Page 1 of 1

Solving MIP with CPLEX. Gap 0%?

Posted: Fri Sep 07, 2018 1:08 pm
by GFA
Dear GAMS-users,

I'm trying to solve a MIP-model using CPLEX, however somehow the model becomes very slow (see attachment).
What I don't understand is why it keeps saying "GAP 0%" while not returning an optimal solution. Does anyone have a clue why the model could be slow?


I set the following options:
OPTION OPTCA=0;
OPTION OPTCR=0;

and in the CPLEX option-file:
numericalemphasis=1
scaind=1
relaxfixedinfeas=1
mipdisplay=2

Regards,
GFA

Re: Solving MIP with CPLEX. Gap 0%?

Posted: Tue Sep 11, 2018 6:55 am
by bussieck
It says "gap 0.00" so something smaller than 5e-3, but apparently not 0 as you requested otherwise it would have stopped. The objective to so humongous that you also can't see the difference between the incumbent and bound. Set the gap to something bigger or wait for Cplex to really prove optimality.

-Michael