Choosen solution out of solution pool
Answered
Hello, I have the following question. I am currently solving a Cullum generation problem and have observed the following. When I solve the subproblems, I sometimes find several optimal solutions (i.e., the same reduced costs) that differ structurally. For example, let's say I minimize variable C in the master problem. Now I find two solutions in the subproblem with the same objective function value but different values for variable C. The first question is, how does Gurubi decide which of the two solutions is better, or which solution is output as the optimal solution?
-
-
Gurobi Staff
To your question “how does Gurobi decide which solution is better?”:
Gurobi doesn’t have a built-in notion that one of these optimal solutions is “better” than the other. Internally, the optimizer will end up with one of the optimal solutions depending on:
- The path the algorithm takes (e.g., pivot choices in simplex, branching decisions in MIP),
- Numerical aspects (tie-breaking on nearly equal values, internal ordering),
If you are interested in alternative solutions for MIPs you can reference our documentation on solution pools here: Solution Pool
2 -
Please sign in to leave a comment.
Comments
1 comment