HiGHS Infeasible when Presolve is Off

[erdembanak]

For some problems, HiGHS returns the problem as infeasible when CPLEX and CBC find the optimal solution (3.1108620188e+11). Since there were issues about presolve, I have switched off presolve but the problem is still infeasible.

infeasible_presolve_off.mps.txt

[jajhall]

Thanks!

[jajhall]

Just run this with Gurobi and it says,

Matrix range [1e+00, 1e+09] Objective range [1e+00, 1e+00] Bounds range [1e+00, 1e+00] RHS range [1e-01, 2e+03] Warning: Model contains large matrix coefficients Consider reformulating model or setting NumericFocus parameter to avoid numerical issues.

and, although it finds a solution with objective 3.110862018838e+11,

Warning: max constraint violation (1.8311e-04) exceeds tolerance

So, since the problem may be infeasible to within the primal tolerance, it may be futile to try to identify whether HiGHS is actually wrong