[Symphony] modelling language for symphony

[Mueller, Martin C]

Hi,

Hans Mittelmann's regularily run benchmark is well-known and includes SYMPHONY and CPLEX: http://plato.asu.edu/ftp/milpf.html (http://scip.zib.de has a graphical display).

However, as a famous diagnostic physician would put it, "benchmarks always lie". For your particular problem the situation might differ completely from the benchmark. Just look at the detailed results of the benchmark above: for any ordered pair of solvers there's at least one instance for which the first one is better than the other one.

Best regards,
MCM


Siemens AG
Corporate Technology
CT PP 7
Otto-Hahn-Ring 6
81739 München, Deutschland
Tel.: +49 (89) 636-42373 
Fax: +49 (89) 636-42284 
Mobil: +49 (170) 9950406
mailto:martin.c.mueller at siemens.com 

Siemens Aktiengesellschaft: Vorsitzender des Aufsichtsrats: Gerhard Cromme; Vorstand: Peter Löscher, Vorsitzender; Wolfgang Dehen, Heinrich Hiesinger, Joe Kaeser, Barbara Kux, Hermann Requardt, Siegfried Russwurm, Peter Y. Solmssen; Sitz der Gesellschaft: Berlin und München, Deutschland; Registergericht: Berlin Charlottenburg, HRB 12300, München, HRB 6684; WEEE-Reg.-Nr. DE 23691322
 
 

> -----Original Message-----
> From: symphony-bounces at list.coin-or.org 
> [mailto:symphony-bounces at list.coin-or.org] On Behalf Of 
> Rundlett, Stuart J (Stuart)
> Sent: Montag, 21. September 2009 21:52
> To: Ashutosh Mahajan; G Chandramouli
> Cc: symphony at list.coin-or.org
> Subject: Re: [Symphony] modelling language for symphony
> 
> 
>  Hello,
> 
> Can you quantify how much slower Symphony is than Cplex for 
> MIP problems? Have there been any benchmark tests done?
> 
> We had been using an old (6.5) version of Cplex (with AMPL) 
> and are now using Symphony 5.1.10 (with GMPL), but I have not 
> been able to do any direct comparisons yet because the 
> software runs on different processors. I have not had any 
> major problem taking models written in AMPL for Cplex  and 
> getting them to run with GMPL for Symphony, although a few 
> changes are needed.
> 
> Thanks.
> 
> Stu Rundlett
> 
> -----Original Message-----
> From: symphony-bounces at list.coin-or.org 
> [mailto:symphony-bounces at list.coin-or.org] On Behalf Of 
> Ashutosh Mahajan
> Sent: Monday, September 21, 2009 1:15 PM
> To: G Chandramouli
> Cc: symphony at list.coin-or.org
> Subject: Re: [Symphony] modelling language for symphony
> 
> On Mon, 21 Sep 2009, G Chandramouli wrote:
> 
> >    Hi,
> >        I am doing my thesis in MIP of location problem. I 
> am using LINGO but
> >    it is very slow. I have 400 binary variables and over 
> 200000 constraints.
> 
> are you giving LINGO all the constraints at once? have you 
> tried starting with
> few constraints and adding only those that are violated later? 
> 
> >    I want to use symphony but I don't have a proper 
> modelling language. AMPL
> >    is costly and can't be afforded. Can I model it on AMPL 
> limited (free)
> >    version using symphony solver.
> 
> i dont think the free version allows you to model a problem 
> with such a size.
> 
> > Can GMPL be used for modelling such a large problem? What 
> about FLOPC++ ? 
> 
> either should work, but i havent used them myself.
> 
> > Any limitations with these languages for
> >    large problems? Also suggest me an alternative , in case 
> above languages
> >    fail
> >    Pls. tell me whether symphony is faster than LINGO and 
> Cplex in case of
> >    parallel systems and individual system.
> 
> we havent compared symphony to LINGO, but we are definitely 
> slower than Cplex.
> also, if you have cplex, you will most probably have access to ampl or
> gams for modelling as well.
> 
> --
> regards
> Ashutosh Mahajan
> http://coral.ie.lehigh.edu/~asm4
> 
> 
> _______________________________________________
> Symphony mailing list
> Symphony at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/symphony
> 
> 
> _______________________________________________
> Symphony mailing list
> Symphony at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/symphony
> 
>