[Cbc] Getting CSV result with variables only greater than epsilon (1e-4), no zeros result

[Haroldo Santos]

Can you send me the MIP?
Em 12/03/2016 20:46, "Noli Sicad" <nsicad at gmail.com> escreveu:

> Hi Haroldo,
>
> I did not work.
>
> I followed your instruction ( i.e. changing 1.0e-8 to 1.0e-4 ) and
> saved the file.
>
> And then:
>
> ~~~~
> make clean
>
> make
>
> sudo make install
>
> ~~~~
>
> However, it did not work. I am still getting the variables with zero
> values in my CSV output.
>
> I tried to solve and output (without zeros vallues) in my LP model but
> I did not work.
>
> I like to this as well in my MIP model (i.e. not getting the zero
> values of the variables).
>
> Any other option to this?
>
> Thanks.
>
> Regards,
>
> Noli
>
>
> On 3/13/16, Haroldo Gambini Santos <haroldo.santos at gmail.com> wrote:
> > Hi Noli,
> >
> > In CbcSolver.cpp,  line 9685,  just change
> >         } else if (fabs(primalColumnSolution[iColumn]) > 1.0e-8) {
> > to
> >         } else if (fabs(primalColumnSolution[iColumn]) > 1.0e-4) {
> >
> > Cheers,
> >
> > Haroldo
> >
> > Em 11-03-2016 21:56, Noli Sicad escreveu:
> >> Hi,
> >>
> >> I like to get results  of variables only greater than epsilon( 1e-4).
> >>
> >> # significantly different from zero
> >> epsilon  = 1e-4;
> >>
> >> How do I do this? Which source file should I modify and how?
> >>
> >> My MIP model submitted to CBC solver is quite big (e.g. greater than
> >> 50,000 columns) and most of the results are just zero (0). Opening in
> >> the spreadsheet is slow when cleaning the result.
> >>
> >> I think it would be good to have option in the CSV output with
> >> variables only greater than epsilon  (1e-4). No zero (0) included in
> >> the CSV output.
> >>
> >> Thanks.
> >>
> >> Regards,
> >>
> >> Noli
> >> _______________________________________________
> >> Cbc mailing list
> >> Cbc at list.coin-or.org
> >> http://list.coin-or.org/mailman/listinfo/cbc
> >>
> >
> > --
> > ==================================================
> > Haroldo Gambini Santos
> > D.Sc, Computer Science
> > Universidade Federal de Ouro Preto
> > http://www.decom.ufop.br/haroldo/
> >
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/cbc/attachments/20160312/7697b99c/attachment.html>