We are solving a relatively large timetabling problem (15,000 decision variables) and we are trying to switch from IBM's CPLEX to a free solver. I am testing both CBC and Professor Nemhauser's MINTO solvers on NEOS website and I am using AMPL to make the models and the data files. After solving my model I'd like to print out my results but "display x;" brings all zero while I know that there should be some ones in the solution (b/c the objective value is not zero and also I have checked the model on CPLEX and it is not infeasible) Also none of the AMPL post processing commands work; for example the following forloop command does not work for {i in PAIR, j in ROOM, t in TIME: x[i,j,t]>0} print i," ", j," ", t, " ",x[i,j,t]; My hunch is that the AMPL implementation in COIN is not a standard implementation but I might be wrong. Do you know what I am doing wrong? Thanks asked 14 Mar '10, 22:11 Mark ♦ 
You might want to try using GLPK which uses an AMPLlike interface known as GMPL. This may prove easier for you. http://www.gnu.org/software/glpk/ If you really want to use COIN than I recommend the MIP solver SYMPHONY. It can read AMPL files. From an FAQ... What modeling languages and file formats does SYMPHONY work with?
answered 15 Mar '10, 12:58 larrydag 1 ♦ Thanks I will give it a try. I was using COIN because I just wanted to quickly test a number of different solvers and that website provided many of them. Thanks again
(16 Mar '10, 02:10)
Mark ♦

If you haven't done so already, you might want to ask on the Cbc mailing list. answered 17 Mar '10, 21:01 Paul Rubin ♦♦ 