[Cbc] lp file format

[Haroldo Gambini Santos]

Em seg, 2017-11-13 às 13:29 +0000, Bjørn Sigurd Johansen (Spider
Solutions AS) escreveu:
> Question 2:
>  
> Ok, so I understand you that in case one has an integer variable that
> only is to hold positive values (including zero), then the
> correct/best definition is (only)
> GENERAL 
> I1
>  
> And, also, specifying the following is *not* necessary:
> BOUNDS
> 0 <= I1
For safety reasons I would also include this bound. I think that some
solvers declare general variables as free varibles by default (glpk)
and others not.
>  
> But one would often include a statement like
> BOUNDS
> I1 <= 10
>  
>  
>  
> From: Haroldo Santos [mailto:haroldo.santos at gmail.com] 
> Sent: mandag 13. november 2017 14:20
> To: Bjørn Sigurd Johansen (Spider Solutions AS) <bjorn.sigurd.johanse
> n at spidersolutions.no>
> Cc: cbc at list.coin-or.org; cbc at coin-or.org
> Subject: Re: [Cbc] lp file format
>  
>  
>  
> On Mon, Nov 13, 2017 at 10:41 AM, Bjørn Sigurd Johansen (Spider
> Solutions AS) <bjorn.sigurd.johansen at spidersolutions.no> wrote:
> > Thx  : )
> >  
> > Three modelling questions:
> >  
> >  
> > Question 1:  Is there any advantage in defining bounds in BOUNDS
> > section rather than in SUBJECT TO section?  I.e. is one of the
> > following preferred for an (integer and/or continuous) variable?
> >  
> > SUBJECT TO
> > I1>=0
> > I1<=10
> > ------------------------- alternatively -------------------------
> > BOUNDS
> > 0 <= I1 <= 10
> >  
> >  
> 
> I think that to define in bounds is better.
>  
> >  
> > Question 2:  What is/is there any purpose of defining no bounds,
> > i.e. using FREE in BOUNDS section, for integer variables?
> >  
> > BOUNDS
> > I1 FREE
> 
> Free variables can assume positive and negative values.
>  
> > GENERAL
> > I1
> > ------------------------- alternatively only ----------------------
> > ---
> > GENERAL
> > I1
> >  
> >  
> >  
> > Question 3:  Is there any difference between the following two?
> > V1>=0
> > V1>=+0
> > (I have seen “+0” being used several places).
> 
>  
> When saving LPs you'll probably see -0 +0  these are all zero. When
> saving the solution you'll probably see very small values (such 1e-
> 12)  which are zero too but appear as small values due to the limited
> precision arithmetic.
> >  
> >  
> > Thanks in advance,
> > Bjorn
> >  
> >  
> >  
> > From: Haroldo Santos [mailto:haroldo.santos at gmail.com] 
> > Sent: mandag 13. november 2017 13:03
> > To: Bjørn Sigurd Johansen (Spider Solutions AS) <bjorn.sigurd.johan
> > sen at spidersolutions.no>
> > Cc: cbc at list.coin-or.org; cbc at coin-or.org
> > Subject: Re: [Cbc] lp file format
> >  
> > Hi,
> >  
> > I think that there is no cbc specific (more precisely Osi) LP
> > format documentation. The LP file format was introduced in CPLEX
> > and I think that CPLEX has the official documentation.
> >  
> >  
> >  
> > On Mon, Nov 13, 2017 at 5:54 AM, Bjørn Sigurd Johansen (Spider
> > Solutions AS) <bjorn.sigurd.johansen at spidersolutions.no> wrote:
> > > Hi!
> > >  
> > > Is there any description on the CBC lp file format available ?
> > >  
> > > Or may the Cplex format description at http://lpsolve.sourceforge
> > > .net/5.0/CPLEX-format.htm be used ?
> > >  
> > > Best,
> > > Bjorn
> > > 
> > > _______________________________________________
> > > Cbc mailing list
> > > Cbc at list.coin-or.org
> > > https://list.coin-or.org/mailman/listinfo/cbc
> > > 
> > 
> > 
> >  
> > --
> > =============================================================
> > Haroldo Gambini Santos
> > Computing Department
> > Universidade Federal de Ouro Preto - UFOP
> > email: haroldo [at ] iceb.ufop.br
> > home/research page: www.decom.ufop.br/haroldo
> > 
> > 
> > It has long been an axiom of mine that the little things are
> > infinitely
> > the most important.
> > -- Sir Arthur Conan Doyle, "A Case of Identity"
> 
> 
>  
> --
> =============================================================
> Haroldo Gambini Santos
> Computing Department
> Universidade Federal de Ouro Preto - UFOP
> email: haroldo [at ] iceb.ufop.br
> home/research page: www.decom.ufop.br/haroldo
> 
> 
> It has long been an axiom of mine that the little things are
> infinitely
> the most important.
> -- Sir Arthur Conan Doyle, "A Case of Identity"
-- 
=============================================================
Haroldo Gambini Santos
Computing Department
Universidade Federal de Ouro Preto - UFOP
email: haroldo [at ] iceb.ufop.br
home/research page: www.decom.ufop.br/haroldo


It has long been an axiom of mine that the little things are 
infinitely the most important.
-- Sir Arthur Conan Doyle, "A Case of Identity"