Originally posted by: XPAN_Zaharia_dragos

Hello everybody,

I am an occasional user of CPLEX. I try to estimate if CPLEX may be helpfull for me in trying to solve an optimisation problem of interest for me.
After carefully learning the OPL language, I arrive to have a model with some 20,000 binary decision variables. These binary decision variables are organised as a matrix of about 140 x 150.
So, two questions now:
1. This number of binary decision variables is something which is a feasible, or it is something which is out of question to be solved in a reasonable time ?
2. Additionally, the matrix is largelly sparse, i.e. only 750 values of the total of 21,000 values are expected to be one, and the rest are zero. Or, after reading the help it seems that there is not evident the way to declare this matrix as sparse. Apparently, the barrier algorithm may be usefull to deal with this kind of matrices, but it is not evident how to set the CPLEX parameters to use only this algorithm.
So globally, do you think that my problem is a soluble one ? And if the answer is yes, may somebody indicate me how to toproceed for declaring the sparse matrix ?

Thanks in advance for any suggestion or comment.
#DecisionOptimization
#OPLusingCPLEXOptimizer