Skip to content

perf[array]: chunked take with bounded compute graph #6589

Merged
joseph-isaacs merged 9 commits intodevelopfrom
ji/chunked-take
Feb 19, 2026
Merged

perf[array]: chunked take with bounded compute graph #6589
joseph-isaacs merged 9 commits intodevelopfrom
ji/chunked-take

Conversation

@joseph-isaacs
Copy link
Contributor

@joseph-isaacs joseph-isaacs commented Feb 18, 2026

Currently take on a chunked array will create dict arrays for each contiguous indices over a chunk. Instead sort the indices to a filter on each chunk and then materalize and take the whole array

closes: #6407

Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@joseph-isaacs joseph-isaacs added the changelog/performance A performance improvement label Feb 18, 2026
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@joseph-isaacs joseph-isaacs marked this pull request as ready for review February 19, 2026 11:15
// Write the result to a Vortex file on disk.
let session = VortexSession::default();
let home = std::env::var("HOME")?;
let out_path = PathBuf::from(format!("{home}/Downloads/gharchive_shuffled.vortex"));
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

if you actually want this test, i can write a script that generates that arrow

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I removed for now. Might be a benchmark we should run, maybe make a issue?

Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@codspeed-hq
Copy link

codspeed-hq bot commented Feb 19, 2026

Merging this PR will not alter performance

✅ 977 untouched benchmarks
⏩ 1426 skipped benchmarks1


Comparing ji/chunked-take (9a4200f) with develop (6bc2c13)

Open in CodSpeed

Footnotes

  1. 1426 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.

Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@joseph-isaacs joseph-isaacs enabled auto-merge (squash) February 19, 2026 12:00
@joseph-isaacs joseph-isaacs merged commit 8fd8b8d into develop Feb 19, 2026
48 checks passed
@joseph-isaacs joseph-isaacs deleted the ji/chunked-take branch February 19, 2026 12:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

changelog/performance A performance improvement

Projects

None yet

Development

Successfully merging this pull request may close these issues.

dict on chunk explodes memory

3 participants