[Ipopt] Infeasible problem detected during optimization

Stefan Vigerske stefan at math.hu-berlin.de
Fri Apr 12 04:58:36 EDT 2019


Hi,

you might want to check what tolerances you are using.
Also as noted at https://coin-or.github.io/Ipopt/OUTPUT.html, 
"constraint violation is measured for the internal problem formulation...".
If the 9e-12 is what you see in the inf_pr column of the Ipopt output, 
maybe consider using option inf_pr_output 
(https://coin-or.github.io/Ipopt/OPTIONS.html#OPT_inf_pr_output) to 
switch to seeing constraint violation that Ipopt uses internally.

Stefan

On 4/11/19 2:36 PM, masoumeh parseh wrote:
> Hi Stefan,
> 
> I saw your answer on ipopt for infeasible problem detection. I am using ipopt with CasADi in Matlab and I get infeasible problem detection. I even used an optimal solution as an initial guess for the solver but still get infeasible problem. When I check where the problem is I see I ave very small constraint violation. For example, 9e-12 and values similar to this. I was wondering if you know of any ipopt option to fix this problem.
> 
> Best,
> 
> /Masoumeh
> _______________________________________________
> Ipopt mailing list
> Ipopt at list.coin-or.org
> https://list.coin-or.org/mailman/listinfo/ipopt
> 



More information about the Ipopt mailing list