CPLEX callbacks

[Amin Moghimy Fam]

Dear all,

I have been searching to implement CPLEX callbacks using pyomo. I have the CPLEX for academic use and I create the solver as below:

opt = SolverFactory('cplex')

Then, I enable the callbacks and try to do something like this:

def StopCriterion(solver, model):
        if solver.get_num_nodes() > -1:
            print(aaaaaaaaaaaaaaaaaaaaa)

opt._allow_callbacks = True

opt.set_callback('MIPInfoCallback', StopCriterion)

However, I fail to get any result. My final goal is to enable a function to be able to terminate cplex at any time and get the most up to date results.

Is there any way to implement it?

[Ahmad Heidari]

Hello Amin, 

I hope you are doing well.

Unfortunately, when the solver does not achieve a feasible solution, it does not produce meaningful results. To address this issue, you may consider adjusting certain solver parameters. For instance, you can specify the following criteria:

solver = SolverFactory('gurobi')
solver.options['NonConvex'] = 2
solver.options['MIPGap'] = 0.1
solver.options['MIPGapAbs'] = 0.1

# ... (other parameters)

However, it is important to note that unless the solver successfully converges, it will continue running and may not terminate automatically.

Please let me know if further explanation is needed.