> Hello, when I solve a problem, "PROBLEM HAS NO PRIMAL FEASIBLE SOLUTION" > appeared immediately after the "Model has been successfully generated". > But the problem should be calculated for a little time. Dose it mean that > my model is not correct?
This only means that your problem is primal infeasible, i.e. it has no primal feasible solution. If you are in doubt, you can disable the lp presolver with the option '--nopresol'. > Another question is what does the sentences mean when glpsol solve the > problem, such as this: > 800: obj = 1.051500000e+002 infeas = 3.231e+001 (25) > 1000: obj = 1.403333333e+002 infeas = 8.000e-001 (2) > * 1038: obj = 1.379333333e+002 infeas = 5.157e-015 (1) > * 1200: obj = 7.906666667e+001 infeas = 4.151e-015 (0) > * 1225: obj = 7.820000000e+001 infeas = 5.274e-016 (0) > OPTIMAL SOLUTION FOUND > Integer optimization begins... > + 1225: mip = not found yet >= -inf (1; 0) > + 30729: >>>>> 7.820000000e+001 >= 7.820000000e+001 < 0.1% (324; 0) > + 30729: mip = 7.820000000e+001 >= tree is empty 0.0% (0; 647) At first glpsol solves the lp relaxation with the simplex method, and then, once optimal basic solution to lp relaxation has been found (the line "OPTIMAL SOLUTION FOUND") it calls the mixed integer optimizer to find integer optimal solution (the line "Integer optimization begins..."). For more details see the glpk reference manual included in the distribution. See also: http://lists.gnu.org/archive/html/help-glpk/2007-06/msg00060.html _______________________________________________ Help-glpk mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-glpk
