Return current best solution CPLEX Python API

Question

Due to MIP problems which take long computation time, how do I instruct cplex to return current best solution when the computation time takes longer than, an hour for example, and the relative gap is at 5% for example? Individually, I believe I can use both functions: model.parameters.timelimit.set() and model.parameters.mip.tolerances.mipgap.set(), but how do I combine both?

Answer 1

You will have to use a callback to enforce both conditions. The mipex4.py example that is shipped with CPLEX shows exactly how to do this.

Here is the callback from the example:

class TimeLimitCallback(MIPInfoCallback):

    def __call__(self):
        if not self.aborted and self.has_incumbent():
            gap = 100.0 * self.get_MIP_relative_gap()
            timeused = self.get_time() - self.starttime
            if timeused > self.timelimit and gap < self.acceptablegap:
                print("Good enough solution at", timeused, "sec., gap =",
                      gap, "%, quitting.")
                self.aborted = True
                self.abort()

And the relevant part of the rest:

c = cplex.Cplex(filename)

timelim_cb = c.register_callback(TimeLimitCallback)
timelim_cb.starttime = c.get_time()
timelim_cb.timelimit = 1
timelim_cb.acceptablegap = 10
timelim_cb.aborted = False

c.solve()

sol = c.solution

print()
# solution.get_status() returns an integer code
print("Solution status = ", sol.get_status(), ":", end=' ')
# the following line prints the corresponding string
print(sol.status[sol.get_status()])

if sol.is_primal_feasible():
    print("Solution value  = ", sol.get_objective_value())
else:
    print("No solution available.")