Skip to content

Improve take on Chunked(FSL) and FSL(Chunked) array alongside implementation of take using PiecewiseSequence - #8881

Merged
robert3005 merged 8 commits into
developfrom
rk/chunkedfsltake
Aug 7, 2026
Merged

Improve take on Chunked(FSL) and FSL(Chunked) array alongside implementation of take using PiecewiseSequence#8881
robert3005 merged 8 commits into
developfrom
rk/chunkedfsltake

fixes

ffef2a3
Select commit
Loading
Failed to load commit list.
CodSpeed HQ / CodSpeed Performance Analysis failed Aug 7, 2026 in 0s

1 benchmark regressed

⚡ 26 improved benchmarks
❌ 1 regressed benchmark
✅ 1902 untouched benchmarks
⏩ 51 skipped benchmarks1

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

Mode Benchmark BASE HEAD Efficiency
❌ Simulation compress_fsst[(10000, 64, 8)] 9.6 ms 10.7 ms -10.25%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=16384/indices=1] 1,079.5 µs 75.1 µs ×14
⚡ Simulation take[value_shape/duplicate99/fsl8/nonnull/chunks=16/indices=10000] 34.5 ms 10.5 ms ×3.3
⚡ Simulation take[value_shape/shuffled/fsl8/nonnull/chunks=16/indices=10000] 36.3 ms 11.7 ms ×3.1
⚡ Simulation take_chunked_fsl_random[32, 64] 1,115.1 µs 503.2 µs ×2.2
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=1024/indices=1] 164.8 µs 74.6 µs ×2.2
⚡ Simulation take[core/sorted/primitive/nonnull/chunks=16/indices=100000] 29.8 ms 14.3 ms ×2.1
⚡ Simulation fsl_large 200.6 µs 106.1 µs +88.99%
⚡ Simulation take_chunked_fsl_random[16, 64] 865.5 µs 462.5 µs +87.12%
⚡ Simulation take_chunked_fsl_random[8, 64] 737.2 µs 438.1 µs +68.27%
⚡ Simulation take[core/sorted/primitive/nonnull/chunks=16/indices=10000] 2.7 ms 1.6 ms +67.89%
⚡ Simulation take_fsl_chunked_elements_sorted[32, 64] 710.6 µs 429.9 µs +65.31%
⚡ Simulation take_fsl_chunked_elements_random[32, 64] 743.4 µs 454.4 µs +63.61%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=1024/indices=1000] 7.7 ms 5.8 ms +33.57%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=1024/indices=256] 2.9 ms 2.2 ms +32.23%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=16384/indices=1000] 12.8 ms 9.9 ms +29.32%
⚡ Simulation take[core/sorted/primitive/nonnull/chunks=1024/indices=100000] 31.3 ms 24.2 ms +29.25%
⚡ Simulation take_fsl_chunked_elements_random[16, 64] 524.7 µs 414 µs +26.73%
⚡ Simulation take_fsl_chunked_elements_sorted[16, 64] 504.8 µs 401.7 µs +25.68%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=16384/indices=256] 4 ms 3.3 ms +22.85%
... ... ... ... ... ...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

Comparing rk/chunkedfsltake (ffef2a3) with develop (c1ae775)

Open in CodSpeed

Footnotes

  1. 51 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

View more details on CodSpeed HQ