Skip to content

Normalize Execution#7278

Merged
gatesn merged 12 commits intodevelopfrom
ngates/normalize-execution
Apr 7, 2026
Merged

Normalize Execution#7278
gatesn merged 12 commits intodevelopfrom
ngates/normalize-execution

Conversation

@gatesn
Copy link
Copy Markdown
Contributor

@gatesn gatesn commented Apr 3, 2026

Allow execution during normalization

@gatesn
Normalize Execution
4545e5c 
Signed-off-by: Nicholas Gates <nick@nickgates.com>
@gatesn gatesn requested a review from joseph-isaacs April 3, 2026 14:53
@gatesn gatesn enabled auto-merge (squash) April 3, 2026 14:53
@gatesn gatesn added the changelog/feature A new feature label Apr 3, 2026
gatesn and others added 3 commits April 3, 2026 11:03
@gatesn
Normalize Execution
58fd4cd 
Signed-off-by: Nicholas Gates <nick@nickgates.com>
@joseph-isaacs
Merge develop into ngates/normalize-execution
7dd87e0 
Resolve conflicts keeping the HashSet<ArrayId>-based allowed encodings
approach from this branch. Remove duplicate with_slots method from
auto-merge and clean up unused imports.

Signed-off-by: Joe Isaacs <joe@spiraldb.com>
@joseph-isaacs
fix
b1a491d 
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@codspeed-hq
Copy link
Copy Markdown

codspeed-hq bot commented Apr 7, 2026 •
edited
Loading

Merging this PR will degrade performance by 23.82%

⚡ 2 improved benchmarks
❌ 24 regressed benchmarks
✅ 1096 untouched benchmarks
⏩ 1530 skipped benchmarks1

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

Mode Benchmark BASE HEAD Efficiency
⚡ Simulation chunked_bool_canonical_into[(1000, 10)] 60.3 µs 49.9 µs +20.85%
❌ Simulation take_search[(0.005, 0.1)] 246.7 µs 320 µs -22.9%
❌ Simulation take_search[(0.005, 0.05)] 131.2 µs 168.1 µs -21.95%
❌ Simulation take_search[(0.005, 0.5)] 1.2 ms 1.5 ms -23.7%
❌ Simulation take_search[(0.005, 1.0)] 2.3 ms 3.1 ms -23.82%
❌ Simulation take_search[(0.01, 1.0)] 2.5 ms 3.3 ms -22.35%
❌ Simulation take_search[(0.1, 1.0)] 3.5 ms 4.3 ms -17.11%
❌ Simulation take_search[(0.1, 0.05)] 212.1 µs 249 µs -14.82%
❌ Simulation take_search[(0.1, 0.1)] 386.1 µs 459.5 µs -15.96%
❌ Simulation take_search[(0.01, 0.5)] 1.3 ms 1.6 ms -22.24%
❌ Simulation take_search[(0.01, 0.1)] 267.5 µs 340.8 µs -21.51%
❌ Simulation take_search_chunked[(0.005, 0.05)] 165.5 µs 197 µs -15.99%
❌ Simulation take_search[(0.1, 0.5)] 1.8 ms 2.2 ms -16.96%
❌ Simulation take_search[(0.01, 0.05)] 142.3 µs 179.2 µs -20.59%
❌ Simulation take_search_chunked[(0.005, 0.1)] 314.2 µs 376.6 µs -16.58%
❌ Simulation take_search_chunked[(0.01, 1.0)] 3.2 ms 3.9 ms -16.07%
❌ Simulation take_search_chunked[(0.01, 0.05)] 178.4 µs 209.9 µs -15%
❌ Simulation take_search_chunked[(0.01, 0.5)] 1.6 ms 1.9 ms -16.01%
❌ Simulation take_search_chunked[(0.1, 0.1)] 449.6 µs 512 µs -12.2%
❌ Simulation take_search_chunked[(0.005, 0.5)] 1.5 ms 1.8 ms -17.05%
... ... ... ... ... ...

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

Comparing ngates/normalize-execution (042c636) with develop (29a5f43)

Open in CodSpeed

Footnotes

  1. 1530 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. ↩

joseph-isaacs and others added 8 commits April 7, 2026 16:23
@joseph-isaacs @claude
Fix vortex-cuda build: use HashSet<ArrayId> instead of ArrayRegistry
d2a8db2 
The NormalizeOptions API expects HashSet<Id>, not ArrayRegistry. Update
CudaFlatLayoutStrategy to match. Also clean up unused imports in
strategy.rs from the merge.

Signed-off-by: Joe Isaacs <joe@spiraldb.com>

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@joseph-isaacs @claude
Fix lint, update public API lock files
a3f6379 
Remove unused imports from merge (ArrayRegistry, ArrayPluginRef,
Primitive) in flat writer tests. Fix redundant clone in normalize test.
Regenerate public API lock files.

Signed-off-by: Joe Isaacs <joe@spiraldb.com>

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@joseph-isaacs @claude @gatesn
Fix lint, update public API lock files
5b7a692 
Remove unused imports from merge (ArrayRegistry, ArrayPluginRef,
Primitive) in flat writer tests. Fix redundant clone in normalize test.
Regenerate public API lock files.

Signed-off-by: Joe Isaacs <joe@spiraldb.com>

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@joseph-isaacs @claude
Fix flat_valid_array_writes test: allow Primitive encoding for Dict c…
bc11e55 
…hildren

Dict arrays contain Primitive codes and values, so the allowed encodings
set must include Primitive in addition to Dict.

Signed-off-by: Joe Isaacs <joe@spiraldb.com>

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@gatesn
DuckDB FSST Export
353189b 
Signed-off-by: Nicholas Gates <nick@nickgates.com>
@gatesn
Merge branch 'ngates/normalize-execution' of github.com:vortex-data/v…
3ed9779 
…ortex into ngates/normalize-execution
@gatesn
DuckDB FSST Export
0f78b72 
Signed-off-by: Nicholas Gates <nick@nickgates.com>
@gatesn
DuckDB FSST Export
042c636 
Signed-off-by: Nicholas Gates <nick@nickgates.com>
@gatesn gatesn merged commit 06065ff into develop Apr 7, 2026
105 of 109 checks passed
@gatesn gatesn deleted the ngates/normalize-execution branch April 7, 2026 16:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@joseph-isaacs joseph-isaacs joseph-isaacs approved these changes

Assignees

No one assigned

Labels

changelog/feature A new feature

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@gatesn @joseph-isaacs