Feasible with CONOPT3, Infeasible with CONOPT4
Posted: Thu Sep 12, 2019 10:17 am
Hello everyone,
lately I bought the latest version of GAMS along with CONOPT4 so as to be able to solve a large NLP problem for which CONOPT3 was taking way too long.
So I tested CONOPT4 on a small problem at first but it was unable to find a feasible solution while CONOPT3 did find one.
I can see CONOPT4 is faster than CONOPT3 for large problems but this issue is a bit frustrating...
I sent a scalar version of my problem to Gams support but no answer so far. Their first guess was poor scaling.
Do you have any idea about this issue ?
Thanks, in advance,
Sisco
lately I bought the latest version of GAMS along with CONOPT4 so as to be able to solve a large NLP problem for which CONOPT3 was taking way too long.
So I tested CONOPT4 on a small problem at first but it was unable to find a feasible solution while CONOPT3 did find one.
I can see CONOPT4 is faster than CONOPT3 for large problems but this issue is a bit frustrating...
I sent a scalar version of my problem to Gams support but no answer so far. Their first guess was poor scaling.
Do you have any idea about this issue ?
Thanks, in advance,
Sisco