Skip to first unread message
Cyan Zhou
Dec 4, 2019, 6:04:42 AM12/4/19
to YALMIP
Hi, I am solving a linear optimization problem using Mosek. The problem can be solved when the size of the problem is small, for example, less than 50 variables and less than 500 constraints. However, when the size of the problem becomes larger, with more than 4000 variables and 80000 constraints, mosek cannot get a feasible solution. Is that because that this size of problem is unsolvable using mosek?
Johan Löfberg
Dec 4, 2019, 6:07:32 AM12/4/19
to YALMIP
Depends on what problem it is. An LP of that size is fairly reasonable, a MILP of that type could be unsolvable within the estimated remaining time of the universe if you are unlucky
Johan Löfberg
Dec 4, 2019, 6:08:40 AM12/4/19
to YALMIP
and what do you mean with "cannot get a feasible solution". If you mean it says it is infeasible, that has nothing to do with size. That just means you've posed a problem which doesn't have a solution.
Cyan Zhou
Dec 4, 2019, 6:25:41 AM12/4/19
to YALMIP
Thank you. Then, why it can be solved with a smaller size? since I only added increase the size the problem and didn't change the problem at all.
Johan Löfberg
Dec 4, 2019, 6:27:50 AM12/4/19
to YALMIP
Well you obviously changed something, as you are solving a bigger problem. Either you now have an infeasible problem, or this larger problem is numerically ill-posed causing the solver to have numerical problemsand failing to find a solution. We cannot tell without more info and debug info
Michal Adamaszek
Dec 4, 2019, 6:40:08 AM12/4/19
to YALMIP
Can you please at the very least post the full log output so that we know what we are talking about?
Reply all
Reply to author
Forward
0 new messages