3 ms·
> 4000 integers and a couple hundreds of constraints we can can typically solve today to proven global optimality within seconds Oh really? I got a problem wit
by freemint 4y ago
> 4000 integers and a couple hundreds of constraints we can can typically solve today to proven global optimality within seconds
Oh really? I got a problem with 1064 binary variables and 8000ish constraints. Surely you can solve it for me. I do just require a solution not even an optimal solution.
Surely you can solve it for me. https://www.file-upload.net/download-14976197/strassic-7-2.mps.html https://www.file-upload.net/download-14976197/strassic-7-2.m...
It is not that simple, it really depends on the problem and the formulation.
- whatever1 4y agoI said typically. These are NP problems you can always find edge cases that are impossible to solve. For these edge cases we can always reformulate the problem, aka use different definition of variables and constraints that can better inform the search. We are talking about real world applications, not toying with pathogenic benchmark toy examples.
- whatever1 4y agoBtw out of curiosity I submited your mps file to the NEOS server https://neos-server.org/neos/ https://neos-server.org/neos/ and solved it with FICO XPRESS to global optimality.
- freemint 4y agoI had Gurobi struggled with it for over a week. Do you have job id & password or the solution file?