I'm experimenting with my model (JuMP with Ipopt solver), and I want to know if my changes are helping or hurting the memory consumption. I have been using the @time to see the memory usage, but I do not know if that is the best method.
I also switched to the limited-memory hessian approximation and did not see any reduction in memory size with the @time macro, which surprised me. My problem is large, so memory is an issue