[Bonmin] Could you help me look at the problem,Thanks!

jx2139 at columbia.edu jx2139 at columbia.edu
Mon Nov 3 18:23:49 EST 2008


Hello,Bonmin Experts ,
I am using Bonmin 0.100 to solve a mixed integer convex problem. I  
found the following log message,then the program exits.
I don't know if it is a bug. Since for other NINLP problems, it works well.
Could you please help me look at the following and give some suggestions.
I modified CppExample in test to implement my problem and I have set  
the options as large as possible as follows to make Bonmin produce  
optimal solution. I used B-Hyb algorithm to solve the problem since it  
is faster for my problem compared to other options. BTW could Bonmin  
get the same solution with B-Hyb and Ecp for mixed integer convex  
problem? I found slight difference between them.

bonmin.options()->SetNumericValue("bonmin.time_limit", 1000000000);
bonmin.options()->SetStringValue("mu_oracle","loqo");
bonmin.options()->SetNumericValue("bonmin.ecp_max_rounds", 100);
bonmin.options()->SetNumericValue("bonmin.oa_dec_time_limit", 10000);
bonmin.options()->SetNumericValue("bonmin.nlp_solve_max_depth", 10000);
bonmin.options()->SetNumericValue("bonmin.node_limit", 3000000000);
bonmin.options()->SetNumericValue("bonmin.iteration_limit",3000000000);
bonmin.options()->SetNumericValue("bonmin.ecp_max_rounds_strong",10000 );
bonmin.options()->SetNumericValue("bonmin.bound_relax_factor",0);
bonmin.options()->SetNumericValue("bonmin.nlp_solve_freguency",100);
bonmin.options()->SetNumericValue("bonmin.solution_limit",3000000000);    .

log message:

NLP0012I
             Num      Status      Obj             It       time
NLP0013I     1        OPT         176.0012190844277       24       0.078
OA0003I New best feasible of 176.162 found after 0.094 sec.
OA0003I New best feasible of 176.012 found after 0.125 sec.
Cbc0012I Integer solution of 176.012 found by nonlinear program after  
0 iterations and 0
nodes (0.13 seconds)
Cbc0031I 2 added rows had average density of 6
Cbc0013I At root node, 0 cuts changed objective from 176.001 to  
176.001 in 1 passes
Cbc0014I Cut generator 0 (NLP solution based oa cuts) - 0 row cuts, 0  
column cuts (0
active)
Cbc0014I Cut generator 1 (Mixed Integer Gomory) - 0 row cuts, 0 column  
cuts (0 active)
Cbc0014I Cut generator 2 (Probing) - 0 row cuts, 0 column cuts (0 active)
Cbc0014I Cut generator 3 (Mixed Integer Rounding) - 0 row cuts, 0  
column cuts (0 active)
Cbc0014I Cut generator 4 (Cover) - 0 row cuts, 0 column cuts (0 active)
Cbc0014I Cut generator 5 (Clique) - 0 row cuts, 0 column cuts (0 active)
Cbc0014I Cut generator 6 (Flow Covers) - 0 row cuts, 0 column cuts (0 active)
Cbc0014I Cut generator 7 (Outer Approximation decomposition.) - 0 row  
cuts, 0 column
cuts (0 active)
Cbc0014I Cut generator 8 (Outer Approximation feasibility check.) - 7  
row cuts, 0 column
cuts (0 active)
OA0003I New best feasible of 176.007 found after 0.203 sec.
Cbc0016I Integer solution of 176.007 found by strong branching after 0  
iterations and 0
nodes (0.19 seconds)

Then the program jsut exit without "Search completed....." I am  
confused and don't know if it is a bug.
Look forward to your reply and thanks a lot.

Best regards,
Jun




More information about the Bonmin mailing list