs ? ?/sec
filter i32 high selectivity (kept 1023/1024) 1.00 9.0±0.51µs ? ?/sec 1.02 9.1±0.51µs ? ?/sec
filter i32 low selectivity (kept 1/1024) 1.24 2.9±0.02µs ? ?/sec 1.00 2.4±0.03µs ? ?/sec
filter optimize (kept 1/2) 1.00 37.0±0.11µs ? ?/sec 1.46 53.9±0.24µs ? ?/sec
filter optimize high selectivity (kept 1023/1024) 1.00 2.6±0.01µs ? ?/sec 1.05 2.7±0.02µs ? ?/sec
filter optimize low selectivity (kept 1/1024) 1.00 1836.2±15.23ns ? ?/sec 1.69 3.1±0.03µs ? ?/sec
filter run array (kept 1/2) 1.26 468.1±7.61µs ? ?/sec 1.00 371.4±4.90µs ? ?/sec
filter run array high selectivity (kept 1023/1024) 1.22 485.2±12.04µs ? ?/sec 1.00 398.7±5.67µs ? ?/sec
filter run array low selectivity (kept 1/1024) 1.31 370.4±7.16µs ? ?/sec 1.00 281.7±3.55µs ? ?/sec
filter single record batch 1.00 38.3±0.42µs ? ?/sec 1.40 53.7±0.28µs ? ?/sec
filter u8 (kept 1/2) 1.00 36.0±0.10µs ? ?/sec 1.27 45.6±0.10µs ? ?/sec
filter u8 high selectivity (kept 1023/1024) 1.00 3.7±0.02µs ? ?/sec 1.01 3.8±0.06µs ? ?/sec
filter u8 low selectivity (kept 1/1024) 1.00 3.0±0.07µs ? ?/sec 1.08 3.2±0.02µs ? ?/sec
```
</p>
</details>
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772837009
🤖: Benchmark completed
<details><summary>Details</summary>
<p>
```
group improve_filter_speed4 main
----- --------------------- ----
filter context decimal128 (kept 1/2) 1.00 45.7±3.77µs ? ?/sec 1.03 47.3±7.87µs ? ?/sec
filter context decimal128 high selectivity (kept 1023/1024) 1.02 49.9±1.44µs ? ?/sec 1.00 48.8±1.68µs ? ?/sec
filter context decimal128 low selectivity (kept 1/1024) 1.00 226.1±1.29ns ? ?/sec 1.06 239.5±5.43ns ? ?/sec
filter context f32 (kept 1/2) 1.00 77.2±0.18µs ? ?/sec 1.13 87.6±0.49µs ? ?/sec
filter context f32 high selectivity (kept 1023/1024) 1.00 10.0±0.42µs ? ?/sec 1.00 10.0±0.38µs ? ?/sec
filter context f32 low selectivity (kept 1/1024) 1.00 416.8±1.60ns ? ?/sec 1.10 456.5±8.38ns ? ?/sec
filter context fsb with value length 20 (kept 1/2) 1.00 60.6±0.23µs ? ?/sec 1.17 70.8±0.54µs ? ?/sec
filter context fsb with value length 20 high selectivity (kept 1023/1024) 1.00 60.6±0.28µs ? ?/sec 1.17 70.8±1.00µs ? ?/sec
filter context fsb with value length 20 low selectivity (kept 1/1024) 1.00 60.6±0.42µs ? ?/sec 1.17 70.6±0.42µs ? ?/sec
filter context fsb with value length 5 (kept 1/2) 1.00 60.7±0.86µs ? ?/sec 1.16 70.7±0.51µs ? ?/sec
filter context fsb with value length 5 high selectivity (kept 1023/1024) 1.00 60.7±0.98µs ? ?/sec 1.17 71.0±2.75µs ? ?/sec
filter context fsb with value length 5 low selectivity (kept 1/1024) 1.00 60.8±1.39µs ? ?/sec 1.16 70.7±0.57µs ? ?/sec
filter context fsb with value length 50 (kept 1/2) 1.00 60.7±0.91µs ? ?/sec 1.17 70.9±1.16µs ? ?/sec
filter context fsb with value length 50 high selectivity (kept 1023/1024) 1.00 60.6±0.61µs ? ?/sec 1.17 70.7±0.46µs ? ?/sec
filter context fsb with value length 50 low selectivity (kept 1/1024) 1.00 60.5±0.22µs ? ?/sec 1.17 70.6±0.62µs ? ?/sec
filter context i32 (kept 1/2) 1.00 16.5±0.12µs ? ?/sec 1.02 16.9±0.91µs ? ?/sec
filter context i32 high selectivity (kept 1023/1024) 1.00 6.4±0.31µs ? ?/sec 1.05 6.7±0.44µs ? ?/sec
filter context i32 low selectivity (kept 1/1024) 1.00 220.8±0.94ns ? ?/sec 1.04 228.8±5.85ns ? ?/sec
filter context i32 w NULLs (kept 1/2) 1.00 77.4±0.40µs ? ?/sec 1.14 88.3±3.68µs ? ?/sec
filter context i32 w NULLs high selectivity (kept 1023/1024) 1.03 10.1±0.37µs ? ?/sec 1.00 9.8±0.36µs ? ?/sec
filter context i32 w NULLs low selectivity (kept 1/1024) 1.00 418.8±1.54ns ? ?/sec 1.09 457.8±13.54ns ? ?/sec
filter context mixed string view (kept 1/2) 1.00 103.1±3.66µs ? ?/sec 1.13 116.8±6.95µs ? ?/sec
filter context mixed string view high selectivity (kept 1023/1024) 1.00 53.8±1.72µs ? ?/sec 1.02 55.0±1.75µs ? ?/sec
filter context mixed string view low selectivity (kept 1/1024) 1.00 611.8±37.83ns ? ?/sec 1.04 636.5±9.12ns ? ?/sec
filter context short string view (kept 1/2) 1.00 105.7±6.47µs ? ?/sec 1.11 117.0±5.34µs ? ?/sec
filter context short string view high selectivity (kept 1023/1024) 1.02 55.7±1.12µs ? ?/sec 1.00 54.8±1.46µs ? ?/sec
filter context short string view low selectivity (kept 1/1024) 1.00 471.3±3.05ns ? ?/sec 1.05 495.7±20.48ns ? ?/sec
filter context string (kept 1/2) 1.00 568.6±6.79µs ? ?/sec 1.03 586.1±13.70µs ? ?/sec
filter context string dictionary (kept 1/2) 1.02 17.1±0.11µs ? ?/sec 1.00 16.8±0.07µs ? ?/sec
filter context string dictionary high selectivity (kept 1023/1024) 1.00 6.8±0.44µs ? ?/sec 1.07 7.3±0.44µs ? ?/sec
filter context string dictionary low selectivity (kept 1/1024) 1.00 626.7±10.80ns ? ?/sec 1.04 652.0±17.39ns ? ?/sec
filter context string dictionary w NULLs (kept 1/2) 1.00 78.5±1.55µs ? ?/sec 1.13 88.7±2.15µs ? ?/sec
filter context string dictionary w NULLs high selectivity (kept 1023/1024) 1.04 10.9±0.59µs ? ?/sec 1.00 10.4±0.60µs ? ?/sec
filter context string dictionary w NULLs low selectivity (kept 1/1024) 1.00 855.7±14.15ns ? ?/sec 1.02 875.3±21.35ns ? ?/sec
filter context string high selectivity (kept 1023/1024) 1.04 651.1±19.07µs ? ?/sec 1.00 623.3±11.69µs ? ?/sec
filter context string low selectivity (kept 1/1024) 1.00 995.6±10.53ns ? ?/sec 1.01 1007.4±6.02ns ? ?/sec
filter context u8 (kept 1/2) 1.00 15.0±0.04µs ? ?/sec 1.00 15.0±0.08µs ? ?/sec
filter context u8 high selectivity (kept 1023/1024) 1.05 1907.5±10.01ns ? ?/sec 1.00 1817.9±24.10ns ? ?/sec
filter context u8 low selectivity (kept 1/1024) 1.00 212.2±0.92ns ? ?/sec 1.05 223.8±3.73ns ? ?/sec
filter context u8 w NULLs (kept 1/2) 1.00 75.8±0.16µs ? ?/sec 1.14 86.7±4.34µs ? ?/sec
filter context u8 w NULLs high selectivity (kept 1023/1024) 1.05 5.3±0.05µs ? ?/sec 1.00 5.1±0.10µs ? ?/sec
filter context u8 w NULLs low selectivity (kept 1/1024) 1.00 411.5±1.36ns ? ?/sec 1.10 453.7±8.10ns ? ?/sec
filter decimal128 (kept 1/2) 1.00 47.6±1.84µs ? ?/sec 1.03 49.0±1.86µs ? ?/sec
filter decimal128 high selectivity (kept 1023/1024) 1.05 53.6±1.47µs ? ?/sec 1.00 51.3±1.44µs ? ?/sec
filter decimal128 low selectivity (kept 1/1024) 1.10 3.2±0.02µs ? ?/sec 1.00 2.9±0.07µs ? ?/sec
filter f32 (kept 1/2) 1.00 164.4±0.74µs ? ?/sec 1.00 165.1±3.99µs ? ?/sec
filter fsb with value length 20 (kept 1/2) 1.08 135.6±0.60µs ? ?/sec 1.00 126.0±4.49µs ? ?/sec
filter fsb with value length 20 high selectivity (kept 1023/1024) 1.00 71.7±1.99µs ? ?/sec 1.01 72.8±2.01µs ? ?/sec
filter fsb with value length 20 low selectivity (kept 1/1024) 1.00 2.6±0.02µs ? ?/sec 1.29 3.4±0.02µs ? ?/sec
filter fsb with value length 5 (kept 1/2) 1.17 141.2±0.55µs ? ?/sec 1.00 120.9±1.04µs ? ?/sec
filter fsb with value length 5 high selectivity (kept 1023/1024) 1.04 11.8±0.86µs ? ?/sec 1.00 11.3±0.76µs ? ?/sec
filter fsb with value length 5 low selectivity (kept 1/1024) 1.00 2.6±0.01µs ? ?/sec 1.26 3.3±0.03µs ? ?/sec
filter fsb with value length 50 (kept 1/2) 1.01 169.4±18.08µs ? ?/sec 1.00 168.4±17.50µs ? ?/sec
filter fsb with value length 50 high selectivity (kept 1023/1024) 1.06 215.9±9.54µs ? ?/sec 1.00 203.9±5.08µs ? ?/sec
filter fsb with value length 50 low selectivity (kept 1/1024) 1.00 2.7±0.02µs ? ?/sec 1.26 3.4±0.02µs ? ?/sec
filter i32 (kept 1/2) 1.00 36.5±0.12µs ? ?/sec 1.18 43.2±0.39µs ? ?/sec
filter i32 high selectivity (kept 1023/1024) 1.00 8.8±0.33µs ? ?/sec 1.01 8.9±0.45µs ? ?/sec
filter i32 low selectivity (kept 1/1024) 1.00 2.4±0.01µs ? ?/sec 1.00 2.4±0.02µs ? ?/sec
filter optimize (kept 1/2) 1.00 37.2±0.08µs ? ?/sec 1.46 54.3±2.24µs ? ?/sec
filter optimize high selectivity (kept 1023/1024) 1.11 3.0±0.01µs ? ?/sec 1.00 2.7±0.02µs ? ?/sec
filter optimize low selectivity (kept 1/1024) 1.00 2.7±0.01µs ? ?/sec 1.14 3.1±0.02µs ? ?/sec
filter run array (kept 1/2) 1.22 453.4±4.51µs ? ?/sec 1.00 371.1±4.51µs ? ?/sec
filter run array high selectivity (kept 1023/1024) 1.21 485.0±7.59µs ? ?/sec 1.00 399.3±8.51µs ? ?/sec
filter run array low selectivity (kept 1/1024) 1.30 369.7±1.21µs ? ?/sec 1.00 285.3±3.38µs ? ?/sec
filter single record batch 1.00 52.7±0.62µs ? ?/sec 1.03 54.3±3.72µs ? ?/sec
filter u8 (kept 1/2) 1.17 53.5±0.10µs ? ?/sec 1.00 45.7±0.83µs ? ?/sec
filter u8 high selectivity (kept 1023/1024) 1.06 4.0±0.02µs ? ?/sec 1.00 3.8±0.04µs ? ?/sec
filter u8 low selectivity (kept 1/1024) 1.01 3.3±0.01µs ? ?/sec 1.00 3.3±0.06µs ? ?/sec
```
</p>
</details>
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on code in PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#discussion_r2708314492
##########
arrow-select/src/filter.rs:
##########
@@ -91,6 +91,24 @@ impl<'a> IndexIterator<'a> {
let iter = filter.values().set_indices();
Self { remaining, iter }
}
+
+ /// Collect this iterator as a [`Vec`]
+ pub fn collect(mut self) -> Vec<usize> {
Review Comment:
Faster collect for `IndexIterator` (as we cannot use TrustedLen yet in stable)
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3773188102
🤖 `./gh_compare_arrow.sh` [gh_compare_arrow.sh](https://github.com/alamb/datafusion-benchmarking/blob/main/scripts/gh_compare_arrow.sh) Running
Linux aal-dev 6.14.0-1018-gcp #19~24.04.1-Ubuntu SMP Wed Sep 24 23:23:09 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
Comparing improve_filter_speed4 (7c22b75fce3a3a7b42a3ea385e2b8df920b549fd) to abd11d74bdc179840085bd99d6a697c098e5c751 [diff](https://github.com/apache/arrow-rs/compare/abd11d74bdc179840085bd99d6a697c098e5c751..7c22b75fce3a3a7b42a3ea385e2b8df920b549fd)
BENCH_NAME=arrow_reader_clickbench
BENCH_COMMAND=cargo bench --features=arrow,async,test_common,experimental --bench arrow_reader_clickbench
BENCH_FILTER=
BENCH_BRANCH_NAME=improve_filter_speed4
Results will be posted here when complete
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3773463325
```
filter optimize (kept 1/2) 1.00 37.0±0.11µs ? ?/sec 1.46 53.9±0.24µs ? ?/sec
filter optimize low selectivity (kept 1/1024) 1.00 1836.2±15.23ns ? ?/sec 1.69 3.1±0.03µs ? ?/sec
```
Nice
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter/BitIndexIterator [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772658420
🤖 Hi @Dandandan, thanks for the request (https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772658318).
[`scrape_comments.py`](https://github.com/alamb/datafusion-benchmarking/blob/main/scripts/scrape_comments.py) only supports whitelisted benchmarks.
- Standard: (none)
- Criterion: array_iter, arrow_reader, arrow_reader_clickbench, arrow_reader_row_filter, arrow_statistics, arrow_writer, bitwise_kernel, boolean_kernels, buffer_bit_ops, cast_kernels, coalesce_kernels, comparison_kernels, concatenate_kernel, csv_writer, encoding, filter_kernels, interleave_kernels, json-reader, metadata, row_format, take_kernels, union_array, variant_builder, variant_kernels, variant_validation, view_types, zip_kernels
Please choose one or more of these with `run benchmark <name>` or `run benchmark <name1> <name2>...`
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3773308741
🤖: Benchmark completed
<details><summary>Details</summary>
<p>
```
group improve_filter_speed4 main
----- --------------------- ----
arrow_reader_clickbench/async/Q1 1.00 2.3±0.03ms ? ?/sec 1.01 2.4±0.03ms ? ?/sec
arrow_reader_clickbench/async/Q10 1.05 12.9±0.41ms ? ?/sec 1.00 12.3±0.14ms ? ?/sec
arrow_reader_clickbench/async/Q11 1.02 14.6±0.35ms ? ?/sec 1.00 14.3±0.37ms ? ?/sec
arrow_reader_clickbench/async/Q12 1.02 25.3±0.41ms ? ?/sec 1.00 24.8±0.24ms ? ?/sec
arrow_reader_clickbench/async/Q13 1.01 30.3±0.40ms ? ?/sec 1.00 30.1±0.35ms ? ?/sec
arrow_reader_clickbench/async/Q14 1.01 27.8±0.36ms ? ?/sec 1.00 27.5±1.31ms ? ?/sec
arrow_reader_clickbench/async/Q19 1.00 5.2±0.13ms ? ?/sec 1.01 5.3±0.17ms ? ?/sec
arrow_reader_clickbench/async/Q20 1.01 147.4±1.36ms ? ?/sec 1.00 145.7±3.81ms ? ?/sec
arrow_reader_clickbench/async/Q21 1.00 166.8±1.04ms ? ?/sec 1.00 166.2±1.40ms ? ?/sec
arrow_reader_clickbench/async/Q22 1.00 318.4±22.35ms ? ?/sec 1.00 319.1±15.68ms ? ?/sec
arrow_reader_clickbench/async/Q23 1.00 402.1±3.37ms ? ?/sec 1.00 402.6±2.48ms ? ?/sec
arrow_reader_clickbench/async/Q24 1.00 33.4±0.32ms ? ?/sec 1.01 33.7±0.31ms ? ?/sec
arrow_reader_clickbench/async/Q27 1.00 99.1±1.32ms ? ?/sec 1.00 98.7±0.69ms ? ?/sec
arrow_reader_clickbench/async/Q28 1.00 97.4±0.99ms ? ?/sec 1.00 97.3±0.52ms ? ?/sec
arrow_reader_clickbench/async/Q30 1.05 31.8±0.91ms ? ?/sec 1.00 30.3±0.62ms ? ?/sec
arrow_reader_clickbench/async/Q36 1.01 108.0±1.47ms ? ?/sec 1.00 107.4±0.99ms ? ?/sec
arrow_reader_clickbench/async/Q37 1.01 84.4±1.72ms ? ?/sec 1.00 83.3±0.42ms ? ?/sec
arrow_reader_clickbench/async/Q38 1.03 33.5±0.98ms ? ?/sec 1.00 32.4±0.32ms ? ?/sec
arrow_reader_clickbench/async/Q39 1.05 47.5±1.53ms ? ?/sec 1.00 45.3±0.53ms ? ?/sec
arrow_reader_clickbench/async/Q40 1.13 30.9±0.43ms ? ?/sec 1.00 27.3±1.16ms ? ?/sec
arrow_reader_clickbench/async/Q41 1.14 24.6±0.57ms ? ?/sec 1.00 21.6±0.41ms ? ?/sec
arrow_reader_clickbench/async/Q42 1.07 11.5±0.15ms ? ?/sec 1.00 10.7±0.10ms ? ?/sec
arrow_reader_clickbench/sync/Q1 1.00 2.1±0.04ms ? ?/sec 1.00 2.1±0.02ms ? ?/sec
arrow_reader_clickbench/sync/Q10 1.06 10.1±0.24ms ? ?/sec 1.00 9.6±0.18ms ? ?/sec
arrow_reader_clickbench/sync/Q11 1.04 11.7±0.34ms ? ?/sec 1.00 11.2±0.19ms ? ?/sec
arrow_reader_clickbench/sync/Q12 1.06 34.4±2.01ms ? ?/sec 1.00 32.4±1.17ms ? ?/sec
arrow_reader_clickbench/sync/Q13 1.03 43.7±1.19ms ? ?/sec 1.00 42.3±3.87ms ? ?/sec
arrow_reader_clickbench/sync/Q14 1.00 35.2±0.33ms ? ?/sec 1.16 40.6±0.97ms ? ?/sec
arrow_reader_clickbench/sync/Q19 1.01 4.2±0.06ms ? ?/sec 1.00 4.1±0.02ms ? ?/sec
arrow_reader_clickbench/sync/Q20 1.00 176.6±1.20ms ? ?/sec 1.00 176.6±1.71ms ? ?/sec
arrow_reader_clickbench/sync/Q21 1.04 235.9±1.39ms ? ?/sec 1.00 227.0±2.50ms ? ?/sec
arrow_reader_clickbench/sync/Q22 1.00 474.6±3.30ms ? ?/sec 1.00 472.9±3.15ms ? ?/sec
arrow_reader_clickbench/sync/Q23 1.00 438.5±16.22ms ? ?/sec 1.01 443.8±15.68ms ? ?/sec
arrow_reader_clickbench/sync/Q24 1.00 44.1±0.60ms ? ?/sec 1.00 44.1±1.03ms ? ?/sec
arrow_reader_clickbench/sync/Q27 1.00 153.1±1.22ms ? ?/sec 1.00 152.8±1.17ms ? ?/sec
arrow_reader_clickbench/sync/Q28 1.00 147.9±0.91ms ? ?/sec 1.01 149.8±1.85ms ? ?/sec
arrow_reader_clickbench/sync/Q30 1.01 30.3±0.30ms ? ?/sec 1.00 30.1±0.24ms ? ?/sec
arrow_reader_clickbench/sync/Q36 1.00 151.5±1.46ms ? ?/sec 1.00 152.0±1.43ms ? ?/sec
arrow_reader_clickbench/sync/Q37 1.00 86.9±0.98ms ? ?/sec 1.00 86.8±0.61ms ? ?/sec
arrow_reader_clickbench/sync/Q38 1.00 28.7±0.51ms ? ?/sec 1.00 28.8±0.76ms ? ?/sec
arrow_reader_clickbench/sync/Q39 1.00 33.3±0.26ms ? ?/sec 1.00 33.2±0.23ms ? ?/sec
arrow_reader_clickbench/sync/Q40 1.00 25.5±0.24ms ? ?/sec 1.02 25.9±0.77ms ? ?/sec
arrow_reader_clickbench/sync/Q41 1.00 28.1±0.36ms ? ?/sec 1.01 28.3±0.21ms ? ?/sec
arrow_reader_clickbench/sync/Q42 1.00 12.1±0.17ms ? ?/sec 1.01 12.3±0.25ms ? ?/sec
```
</p>
</details>
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3773187867
🤖: Benchmark completed
<details><summary>Details</summary>
<p>
```
group improve_filter_speed4 main
----- --------------------- ----
arrow_array_reader/BYTE_ARRAY/Decimal128Array/plain encoded, mandatory, no NULLs 1.00 1226.0±6.00µs ? ?/sec 1.01 1239.8±5.52µs ? ?/sec
arrow_array_reader/BYTE_ARRAY/Decimal128Array/plain encoded, optional, half NULLs 1.00 1267.9±8.62µs ? ?/sec 1.01 1283.7±6.79µs ? ?/sec
arrow_array_reader/BYTE_ARRAY/Decimal128Array/plain encoded, optional, no NULLs 1.00 1231.5±6.85µs ? ?/sec 1.01 1249.6±7.74µs ? ?/sec
arrow_array_reader/BinaryArray/dictionary encoded, mandatory, no NULLs 1.00 495.7±10.00µs ? ?/sec 1.02 507.0±12.37µs ? ?/sec
arrow_array_reader/BinaryArray/dictionary encoded, optional, half NULLs 1.00 668.7±11.45µs ? ?/sec 1.01 672.1±20.25µs ? ?/sec
arrow_array_reader/BinaryArray/dictionary encoded, optional, no NULLs 1.00 502.6±13.18µs ? ?/sec 1.02 513.1±10.51µs ? ?/sec
arrow_array_reader/BinaryArray/plain encoded, mandatory, no NULLs 1.02 556.9±15.26µs ? ?/sec 1.00 543.6±8.24µs ? ?/sec
arrow_array_reader/BinaryArray/plain encoded, optional, half NULLs 1.00 727.3±4.32µs ? ?/sec 1.01 732.9±8.67µs ? ?/sec
arrow_array_reader/BinaryArray/plain encoded, optional, no NULLs 1.01 563.9±3.29µs ? ?/sec 1.00 560.3±12.67µs ? ?/sec
arrow_array_reader/BinaryViewArray/dictionary encoded, mandatory, no NULLs 1.10 260.3±5.50µs ? ?/sec 1.00 237.1±2.89µs ? ?/sec
arrow_array_reader/BinaryViewArray/dictionary encoded, optional, half NULLs 1.20 260.4±1.20µs ? ?/sec 1.00 217.2±1.14µs ? ?/sec
arrow_array_reader/BinaryViewArray/dictionary encoded, optional, no NULLs 1.06 257.5±3.50µs ? ?/sec 1.00 244.0±3.87µs ? ?/sec
arrow_array_reader/BinaryViewArray/plain encoded, mandatory, no NULLs 1.00 338.9±3.16µs ? ?/sec 1.03 350.4±4.32µs ? ?/sec
arrow_array_reader/BinaryViewArray/plain encoded, mandatory, no NULLs, short string 1.00 325.4±2.54µs ? ?/sec 1.00 326.0±4.38µs ? ?/sec
arrow_array_reader/BinaryViewArray/plain encoded, optional, half NULLs 1.11 314.9±2.07µs ? ?/sec 1.00 284.2±12.47µs ? ?/sec
arrow_array_reader/BinaryViewArray/plain encoded, optional, no NULLs 1.00 346.7±5.64µs ? ?/sec 1.04 361.1±8.28µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/byte_stream_split encoded, mandatory, no NULLs 1.00 1080.6±9.60µs ? ?/sec 1.05 1133.0±40.01µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/byte_stream_split encoded, optional, half NULLs 1.00 901.2±28.04µs ? ?/sec 1.09 986.1±4.32µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/byte_stream_split encoded, optional, no NULLs 1.00 1088.4±11.75µs ? ?/sec 1.05 1138.4±21.67µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/plain encoded, mandatory, no NULLs 1.00 436.9±7.05µs ? ?/sec 1.05 460.1±7.47µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/plain encoded, optional, half NULLs 1.00 545.5±4.96µs ? ?/sec 1.19 649.8±24.99µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Decimal128Array/plain encoded, optional, no NULLs 1.00 446.4±8.85µs ? ?/sec 1.05 468.0±16.34µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/byte_stream_split encoded, mandatory, no NULLs 1.00 160.8±4.28µs ? ?/sec 1.26 203.1±8.41µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/byte_stream_split encoded, optional, half NULLs 1.00 300.3±2.49µs ? ?/sec 1.05 316.4±4.43µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/byte_stream_split encoded, optional, no NULLs 1.00 166.3±4.67µs ? ?/sec 1.25 207.8±2.74µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/plain encoded, mandatory, no NULLs 1.00 75.8±0.55µs ? ?/sec 1.55 117.6±2.75µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/plain encoded, optional, half NULLs 1.00 257.0±2.70µs ? ?/sec 1.07 274.9±15.49µs ? ?/sec
arrow_array_reader/FIXED_LEN_BYTE_ARRAY/Float16Array/plain encoded, optional, no NULLs 1.00 80.8±0.57µs ? ?/sec 1.52 122.5±1.03µs ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/byte_stream_split encoded, mandatory, no NULLs 1.00 737.5±8.44µs ? ?/sec 1.01 746.0±14.40µs ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/byte_stream_split encoded, optional, half NULLs 1.00 536.0±2.24µs ? ?/sec 1.11 596.0±15.84µs ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/byte_stream_split encoded, optional, no NULLs 1.00 749.1±25.38µs ? ?/sec 1.01 754.8±20.71µs ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/plain encoded, mandatory, no NULLs 1.00 64.2±6.14µs ? ?/sec 1.33 85.7±3.60µs ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/plain encoded, optional, half NULLs 1.00 204.4±2.01µs ? ?/sec 1.26 257.3±2.39µs ? ?/sec
arrow_array_reader/FixedLenByteArray(16)/plain encoded, optional, no NULLs 1.00 72.0±5.09µs ? ?/sec 1.30 93.5±2.92µs ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/byte_stream_split encoded, mandatory, no NULLs 1.00 93.9±0.38µs ? ?/sec 1.00 94.2±0.34µs ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/byte_stream_split encoded, optional, half NULLs 1.11 231.8±1.33µs ? ?/sec 1.00 208.3±1.54µs ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/byte_stream_split encoded, optional, no NULLs 1.00 99.8±4.09µs ? ?/sec 1.00 100.1±2.54µs ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/plain encoded, mandatory, no NULLs 1.00 9.5±0.42µs ? ?/sec 1.02 9.6±0.45µs ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/plain encoded, optional, half NULLs 1.16 190.6±18.18µs ? ?/sec 1.00 164.4±1.83µs ? ?/sec
arrow_array_reader/FixedLenByteArray(2)/plain encoded, optional, no NULLs 1.01 14.6±0.38µs ? ?/sec 1.00 14.5±0.38µs ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/byte_stream_split encoded, mandatory, no NULLs 1.00 183.7±0.74µs ? ?/sec 1.00 184.0±2.49µs ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/byte_stream_split encoded, optional, half NULLs 1.12 368.5±8.07µs ? ?/sec 1.00 329.2±1.84µs ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/byte_stream_split encoded, optional, no NULLs 1.00 190.1±3.61µs ? ?/sec 1.00 189.7±1.04µs ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/plain encoded, mandatory, no NULLs 1.00 13.9±0.38µs ? ?/sec 1.03 14.3±0.29µs ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/plain encoded, optional, half NULLs 1.16 282.7±2.46µs ? ?/sec 1.00 244.6±2.12µs ? ?/sec
arrow_array_reader/FixedLenByteArray(4)/plain encoded, optional, no NULLs 1.02 20.2±0.31µs ? ?/sec 1.00 19.8±0.48µs ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/byte_stream_split encoded, mandatory, no NULLs 1.00 363.5±3.59µs ? ?/sec 1.05 381.0±63.72µs ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/byte_stream_split encoded, optional, half NULLs 1.00 343.0±5.73µs ? ?/sec 1.14 391.0±7.43µs ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/byte_stream_split encoded, optional, no NULLs 1.00 371.5±2.82µs ? ?/sec 1.01 374.1±4.94µs ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/plain encoded, mandatory, no NULLs 1.00 24.7±0.50µs ? ?/sec 1.18 29.1±1.17µs ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/plain encoded, optional, half NULLs 1.00 172.9±1.11µs ? ?/sec 1.27 219.6±3.49µs ? ?/sec
arrow_array_reader/FixedLenByteArray(8)/plain encoded, optional, no NULLs 1.00 32.2±0.50µs ? ?/sec 1.18 38.1±1.62µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/binary packed skip, mandatory, no NULLs 1.00 110.7±0.72µs ? ?/sec 1.00 111.1±0.57µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/binary packed skip, optional, half NULLs 1.00 120.4±0.70µs ? ?/sec 1.08 129.7±0.86µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/binary packed skip, optional, no NULLs 1.00 113.6±0.67µs ? ?/sec 1.00 113.7±0.61µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/binary packed, mandatory, no NULLs 1.00 160.0±0.94µs ? ?/sec 1.01 161.0±5.60µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/binary packed, optional, half NULLs 1.00 200.7±2.56µs ? ?/sec 1.11 222.9±3.27µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/binary packed, optional, no NULLs 1.00 165.3±1.09µs ? ?/sec 1.01 167.5±6.46µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/byte_stream_split encoded, mandatory, no NULLs 1.00 74.8±0.34µs ? ?/sec 1.02 76.0±0.44µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/byte_stream_split encoded, optional, half NULLs 1.00 156.1±2.97µs ? ?/sec 1.13 176.8±4.49µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/byte_stream_split encoded, optional, no NULLs 1.00 80.9±0.51µs ? ?/sec 1.00 81.2±0.70µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/dictionary encoded, mandatory, no NULLs 1.00 134.7±0.57µs ? ?/sec 1.06 143.4±1.01µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/dictionary encoded, optional, half NULLs 1.00 191.3±7.38µs ? ?/sec 1.12 213.6±1.83µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/dictionary encoded, optional, no NULLs 1.00 141.3±1.28µs ? ?/sec 1.05 148.6±3.33µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/plain encoded, mandatory, no NULLs 1.00 70.0±0.42µs ? ?/sec 1.02 71.8±2.24µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/plain encoded, optional, half NULLs 1.00 153.4±1.44µs ? ?/sec 1.15 175.9±4.29µs ? ?/sec
arrow_array_reader/INT32/Decimal128Array/plain encoded, optional, no NULLs 1.00 76.6±0.90µs ? ?/sec 1.01 77.3±0.51µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/binary packed skip, mandatory, no NULLs 1.01 109.8±1.04µs ? ?/sec 1.00 108.8±1.57µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/binary packed skip, optional, half NULLs 1.01 131.8±1.96µs ? ?/sec 1.00 130.4±5.41µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/binary packed skip, optional, no NULLs 1.00 111.1±0.80µs ? ?/sec 1.01 111.8±0.61µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/binary packed, mandatory, no NULLs 1.00 162.0±5.77µs ? ?/sec 1.00 162.3±1.00µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/binary packed, optional, half NULLs 1.02 233.6±1.40µs ? ?/sec 1.00 228.5±2.45µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/binary packed, optional, no NULLs 1.00 166.9±2.17µs ? ?/sec 1.01 168.6±5.88µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/byte_stream_split encoded, mandatory, no NULLs 1.01 200.6±1.01µs ? ?/sec 1.00 199.3±3.45µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/byte_stream_split encoded, optional, half NULLs 1.02 248.7±0.88µs ? ?/sec 1.00 243.4±3.32µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/byte_stream_split encoded, optional, no NULLs 1.01 206.9±1.18µs ? ?/sec 1.00 205.1±0.76µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/dictionary encoded, mandatory, no NULLs 1.07 152.5±6.11µs ? ?/sec 1.00 141.9±0.78µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/dictionary encoded, optional, half NULLs 1.06 223.5±1.40µs ? ?/sec 1.00 211.3±2.37µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/dictionary encoded, optional, no NULLs 1.09 158.6±1.30µs ? ?/sec 1.00 145.5±0.67µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/plain encoded, mandatory, no NULLs 1.02 108.1±1.76µs ? ?/sec 1.00 105.7±1.75µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/plain encoded, optional, half NULLs 1.02 198.3±1.96µs ? ?/sec 1.00 195.2±1.18µs ? ?/sec
arrow_array_reader/INT64/Decimal128Array/plain encoded, optional, no NULLs 1.02 115.8±1.49µs ? ?/sec 1.00 113.9±1.39µs ? ?/sec
arrow_array_reader/Int16Array/binary packed skip, mandatory, no NULLs 1.00 78.8±1.04µs ? ?/sec 1.00 78.8±0.77µs ? ?/sec
arrow_array_reader/Int16Array/binary packed skip, optional, half NULLs 1.00 92.0±1.26µs ? ?/sec 1.13 103.9±0.52µs ? ?/sec
arrow_array_reader/Int16Array/binary packed skip, optional, no NULLs 1.00 81.1±0.95µs ? ?/sec 1.01 81.6±0.75µs ? ?/sec
arrow_array_reader/Int16Array/binary packed, mandatory, no NULLs 1.00 108.1±1.11µs ? ?/sec 1.00 108.3±0.57µs ? ?/sec
arrow_array_reader/Int16Array/binary packed, optional, half NULLs 1.00 152.4±2.77µs ? ?/sec 1.16 176.1±6.68µs ? ?/sec
arrow_array_reader/Int16Array/binary packed, optional, no NULLs 1.00 112.4±0.81µs ? ?/sec 1.01 113.3±2.08µs ? ?/sec
arrow_array_reader/Int16Array/byte_stream_split encoded, mandatory, no NULLs 1.04 42.4±0.16µs ? ?/sec 1.00 40.7±0.15µs ? ?/sec
arrow_array_reader/Int16Array/byte_stream_split encoded, optional, half NULLs 1.00 116.7±1.58µs ? ?/sec 1.23 143.7±12.77µs ? ?/sec
arrow_array_reader/Int16Array/byte_stream_split encoded, optional, no NULLs 1.03 46.7±0.40µs ? ?/sec 1.00 45.5±0.15µs ? ?/sec
arrow_array_reader/Int16Array/dictionary encoded, mandatory, no NULLs 1.00 101.8±0.63µs ? ?/sec 1.08 110.2±1.06µs ? ?/sec
arrow_array_reader/Int16Array/dictionary encoded, optional, half NULLs 1.00 153.5±1.62µs ? ?/sec 1.17 179.2±1.44µs ? ?/sec
arrow_array_reader/Int16Array/dictionary encoded, optional, no NULLs 1.00 107.5±1.56µs ? ?/sec 1.07 114.6±0.60µs ? ?/sec
arrow_array_reader/Int16Array/plain encoded, mandatory, no NULLs 1.00 36.7±0.25µs ? ?/sec 1.00 36.7±0.15µs ? ?/sec
arrow_array_reader/Int16Array/plain encoded, optional, half NULLs 1.00 114.7±0.45µs ? ?/sec 1.20 138.1±0.64µs ? ?/sec
arrow_array_reader/Int16Array/plain encoded, optional, no NULLs 1.00 41.5±0.50µs ? ?/sec 1.00 41.6±0.49µs ? ?/sec
arrow_array_reader/Int32Array/binary packed skip, mandatory, no NULLs 1.00 84.3±0.77µs ? ?/sec 1.00 84.2±0.55µs ? ?/sec
arrow_array_reader/Int32Array/binary packed skip, optional, half NULLs 1.00 90.8±2.13µs ? ?/sec 1.13 102.4±0.73µs ? ?/sec
arrow_array_reader/Int32Array/binary packed skip, optional, no NULLs 1.01 87.8±1.30µs ? ?/sec 1.00 87.2±2.09µs ? ?/sec
arrow_array_reader/Int32Array/binary packed, mandatory, no NULLs 1.00 108.3±0.73µs ? ?/sec 1.00 108.7±1.85µs ? ?/sec
arrow_array_reader/Int32Array/binary packed, optional, half NULLs 1.00 143.2±1.29µs ? ?/sec 1.17 167.0±1.24µs ? ?/sec
arrow_array_reader/Int32Array/binary packed, optional, no NULLs 1.00 112.1±0.51µs ? ?/sec 1.01 113.2±0.80µs ? ?/sec
arrow_array_reader/Int32Array/byte_stream_split encoded, mandatory, no NULLs 1.03 23.6±0.38µs ? ?/sec 1.00 23.0±0.19µs ? ?/sec
arrow_array_reader/Int32Array/byte_stream_split encoded, optional, half NULLs 1.00 99.5±4.74µs ? ?/sec 1.22 121.4±1.14µs ? ?/sec
arrow_array_reader/Int32Array/byte_stream_split encoded, optional, no NULLs 1.02 28.2±0.90µs ? ?/sec 1.00 27.7±0.28µs ? ?/sec
arrow_array_reader/Int32Array/dictionary encoded, mandatory, no NULLs 1.00 83.2±0.81µs ? ?/sec 1.10 91.4±1.51µs ? ?/sec
arrow_array_reader/Int32Array/dictionary encoded, optional, half NULLs 1.00 134.0±1.88µs ? ?/sec 1.20 160.3±5.60µs ? ?/sec
arrow_array_reader/Int32Array/dictionary encoded, optional, no NULLs 1.00 88.5±0.76µs ? ?/sec 1.07 95.1±0.52µs ? ?/sec
arrow_array_reader/Int32Array/plain encoded, mandatory, no NULLs 1.03 15.8±1.28µs ? ?/sec 1.00 15.2±0.47µs ? ?/sec
arrow_array_reader/Int32Array/plain encoded, optional, half NULLs 1.00 96.5±1.48µs ? ?/sec 1.24 120.0±6.07µs ? ?/sec
arrow_array_reader/Int32Array/plain encoded, optional, no NULLs 1.00 21.1±0.55µs ? ?/sec 1.00 21.1±0.39µs ? ?/sec
arrow_array_reader/Int64Array/binary packed skip, mandatory, no NULLs 1.00 80.2±1.13µs ? ?/sec 1.00 80.2±0.71µs ? ?/sec
arrow_array_reader/Int64Array/binary packed skip, optional, half NULLs 1.01 102.2±3.33µs ? ?/sec 1.00 101.1±4.06µs ? ?/sec
arrow_array_reader/Int64Array/binary packed skip, optional, no NULLs 1.00 82.5±0.40µs ? ?/sec 1.00 82.7±0.87µs ? ?/sec
arrow_array_reader/Int64Array/binary packed, mandatory, no NULLs 1.00 105.9±1.58µs ? ?/sec 1.02 108.1±1.91µs ? ?/sec
arrow_array_reader/Int64Array/binary packed, optional, half NULLs 1.06 173.6±1.75µs ? ?/sec 1.00 163.6±1.24µs ? ?/sec
arrow_array_reader/Int64Array/binary packed, optional, no NULLs 1.01 112.1±1.76µs ? ?/sec 1.00 110.7±2.31µs ? ?/sec
arrow_array_reader/Int64Array/byte_stream_split encoded, mandatory, no NULLs 1.00 145.2±0.88µs ? ?/sec 1.01 146.8±1.61µs ? ?/sec
arrow_array_reader/Int64Array/byte_stream_split encoded, optional, half NULLs 1.02 192.3±5.18µs ? ?/sec 1.00 189.0±6.49µs ? ?/sec
arrow_array_reader/Int64Array/byte_stream_split encoded, optional, no NULLs 1.00 150.7±0.43µs ? ?/sec 1.01 152.1±0.45µs ? ?/sec
arrow_array_reader/Int64Array/dictionary encoded, mandatory, no NULLs 1.10 97.1±0.66µs ? ?/sec 1.00 88.0±0.93µs ? ?/sec
arrow_array_reader/Int64Array/dictionary encoded, optional, half NULLs 1.05 164.1±1.53µs ? ?/sec 1.00 155.5±0.80µs ? ?/sec
arrow_array_reader/Int64Array/dictionary encoded, optional, no NULLs 1.11 103.4±1.56µs ? ?/sec 1.00 93.0±0.65µs ? ?/sec
arrow_array_reader/Int64Array/plain encoded, mandatory, no NULLs 1.21 46.5±2.76µs ? ?/sec 1.00 38.4±0.82µs ? ?/sec
arrow_array_reader/Int64Array/plain encoded, optional, half NULLs 1.03 137.1±2.61µs ? ?/sec 1.00 133.4±0.85µs ? ?/sec
arrow_array_reader/Int64Array/plain encoded, optional, no NULLs 1.26 54.2±3.54µs ? ?/sec 1.00 43.2±0.87µs ? ?/sec
arrow_array_reader/Int8Array/binary packed skip, mandatory, no NULLs 1.00 82.6±0.70µs ? ?/sec 1.00 82.8±0.68µs ? ?/sec
arrow_array_reader/Int8Array/binary packed skip, optional, half NULLs 1.00 92.1±1.20µs ? ?/sec 1.13 104.2±1.58µs ? ?/sec
arrow_array_reader/Int8Array/binary packed skip, optional, no NULLs 1.00 85.0±0.34µs ? ?/sec 1.00 85.3±0.24µs ? ?/sec
arrow_array_reader/Int8Array/binary packed, mandatory, no NULLs 1.00 110.3±1.83µs ? ?/sec 1.01 111.3±8.03µs ? ?/sec
arrow_array_reader/Int8Array/binary packed, optional, half NULLs 1.00 149.2±2.40µs ? ?/sec 1.16 172.9±3.03µs ? ?/sec
arrow_array_reader/Int8Array/binary packed, optional, no NULLs 1.00 114.4±1.45µs ? ?/sec 1.00 115.0±0.65µs ? ?/sec
arrow_array_reader/Int8Array/byte_stream_split encoded, mandatory, no NULLs 1.00 32.9±0.17µs ? ?/sec 1.05 34.6±0.98µs ? ?/sec
arrow_array_reader/Int8Array/byte_stream_split encoded, optional, half NULLs 1.00 110.2±1.55µs ? ?/sec 1.21 133.3±1.10µs ? ?/sec
arrow_array_reader/Int8Array/byte_stream_split encoded, optional, no NULLs 1.00 37.5±0.58µs ? ?/sec 1.05 39.3±0.48µs ? ?/sec
arrow_array_reader/Int8Array/dictionary encoded, mandatory, no NULLs 1.00 94.1±3.78µs ? ?/sec 1.09 102.3±2.19µs ? ?/sec
arrow_array_reader/Int8Array/dictionary encoded, optional, half NULLs 1.00 144.9±0.53µs ? ?/sec 1.18 171.3±2.47µs ? ?/sec
arrow_array_reader/Int8Array/dictionary encoded, optional, no NULLs 1.00 98.6±1.09µs ? ?/sec 1.08 106.9±0.40µs ? ?/sec
arrow_array_reader/Int8Array/plain encoded, mandatory, no NULLs 1.00 28.9±0.32µs ? ?/sec 1.00 28.7±0.19µs ? ?/sec
arrow_array_reader/Int8Array/plain encoded, optional, half NULLs 1.00 107.1±0.63µs ? ?/sec 1.22 131.0±5.92µs ? ?/sec
arrow_array_reader/Int8Array/plain encoded, optional, no NULLs 1.00 33.5±0.24µs ? ?/sec 1.01 33.8±1.00µs ? ?/sec
arrow_array_reader/ListArray/plain encoded optional strings half NULLs 1.01 6.0±0.03ms ? ?/sec 1.00 6.0±0.04ms ? ?/sec
arrow_array_reader/ListArray/plain encoded optional strings no NULLs 1.00 11.6±0.12ms ? ?/sec 1.01 11.7±0.17ms ? ?/sec
arrow_array_reader/StringArray/dictionary encoded, mandatory, no NULLs 1.00 500.9±5.28µs ? ?/sec 1.04 521.9±29.98µs ? ?/sec
arrow_array_reader/StringArray/dictionary encoded, optional, half NULLs 1.00 667.8±3.48µs ? ?/sec 1.01 672.4±13.03µs ? ?/sec
arrow_array_reader/StringArray/dictionary encoded, optional, no NULLs 1.00 501.5±5.15µs ? ?/sec 1.02 509.2±6.19µs ? ?/sec
arrow_array_reader/StringArray/plain encoded, mandatory, no NULLs 1.05 680.5±8.22µs ? ?/sec 1.00 647.3±20.28µs ? ?/sec
arrow_array_reader/StringArray/plain encoded, optional, half NULLs 1.01 787.7±10.32µs ? ?/sec 1.00 779.7±10.31µs ? ?/sec
arrow_array_reader/StringArray/plain encoded, optional, no NULLs 1.05 693.8±13.21µs ? ?/sec 1.00 659.4±5.62µs ? ?/sec
arrow_array_reader/StringDictionary/dictionary encoded, mandatory, no NULLs 1.01 333.6±3.80µs ? ?/sec 1.00 330.6±3.28µs ? ?/sec
arrow_array_reader/StringDictionary/dictionary encoded, optional, half NULLs 1.03 386.7±6.52µs ? ?/sec 1.00 375.4±22.05µs ? ?/sec
arrow_array_reader/StringDictionary/dictionary encoded, optional, no NULLs 1.02 343.5±15.64µs ? ?/sec 1.00 337.2±8.69µs ? ?/sec
arrow_array_reader/StringViewArray/dictionary encoded, mandatory, no NULLs 1.04 248.8±3.34µs ? ?/sec 1.00 238.5±3.39µs ? ?/sec
arrow_array_reader/StringViewArray/dictionary encoded, optional, half NULLs 1.20 261.5±3.41µs ? ?/sec 1.00 218.0±4.38µs ? ?/sec
arrow_array_reader/StringViewArray/dictionary encoded, optional, no NULLs 1.06 258.3±3.04µs ? ?/sec 1.00 244.6±5.91µs ? ?/sec
arrow_array_reader/StringViewArray/plain encoded, mandatory, no NULLs 1.00 475.4±16.42µs ? ?/sec 1.05 499.0±13.60µs ? ?/sec
arrow_array_reader/StringViewArray/plain encoded, optional, half NULLs 1.07 376.6±2.00µs ? ?/sec 1.00 353.0±3.00µs ? ?/sec
arrow_array_reader/StringViewArray/plain encoded, optional, no NULLs 1.00 481.6±1.99µs ? ?/sec 1.06 509.5±13.58µs ? ?/sec
arrow_array_reader/UInt16Array/binary packed skip, mandatory, no NULLs 1.00 94.5±3.81µs ? ?/sec 1.00 94.1±0.55µs ? ?/sec
arrow_array_reader/UInt16Array/binary packed skip, optional, half NULLs 1.00 100.3±1.61µs ? ?/sec 1.12 112.2±0.91µs ? ?/sec
arrow_array_reader/UInt16Array/binary packed skip, optional, no NULLs 1.00 97.0±0.83µs ? ?/sec 1.01 97.5±1.77µs ? ?/sec
arrow_array_reader/UInt16Array/binary packed, mandatory, no NULLs 1.00 127.9±2.17µs ? ?/sec 1.00 127.6±0.69µs ? ?/sec
arrow_array_reader/UInt16Array/binary packed, optional, half NULLs 1.00 163.4±2.42µs ? ?/sec 1.15 187.3±4.12µs ? ?/sec
arrow_array_reader/UInt16Array/binary packed, optional, no NULLs 1.00 131.9±0.61µs ? ?/sec 1.00 132.2±0.84µs ? ?/sec
arrow_array_reader/UInt16Array/byte_stream_split encoded, mandatory, no NULLs 1.04 42.5±0.46µs ? ?/sec 1.00 40.8±0.09µs ? ?/sec
arrow_array_reader/UInt16Array/byte_stream_split encoded, optional, half NULLs 1.00 118.1±0.81µs ? ?/sec 1.19 140.8±0.59µs ? ?/sec
arrow_array_reader/UInt16Array/byte_stream_split encoded, optional, no NULLs 1.03 46.8±0.78µs ? ?/sec 1.00 45.5±0.43µs ? ?/sec
arrow_array_reader/UInt16Array/dictionary encoded, mandatory, no NULLs 1.00 101.9±1.02µs ? ?/sec 1.08 109.8±0.86µs ? ?/sec
arrow_array_reader/UInt16Array/dictionary encoded, optional, half NULLs 1.00 153.3±2.20µs ? ?/sec 1.17 178.7±0.78µs ? ?/sec
arrow_array_reader/UInt16Array/dictionary encoded, optional, no NULLs 1.00 106.7±0.74µs ? ?/sec 1.08 115.2±3.02µs ? ?/sec
arrow_array_reader/UInt16Array/plain encoded, mandatory, no NULLs 1.00 36.9±0.20µs ? ?/sec 1.00 36.8±0.23µs ? ?/sec
arrow_array_reader/UInt16Array/plain encoded, optional, half NULLs 1.00 114.6±0.42µs ? ?/sec 1.21 138.5±1.24µs ? ?/sec
arrow_array_reader/UInt16Array/plain encoded, optional, no NULLs 1.00 41.6±0.20µs ? ?/sec 1.00 41.7±0.19µs ? ?/sec
arrow_array_reader/UInt32Array/binary packed skip, mandatory, no NULLs 1.01 84.4±2.12µs ? ?/sec 1.00 84.0±0.31µs ? ?/sec
arrow_array_reader/UInt32Array/binary packed skip, optional, half NULLs 1.00 90.3±0.49µs ? ?/sec 1.13 102.4±0.41µs ? ?/sec
arrow_array_reader/UInt32Array/binary packed skip, optional, no NULLs 1.00 86.6±0.80µs ? ?/sec 1.00 86.8±0.28µs ? ?/sec
arrow_array_reader/UInt32Array/binary packed, mandatory, no NULLs 1.00 109.0±1.40µs ? ?/sec 1.00 108.7±1.23µs ? ?/sec
arrow_array_reader/UInt32Array/binary packed, optional, half NULLs 1.00 143.6±2.71µs ? ?/sec 1.17 167.7±1.22µs ? ?/sec
arrow_array_reader/UInt32Array/binary packed, optional, no NULLs 1.00 112.5±0.48µs ? ?/sec 1.01 113.4±1.59µs ? ?/sec
arrow_array_reader/UInt32Array/byte_stream_split encoded, mandatory, no NULLs 1.00 22.1±0.42µs ? ?/sec 1.00 22.1±0.43µs ? ?/sec
arrow_array_reader/UInt32Array/byte_stream_split encoded, optional, half NULLs 1.00 99.6±0.73µs ? ?/sec 1.23 122.6±5.58µs ? ?/sec
arrow_array_reader/UInt32Array/byte_stream_split encoded, optional, no NULLs 1.00 26.7±0.49µs ? ?/sec 1.01 27.1±0.37µs ? ?/sec
arrow_array_reader/UInt32Array/dictionary encoded, mandatory, no NULLs 1.00 83.1±0.65µs ? ?/sec 1.10 91.4±1.30µs ? ?/sec
arrow_array_reader/UInt32Array/dictionary encoded, optional, half NULLs 1.00 134.2±2.07µs ? ?/sec 1.19 160.2±1.26µs ? ?/sec
arrow_array_reader/UInt32Array/dictionary encoded, optional, no NULLs 1.00 89.8±5.76µs ? ?/sec 1.08 96.8±2.94µs ? ?/sec
arrow_array_reader/UInt32Array/plain encoded, mandatory, no NULLs 1.07 19.4±1.37µs ? ?/sec 1.00 18.1±0.84µs ? ?/sec
arrow_array_reader/UInt32Array/plain encoded, optional, half NULLs 1.00 97.4±9.86µs ? ?/sec 1.23 119.7±2.56µs ? ?/sec
arrow_array_reader/UInt32Array/plain encoded, optional, no NULLs 1.01 23.7±1.12µs ? ?/sec 1.00 23.5±0.93µs ? ?/sec
arrow_array_reader/UInt64Array/binary packed skip, mandatory, no NULLs 1.00 80.1±0.64µs ? ?/sec 1.01 80.5±0.43µs ? ?/sec
arrow_array_reader/UInt64Array/binary packed skip, optional, half NULLs 1.00 101.5±0.74µs ? ?/sec 1.00 101.1±0.63µs ? ?/sec
arrow_array_reader/UInt64Array/binary packed skip, optional, no NULLs 1.00 82.4±1.08µs ? ?/sec 1.01 83.3±0.72µs ? ?/sec
arrow_array_reader/UInt64Array/binary packed, mandatory, no NULLs 1.00 107.9±1.94µs ? ?/sec 1.01 108.6±5.56µs ? ?/sec
arrow_array_reader/UInt64Array/binary packed, optional, half NULLs 1.04 172.5±0.97µs ? ?/sec 1.00 166.1±1.32µs ? ?/sec
arrow_array_reader/UInt64Array/binary packed, optional, no NULLs 1.01 112.1±1.37µs ? ?/sec 1.00 111.0±2.56µs ? ?/sec
arrow_array_reader/UInt64Array/byte_stream_split encoded, mandatory, no NULLs 1.00 146.0±0.65µs ? ?/sec 1.01 146.9±0.58µs ? ?/sec
arrow_array_reader/UInt64Array/byte_stream_split encoded, optional, half NULLs 1.02 191.1±1.85µs ? ?/sec 1.00 188.1±1.00µs ? ?/sec
arrow_array_reader/UInt64Array/byte_stream_split encoded, optional, no NULLs 1.00 151.3±2.99µs ? ?/sec 1.01 152.3±1.04µs ? ?/sec
arrow_array_reader/UInt64Array/dictionary encoded, mandatory, no NULLs 1.10 97.7±4.42µs ? ?/sec 1.00 88.5±0.72µs ? ?/sec
arrow_array_reader/UInt64Array/dictionary encoded, optional, half NULLs 1.05 164.3±3.41µs ? ?/sec 1.00 157.2±2.29µs ? ?/sec
arrow_array_reader/UInt64Array/dictionary encoded, optional, no NULLs 1.12 104.1±1.15µs ? ?/sec 1.00 92.8±0.57µs ? ?/sec
arrow_array_reader/UInt64Array/plain encoded, mandatory, no NULLs 1.30 49.3±2.74µs ? ?/sec 1.00 38.0±0.80µs ? ?/sec
arrow_array_reader/UInt64Array/plain encoded, optional, half NULLs 1.02 137.1±2.06µs ? ?/sec 1.00 134.3±6.78µs ? ?/sec
arrow_array_reader/UInt64Array/plain encoded, optional, no NULLs 1.11 52.6±2.88µs ? ?/sec 1.00 47.3±0.84µs ? ?/sec
arrow_array_reader/UInt8Array/binary packed skip, mandatory, no NULLs 1.00 90.6±2.97µs ? ?/sec 1.02 92.6±4.86µs ? ?/sec
arrow_array_reader/UInt8Array/binary packed skip, optional, half NULLs 1.00 96.6±2.43µs ? ?/sec 1.12 107.9±0.92µs ? ?/sec
arrow_array_reader/UInt8Array/binary packed skip, optional, no NULLs 1.00 92.7±0.74µs ? ?/sec 1.02 94.2±1.10µs ? ?/sec
arrow_array_reader/UInt8Array/binary packed, mandatory, no NULLs 1.00 119.9±2.17µs ? ?/sec 1.00 119.9±1.21µs ? ?/sec
arrow_array_reader/UInt8Array/binary packed, optional, half NULLs 1.00 155.5±1.71µs ? ?/sec 1.15 178.7±3.09µs ? ?/sec
arrow_array_reader/UInt8Array/binary packed, optional, no NULLs 1.00 124.1±1.27µs ? ?/sec 1.00 124.6±1.80µs ? ?/sec
arrow_array_reader/UInt8Array/byte_stream_split encoded, mandatory, no NULLs 1.01 35.1±1.37µs ? ?/sec 1.00 34.6±0.38µs ? ?/sec
arrow_array_reader/UInt8Array/byte_stream_split encoded, optional, half NULLs 1.00 110.0±0.32µs ? ?/sec 1.21 132.8±1.54µs ? ?/sec
arrow_array_reader/UInt8Array/byte_stream_split encoded, optional, no NULLs 1.00 38.8±0.34µs ? ?/sec 1.01 39.2±0.76µs ? ?/sec
arrow_array_reader/UInt8Array/dictionary encoded, mandatory, no NULLs 1.00 94.4±1.36µs ? ?/sec 1.08 102.0±0.39µs ? ?/sec
arrow_array_reader/UInt8Array/dictionary encoded, optional, half NULLs 1.00 145.5±1.02µs ? ?/sec 1.18 171.5±0.75µs ? ?/sec
arrow_array_reader/UInt8Array/dictionary encoded, optional, no NULLs 1.00 99.1±0.75µs ? ?/sec 1.08 107.2±0.43µs ? ?/sec
arrow_array_reader/UInt8Array/plain encoded, mandatory, no NULLs 1.00 28.8±0.32µs ? ?/sec 1.00 28.7±0.11µs ? ?/sec
arrow_array_reader/UInt8Array/plain encoded, optional, half NULLs 1.00 107.7±3.40µs ? ?/sec 1.21 130.4±0.52µs ? ?/sec
arrow_array_reader/UInt8Array/plain encoded, optional, no NULLs 1.00 33.3±0.33µs ? ?/sec 1.01 33.7±0.28µs ? ?/sec
arrow_array_reader/struct/Int32Array/plain encoded, mandatory struct, optional data, half NULLs 1.00 96.1±1.24µs ? ?/sec 1.24 119.5±1.03µs ? ?/sec
arrow_array_reader/struct/Int32Array/plain encoded, mandatory struct, optional data, no NULLs 1.00 23.2±1.23µs ? ?/sec 1.04 24.2±1.18µs ? ?/sec
arrow_array_reader/struct/Int32Array/plain encoded, optional struct, optional data, half NULLs 1.00 223.6±1.57µs ? ?/sec 1.10 245.8±3.27µs ? ?/sec
arrow_array_reader/struct/Int32Array/plain encoded, optional struct, optional data, no NULLs 1.00 126.1±1.55µs ? ?/sec 1.00 125.6±1.53µs ? ?/sec
```
</p>
</details>
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter/BitIndexIterator/iter_set_bits_rev [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772664609
run benchmark filter_kernels
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter/BitIndexIterator [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772658318
run benchmarks filter_kernels
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb-ghbot (via GitHub)" <gi...@apache.org>.
alamb-ghbot commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772837229
🤖 `./gh_compare_arrow.sh` [gh_compare_arrow.sh](https://github.com/alamb/datafusion-benchmarking/blob/main/scripts/gh_compare_arrow.sh) Running
Linux aal-dev 6.14.0-1018-gcp #19~24.04.1-Ubuntu SMP Wed Sep 24 23:23:09 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
Comparing improve_filter_speed4 (9a861829443c45018037f0081ed609bdfcb69177) to abd11d74bdc179840085bd99d6a697c098e5c751 [diff](https://github.com/apache/arrow-rs/compare/abd11d74bdc179840085bd99d6a697c098e5c751..9a861829443c45018037f0081ed609bdfcb69177)
BENCH_NAME=arrow_reader
BENCH_COMMAND=cargo bench --features=arrow,async,test_common,experimental --bench arrow_reader
BENCH_FILTER=
BENCH_BRANCH_NAME=improve_filter_speed4
Results will be posted here when complete
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772899695
> Benchmark completed
Seems it might not give the speed up on intel cpu (or at least not on the default cpu target). `Optimize` seems quite a bit faster though (waiting on the second run which should be faster).
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on code in PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#discussion_r2708452569
##########
arrow-select/src/filter.rs:
##########
@@ -91,6 +91,24 @@ impl<'a> IndexIterator<'a> {
let iter = filter.values().set_indices();
Self { remaining, iter }
}
+
+ /// Collect this iterator as a [`Vec`]
+ pub fn collect(mut self) -> Vec<usize> {
+ let len = self.remaining;
+ let mut result = Vec::with_capacity(len);
+ let ptr: *mut usize = result.as_mut_ptr();
+ for i in 0..len {
+ // SAFETY: we have allocated enough space in `result`
Review Comment:
I think it also relies on self.iter.next() being correctly synchronized with self.remaining.
Maybe we can add a debug assert to validate that assertion ?
##########
arrow-buffer/src/util/bit_iterator.rs:
##########
@@ -305,12 +305,12 @@ impl<'a> BitIndexIterator<'a> {
impl Iterator for BitIndexIterator<'_> {
type Item = usize;
- #[inline]
+ #[inline(always)]
fn next(&mut self) -> Option<Self::Item> {
loop {
if self.current_chunk != 0 {
let bit_pos = self.current_chunk.trailing_zeros();
- self.current_chunk ^= 1 << bit_pos;
+ self.current_chunk &= self.current_chunk - 1;
Review Comment:
fascinating
##########
arrow-select/src/filter.rs:
##########
@@ -91,6 +91,24 @@ impl<'a> IndexIterator<'a> {
let iter = filter.values().set_indices();
Self { remaining, iter }
}
+
+ /// Collect this iterator as a [`Vec`]
+ pub fn collect(mut self) -> Vec<usize> {
Review Comment:
Maybe it is worth a comment explaining that we can avoid this code when TrustedLengh is stable
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on code in PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#discussion_r2708505356
##########
arrow-select/src/filter.rs:
##########
@@ -91,6 +91,24 @@ impl<'a> IndexIterator<'a> {
let iter = filter.values().set_indices();
Self { remaining, iter }
}
+
+ /// Collect this iterator as a [`Vec`]
+ pub fn collect(mut self) -> Vec<usize> {
+ let len = self.remaining;
+ let mut result = Vec::with_capacity(len);
+ let ptr: *mut usize = result.as_mut_ptr();
+ for i in 0..len {
+ // SAFETY: we have allocated enough space in `result`
Review Comment:
Done!
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to ~1.5x) `FilterBuilder::Optimize`/`BitIndexIterator`/`iter_set_bits_rev` [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan merged PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter/BitIndexIterator/iter_set_bits_rev [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#issuecomment-3772668148
run benchmark arrow_reader arrow_reader_clickbench
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] Speedup filter (up to 1.5x) filter/BitIndexIterator/iter_set_bits_rev [arrow-rs]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on code in PR #9229:
URL: https://github.com/apache/arrow-rs/pull/9229#discussion_r2708247266
##########
parquet/src/arrow/buffer/bit_util.rs:
##########
@@ -36,14 +36,15 @@ pub fn iter_set_bits_rev(bytes: &[u8]) -> impl Iterator<Item = usize> + '_ {
.chain(unaligned.chunks().iter().cloned())
.chain(unaligned.suffix());
- iter.rev().flat_map(move |mut chunk| {
+ iter.rev().flat_map(move |chunk| {
let chunk_idx = chunk_end_idx - 64;
chunk_end_idx = chunk_idx;
+ let mut rev_chunk = chunk.reverse_bits();
std::iter::from_fn(move || {
- if chunk != 0 {
- let bit_pos = 63 - chunk.leading_zeros();
- chunk ^= 1 << bit_pos;
- return Some(chunk_idx + (bit_pos as usize));
+ if rev_chunk != 0 {
+ let bit_pos = rev_chunk.trailing_zeros();
+ rev_chunk &= rev_chunk - 1;
Review Comment:
Applying the same optimization here
--
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: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org