Yingjie, > I did a bench mark on my Mac Book, and there is more than 10% time > saving with glpk + row generation than the well respected special > method!
Thank you for benchmarking. I think that the 22-node example is too easy for both glpk and the b&b code you used. Solving larger instances having 40-50 nodes would be more interesting. (Playing with lolib I noticed that instances with 0-1 weights are much harder for solving with b&b.) FYI: The new version of LOLIB is available at: http://heur.uv.es/optsicom/LOLIB/ Andrew Makhorin _______________________________________________ Help-glpk mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-glpk
