Fix erroneous early solve termination - #2037
Conversation
|
/ok to test 263179d |
|
Navigate logical layers of code changes, visualize relationships, and explore their blast radius. No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Repository: NVIDIA/cuopt/.coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 11 remain after this review. 📝 WalkthroughWalkthroughThe branch-and-bound numerical lower bound now starts at positive infinity instead of an unspecified value. ChangesBranch-and-bound initialization
Priority: ⬇️ Low Estimated code review effort: 1 (Trivial) | ~3 minutes Change: Bug fix Suggested reviewers: Merge Risk: ⚪ Minimal · up to The initialization does not create a premature-solve path; bound consumers handle the infinity sentinel. No actionable merge-blocking issue remains. 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
CI Test Summary✅ All 32 test job(s) passed. |
|
/merge |
#2003 introduced an early termination checking path when solutions arrive to the B&B by checking against the dual bound.
However, for early heuristics that run before B&B begins its solve, the dual bound may in some cases be uninitialized, which can cause early termination even though the gap is still large.
This PR introduces a fix by initializing the numerical lower bound to inf before B&B solve starts.
Description
Issue
Checklist