aj/OpenNest-Engines
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

refactor(gpt6astra): use host geometry, tolerances and layout checks

Browse Source 
Gpt6Astra reverse-engineered the validator: hand-tuned paddings and a
copied check sequence (ValidationOverlap) to match its rounding. It now
reads parts with JobPartGeometry, takes clearance from NestTolerances,
checks candidates with NestLayoutCheck.Clears, and assembles results with
NestJobResultBuilder and NestJobCost; its tests use the shared kit. Its
contact search, beam search and extra Automatic angles are unchanged.

Synthetic benchmark (5 jobs, salvage 0.5): all valid, 2 sheets each,
cost 5574.07 -> 5470.07; time 1871 -> 2400 ms from the stricter shared
check on arc-heavy jobs.

Co-Authored-By: Codex <noreply@openai.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
aj
2026-09-25 09:29:27 -04:00
co-authored by Codex Claude Opus 5.5
parent c691381f30
commit d0c6af783b
10 changed files with 120 additions and 217 deletions
Download Patch File Download Diff File Expand all files Collapse all files
OpenNest.Engine.Gpt6Astra/benchmarks/README.md
+1 -1
View File
@@ -9,7 +9,7 @@ claim of performance on an unseen competition dataset.
## Synthetic cases
Both versions were run on exactly the same programmatically generated geometry and stock.
The driver validates materialized output with `OpenNest.Benchmark.NestValidator`, including
The driver validates materialized output with `NestLayoutCheck.Violations`, including
quantity, stock settings, rotation, material overlap and spacing. Timings cover `Solve` only,
exclude external validation, and are single-run observations rather than stable distributions.
Every contact result is valid and complete. The baseline is valid but incomplete on `plate-cap`.