Skip to content

Drop direct arrow-select dependency - #10826

Merged
robert3005 merged 3 commits into
developfrom
rk/remove-arrow-cast-select
Oct 26, 2026
Merged

robert3005 merged 3 commits into
developfrom
rk/remove-arrow-cast-select

Conversation

@robert3005

Copy link
Copy Markdown
Contributor

This is still transitively picked up through arrow_cast and parquet variant but the logic is simpler without it

Replace arrow_select::concat with a MutableArrayData-based helper, and
replace the arrow_cast fallbacks for run-end and dictionary export with
Vortex-native run detection and dict_encode.

Signed-off-by: Robert Kruszewski <github@robertk.io>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FDwK8nfgMsXfEfYoKyoAoB
Rework the replacements for the arrow-cast and arrow-select calls:

- Run-end export of a non-run-end array uses RunEnd::encode for
  primitives, re-wraps encoded storage for extension types, and for every
  other dtype finds run boundaries with one vectorized Eq comparison of
  adjacent rows combined with the validity bitmap.
- Dictionary export of a non-dictionary array uses dict_encode for
  primitives and strings/binary, a hash encoder for decimals of every
  width, and re-wraps encoded storage for extension types, so it covers
  the value types arrow-cast supported. Nulls live on the keys.
- Chunked lists canonicalize to a zero-copy ListView over chunked
  elements instead of being exported per chunk and concatenated.
- Map scalar entries are appended with arrow-data's MutableArrayData,
  dropping the separate concat module.

Signed-off-by: Robert Kruszewski <github@robertk.io>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FDwK8nfgMsXfEfYoKyoAoB
Keep arrow-cast in vortex-arrow and restore its run-end and dictionary
export fallbacks. Only the arrow-select uses are replaced: map scalar
entries are appended with arrow-data's MutableArrayData, and chunked
lists canonicalize to a zero-copy ListView instead of being exported per
chunk and concatenated.

Signed-off-by: Robert Kruszewski <github@robertk.io>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FDwK8nfgMsXfEfYoKyoAoB
@robert3005 robert3005 added the changelog/chore A trivial change label Oct 26, 2026
@codspeed

codspeed Bot commented Oct 26, 2026

Copy link
Copy Markdown

Merging this PR will degrade performance by 11.39%

4 warnings, 1 note

⚠️ Walltime results may be unreliable

Some benchmarks used the Walltime instrument outside CodSpeed Macro Runners, which can make the results inconsistent.

Learn more about Macro Runners

See affected benchmarks

⚠️ Unable to measure the execution of 2 benchmarks

These benchmarks measured no execution time, so their results are not comparable and count as unchanged. This usually happens when the compiler removes the code under test.

Prevent compiler optimizations

See affected benchmarks

⚠️ 20 benchmarks are dominated by system calls

System calls cannot be simulated consistently and are left out of the measure. The reported value understates the real cost. Switch to the Walltime instrument to measure them.

Learn more about system calls

See affected benchmarks

⚠️ Base and head ran on different runner types

Some benchmarks with significant performance changes were compared across different runner types, so the changes might not come from the code. Run base and head on the same runner type, or use CodSpeed Macro Runners for more stable results.

Learn more about Macro Runners

Open the report in CodSpeed to investigate

ℹ️ 40 benchmarks have no flamegraph

The flamegraph wasn't generated: the symbols of the profiling trace could not be resolved, because no information about the language frames was collected.

See affected benchmarks

❌ 4 regressed benchmarks
✅ 2435 untouched benchmarks
⏩ 12 skipped benchmarks1

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

Mode Benchmark BASE HEAD Efficiency
❌ CPU Simulation test_scan_filter_to_arrow_array[int_float] 3.5 ms 4 ms -13.93%
❌ CPU Simulation test_repeated_scan_filter[int_float] 3.4 ms 3.8 ms -11.06%
❌ CPU Simulation test_repeated_scan_filter[int] 2.4 ms 2.7 ms -10.3%
❌ CPU Simulation test_scan_filter_to_arrow_array[int] 2.5 ms 2.8 ms -10.23%

Tip

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


Comparing rk/remove-arrow-cast-select (5d7026c) with develop (5b45ae4)

Open in CodSpeed

Footnotes

  1. 12 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. ↩

@robert3005
robert3005 enabled auto-merge (squash) October 26, 2026 18:17
@robert3005
robert3005 merged commit 2c730c4 into develop Oct 26, 2026
110 of 112 checks passed
@robert3005
robert3005 deleted the rk/remove-arrow-cast-select branch October 26, 2026 19:02
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

changelog/chore A trivial change

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants