4
$\begingroup$

I am modeling a big linear programming problem size with about 50,000 variables using OpenSolver.

Open Solver starts setting up the problem at a constant pace of about 100 variables per second, which is quite good.

However, after setting up about 24,000 variables, Open Solver suddenly changes this pace to about one variable per second.

Any help or hint on this?

$\endgroup$
3
  • 1
    $\begingroup$ Is the density of the constraint matrix fairly uniform, or are the first 24K columns sparse and then things get dense? Also, might it be bumping into memory constraints? (Can you tell how much memory it is using and how much it is allocated?) $\endgroup$
    – prubin
    Aug 24 at 19:59
  • 1
    $\begingroup$ I saw this related question a while back: stackoverflow.com/questions/34096450/… $\endgroup$ Aug 25 at 6:44
  • $\begingroup$ @decision maker, do you have any force to use excel? As excel does not perform suitable for the large set of data at all, for optimization or even non-optimization problem, There are some alternatives open-source platforms, like Pyomo, which can be easily connected to excel for solving large-scale models. $\endgroup$
    – A.Omidi
    Aug 25 at 11:30
1
$\begingroup$

I managed to understand why the pace of building up the model using Open Solver slows down after a while.

The reason is because there were several (in)equations without values, which was why model building was initially fast.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.