[SCIP] Cuts from cons_linear and cons_setpcc

[Matthias Walter]

Hi Ambros,

Thanks for the pointer - indeed, for cons_setppc, this was the origin. 
For cons_linear, most of these cuts were actually constructed in the 
call to SCIPseparateRelaxedKnapsack which is indeed called for a general 
linear constraint.

Best,
Matthias

On 30.10.20 17:12, Ambros Gleixner wrote:
> Hi Matthias,
>
> This could be due to conflict analysis.  Both constraint handlers have 
> a routine to try and create conflict constraints and by default these 
> are separated.  This is governed by parameter
>
>   conflict/separate
>
> so you can check whether setting this parameter to FALSE makes these 
> constraints disappear.  Or is it something else?
>
> Cheers,
> Ambros
>
>
>
> Am 29.10.20 um 09:43 schrieb Matthias Walter:
>> Hi all,
>>
>> I observed that  SCIP separated some cutting planes with origin 
>> "cons_linear" and "cons_setppc". More precisely,
>>
>> SCIPsepastoreApplyCuts
>>
>> was called and the SCIP_ROW*s in sepastore->cuts contain such 
>> inequalities. However, inspection revealed that these inequalities 
>> are not the ones given in the model. My question is, where do they 
>> come from? Is there some functionality in these constraint handlers 
>> that creates cuts (say, by tightening one) or were they added during 
>> presolve? I didn't find anything on separation of new cuts in the 
>> docs of these two handlers.
>>
>> Best,
>> Matthias
>> _______________________________________________
>> Scip mailing list
>> Scip at zib.de
>> https://listserv.zib.de/mailman/listinfo/scip
> _______________________________________________
> Scip mailing list
> Scip at zib.de
> https://listserv.zib.de/mailman/listinfo/scip