Fix 2446

[fwesselm]

Fix #2446

• HPresolve::rowPresolve implements "simple probing on a single equation" (see section 3.6, Achterberg et al., Presolve Reductions in Mixed Integer Programming, INFORMS Journal on Computing 32(2):473-506).
• The current code uses bounds on constraint activities sumLower and sumUpper that were obtained using implied column bounds (vectors implColLower and implColUpper), but then uses the standard bounds (vectors col_lower_ and col_upper_) to perform substitutions. I think this is inconsistent.
• I think there are two potential ways to resolve this: (a) use vectors sumLowerOrig and sumUpperOrig that were computed using only the standard bound vectors, or (b) keep using sumLower and sumUpper, but correct the bounds used in the substitutions. I implemented the latter in this PR.
• A test has been added.
• I think this should be qualified on a large set of MIPs and, unfortunately, I am not able to do this currently.
• Once I have the test results, I will report back.

[codecov]

Codecov Report

❌ Patch coverage is 98.64865% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 79.46%. Comparing base (ebd83e1) to head (39c52a2).
⚠️ Report is 56 commits behind head on latest.

Files with missing lines	Patch %	Lines
highs/presolve/HPresolve.cpp	98.48%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           latest    #2457      +/-   ##
==========================================
- Coverage   79.55%   79.46%   -0.10%     
==========================================
  Files         346      346              
  Lines       85401    85401              
==========================================
- Hits        67942    67862      -80     
- Misses      17459    17539      +80     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Opt-Mucca]

Changes look good!

[fwesselm]

The results look OK. The node counts on three problems are affected by this bug fix (blp-ic97, neos-1593097, p80x400b).