<div dir="ltr"><div>Hi Ramesh -</div><div><br></div><div>Yes, Ipopt will run in quasi-Newton mode by approximating the Hessian. Set the option "hessian_approximation" to "limited-memory". Now instead of calling eval_h, Ipopt will approximate the Hessian using the gradient information from a limited number of previous iterations. I believe the approximation is the BFGS method but I may be wrong. Note also that you can set the number of previous iterations used by setting a value for the option "limited_memory_max_history".</div><div><br></div><div>- Seth<br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Apr 17, 2020 at 9:15 AM Ramesh Kolluru <<a href="mailto:rameshkolluru@icloud.com">rameshkolluru@icloud.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Dear All<br>
<br>
I am trying to work around IPOPT in fortran stable version. I could compile and run few examples where I could provide the function, gradients, constraints and their gradients along with the Hessian of Lagrangian.<br>
<br>
I wanted to know if IPOPT can be used with out providing the Hessian information as this belong to class of quasi Newton methods. Can some one provide me an example of how to use IPOPT without Hessian information.<br>
<br>
Regards <br>
_______________________________________________<br>
Ipopt mailing list<br>
<a href="mailto:Ipopt@list.coin-or.org" target="_blank">Ipopt@list.coin-or.org</a><br>
<a href="https://list.coin-or.org/mailman/listinfo/ipopt" rel="noreferrer" target="_blank">https://list.coin-or.org/mailman/listinfo/ipopt</a><br>
</blockquote></div>