[Couenne] false infeasible

[Pietro Belotti]

Hi,

size should not be a problem, but it might bring about numerical
problems. It would be useful to look at the instance that is
erroneously declared infeasible, or even better the smallest instance
that is declared infeasible.

Regards,
Pietro

On Tue, Nov 24, 2015 at 9:20 AM, ALI ZAZOU Abdelkrim
<abdelkrim.ali-zazou at ensma.fr> wrote:
> Hi,
>
> i'm wondering if couenne can handle large MIQCP problems (12k Variables),
> because i implemented a small problem of the exact same formulation as the
> large one and it solves well but when i try to solve the large one it gave
> me the information that :
> "=C Couenne: Warning, tightened NLP is infeasible
>  =CReformulation finds model infeasible."
> And in reality this problem can be solve with Baron, so does it's a problem
> of size?
>
> thank you very much for your informations.
>
> Abdelkrim
> _______________________________________________
> Couenne mailing list
> Couenne at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/couenne