[Scip] Strange behavior

[Stefan Heinz]

Hi Aurélien,


On 07/04/2012 10:30 AM, Aurélien Questel wrote:
> Hi everyone,
> thank you all for your hints so far.
> I've been able to finish my BCP and it gives very good results :)
That sounds good :-)
> Now I would like to add a primal heuristic,
> but I wanted to know if a primal heuristic can add some columns in the PL.
> Is it ok or must we use only the already priced columns ?
That is possible. It make no difference for SCIP when you pop up with 
new columns/variables. The only thing you have to ensure (and what you 
are already doing) that during the LP pricing loop you create all 
columns such that LP is solved to optimality.

Best Stefan
> Regards,
>
> Aurélien
>
> Le 20 juin 2012 à 13:24, Stefan Vigerske a écrit :
>
>> Hi,
>>
>> if it's not the check method, then it could be that your enforcement callback is accepting a solution as feasible that is not. Can you debug this?
>> Or you erroneously decide in your constraint handler at some point that your constraint is redundant and can be deleted.
>>
>> Best,
>> Stefan
>
> _______________________________________________
> Scip mailing list
> Scip at zib.de
> http://listserv.zib.de/mailman/listinfo/scip