[Coin-discuss] Time limit on presolve?

[Stefan Vigerske]

Hi,

El Dono wrote:
> I can't send you the mps file because I'm just a student and that file 
> is part of a project of my school, so I should ask my teacher and other 
> professors working on it and so on...
> I can tell you some more details on mine tries: I've tried both calling 
> the branchAndBound and the initialSolve method and they do the same, 
> "primal infeasible" message, so I think the problem is not on the branch 
> and bound part. I've tried even the OsiClpSolverInterface but I get the 
> same message even with that solver (both with b&b and initialSolve) so I 
> think the problem is not in the solver choosen. So the only thing left 
> is the Osi Interface itself!
> I hope you can tell me something more with these new facts.

Not really.
 From what you tell it looks for me like a problem with the mps reader. 
Or maybe there are some very large numbers in your mps file which are 
confusing the interface.
Glpk seem to have a routine to also write a mps file. Maybe you use glpk 
as mps reader and writer and see if the mps formulation of your model 
created by glpk works better with the osi solvers.
Or you can check which difference are between your original file and the 
created one.

That's all what I can think of currently.
Best,
Stefan