Xypron, Unfortunately, it's not straightforward. It also occurs inside my column generator (each column is generated by optimizing a MIP instance). I tried once to create a synthetic test by generating a series of MIP instances but couldn't reproduce the behaviour.
I'll give it another try when I have time. Thanks, Pavel On Fri, Sep 3, 2010 at 6:04 PM, Xypron <[email protected]> wrote: > Hello Pavel, > > could you,please, supply an example to reproduce the behavior. > > Best regards > > Xypron > > > Pavel Klinov wrote: >> >> Hello Xypron and others, >> >> I have also observed that behaviour (asked here [1]). The performance >> can degrade quite seriously unless I reload the problem instance. I'd >> appreciate if someone could suggest another way of performing the same >> steps (rounding or whatever) but w/o writing/reading data on disk. >> >> Thanks, >> Pavel >> >> On Fri, Sep 3, 2010 at 5:53 PM, Xypron<[email protected]> wrote: >> >>>>> >>>>> The weirdness is not that it just gets stuck in an infinite loop. The >>>>> weird >>>>> part is that it works fine if instead of calling >>>>> lpx_simplex/lpx_integer >>>>> on >>>>> the problem that is already in memory, I write the problem to a file >>>>> and >>>>> load it again. >>>>> >>> >>> >> >> > > -- cheers, --pavel http://www.cs.man.ac.uk/~klinovp _______________________________________________ Help-glpk mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-glpk
