> Just run fbavari pp.freemps if you have perl glpk installed. > If one uncomment $lp->adv_basis, one will get different results.
> Another question: > Way I, > load model > for loop > change obj coef > simplex > end > Way II, > for loop > load model > change obj coef > simplex > end > This two ways will bring different results. Anyone know that? Because your instance has multiple optimal solutions, that is indicated by zero reduced cost of non-basic variables; so in one case the solver finds one optimal solution while in other case it finds another depending on which the starting basis is used. _______________________________________________ Help-glpk mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-glpk
