Closed chriscoey closed 7 years ago
I set a limit of 300s on pp_n1000_d10000
12833 7635 216110.7073 969 216113.6626 216110.6643 78686 0.00% * 13000+ 7790 216113.6446 216110.6643 0.00% 13005 7797 216110.6772 990 216113.6446 216110.6644 78861 0.00% MIP-solver-driven algorithm summary: - Status = CutsFailure - Best feasible = +2.161290e+05 - Best bound = +2.161107e+05 - Relative opt. gap = 8.483e-05 - Total time (s) = 3.00e+02 Timers (s): - Setup = 9.65e-02 -- Transform data = 1.40e-02 -- Create conic data = 5.33e-02 -- Create MIP data = 2.92e-02 - Algorithm = 3.00e+02 -- Solve relaxation = 2.83e-01 -- Get relaxation cuts = 0.00e+00 -- MIP solver driving = 3.00e+02 -- Solve subproblems = 3.32e+00 -- Get subproblem cuts = 1.87e+00 -- Get primal cuts = 2.38e-03 Counters: - Lazy callbacks = 234 -- Integer repeats = 121 -- Conic subproblems = 113 --- Infeasible = 0 --- Optimal = 113 --- Suboptimal = 0 --- UserLimit = 0 --- ConicFailure = 0 --- Other status = 0 -- Feasible solutions = 119 --- From subproblems = 113 --- In lazy callback = 6 - Heuristic callbacks = 13034 -- Solutions passed = 46 Solution returned by conic solver Outer-approximation cuts added: Cone | Relax. | Violated | Nonviol. Second order | 0 | 167351 | 0 0 numerically unstable cone duals encountered Distance to feasibility (negative indicates strict feasibility): Cone | Variable | Constraint Linear | -1.20e+01 | 0.00e+00 Rotated S.O. | -2.39e-07 | NA Distance to integrality of integer/binary variables: integer | 0.00e+00 #STATUS# CutsFailure #OBJVAL# 216128.99852783338 #OBJBOUND# 216110.66440379724 #TIMESOLVER# 300.0157709121704 #TIMEALL# 300.1188051700592
it happened because the relaxation solve gave CutsFailure. we did not get that in MSD. and the return status is thus misleading.
I set a limit of 300s on pp_n1000_d10000