[Cbc] clique processing/discovery in cbc

[John Forrest]

Fei,

There are switches to enable preprocessing combine cliques and there is 
some code to find cliques.  It is possible that the mechanism for using 
the cliques is broken.  Also CglClique should find them.  If you can 
send me your example (and point out one clique unless it is glaringly 
obvious) I can find problem (or hidden/forgotten switches).

John Forrest

On 12/02/14 01:51, Fei Peng wrote:
> Hi,
>
> Does cbc have the ability to construct the clique table in the 
> preprocessing step? I have a small problem where a number of cliques 
> can easily be found just by inspection, however when solved with Cbc 
> no clique cuts were added throughout the solution process. I do have 
> other types of cuts but I do think having the clique cuts may make a 
> difference (and another solver such as cplex generates them for the 
> same problem). Just wondering if there is a switch for constructing 
> the clique table in the pre-peocessing or another step -- I tried the 
> findCliques() routine but it caused segfaults..
>
> Thanks!
> Fei
> _______________________________________________
> Cbc mailing list
> Cbc at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/cbc
>
>