[Cbc] Always, always Cutlength

[John Forrest]

Yves,

Will look into it.  Any change can (and does on your problems) change 
the accuracy - maybe on cuts derived from original constraints plus cuts.

But you have it the wrong way round.  cutlength XX says throw away all 
cuts with length > XX.  The smaller XX, the more cuts will be thrown 
away and so accuracy will be no worse.  As to efficiency - it could go 
either way.  GMI cuts seem to be more accurate and Lagrangean Gomory 
cuts should be more accurate.

So for production purposes I would set cutlength 20.

John

On 19/12/14 15:01, Yves Touchard wrote:
> John,
>
> I am sorry to disturb you (once more) about the cutlength value.
>
> From my understanding, increasing this value decreases efficiency and 
> increases accuracy.
> And decreasing this value, increases efficiency and decreases accuracy.
>
> So, if i am not wrong, please, could you explain me the following 
> behaviour:
> [ /bin/cbc t_18.lp threads 104 cutlength //XX//divesolves 301 ratio 
> 0//.05//multiple 4 branch printingOptions rows solution t_18..sol/ ]
> [ /t18_lp file is there: 
> //https://www.dropbox.com/s/748azg1miloa3iw/t_18.lp?dl=0/ ]
>
>   * With a XX cutlength value of 5, 10, 15, 20, 25, 29, 31, 35 and 50
>     the optimal solution is found.
>   * With a XX cutlength value of 30 and 45, the problem is infeasible
>     (but become feasible with a multiple value of 1).
>
> Regards,
>
> Yves
>
> ------------------------------------------------------------------------------
> This message and any attachments (the "message") are confidential and 
> intended solely for the addressee(s). Any unauthorized use or 
> dissemination is prohibited. E-mails are susceptible to alteration. 
> Neither DxO Labs nor any of its subsidiaries or affiliates shall be 
> liable for the message if altered, changed or falsified.
> Ce message et toutes les pièces jointes (ci-après le "message") sont 
> confidentiels et établis à l'intention exclusive de ses destinataires. 
> Toute utilisation ou diffusion non autorisé est interdite. Tout 
> message électronique est susceptible d'altération. DxO Labs et ses 
> filiales déclinent toute responsabilité au titre de ce message s'il a 
> été altéré, modifié ou falsifié.
>
>
> _______________________________________________
> Cbc mailing list
> Cbc at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/cbc

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/cbc/attachments/20141219/7240f124/attachment.html>