[Ipopt] ipopt not descending

[Filip Jorissen]

Hi all,

I was under the impression that IPOPT tries to minimize the objective and primal infeasibility. A step is accepted when either the objective or primal infeasibility reduces sufficiently. However, I’ve often spotted iterations where this does not seem to be the case, e.g. below:

iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
   6  3.1996216e-02 2.92e-05 6.24e-02  -3.3 1.65e-02    -  1.00e+00 1.00e+00h  1
   7  3.2386862e-02 1.70e-04 1.17e-01  -3.3 6.66e-02    -  1.00e+00 1.00e+00h  1

The objective, primal infeasibility and dual infeasibility all increase, alpha=1. Why does IPOPT accept this step?

Thanks!

Filip
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/ipopt/attachments/20190215/5f319a42/attachment.html>