[Ipopt] Testing IPOPT on a black block problem

Sebastian Nowozin nowozin at gmail.com
Mon Apr 13 12:17:21 EDT 2009


Hello Michael,


Michael Stevens wrote:

> I guess there are two possibilities:
> 
> a) IPOPT always needs an analytical gradient
> 
> b) The eval_grad_f function is not called if IPOPT options are set for a
> Quasi-Newton solution.
> 
> Any help would be greatly appreciated.

IpOpt works best with analytical gradient.  If you cannot provide one,
you can implement a simple finite differencing approximation yourself in
eval_grad_f.  The method will always be called when using IpOpt.

Also note that IpOpt provides a check-gradient option which simplifies
checking the correctness of the gradient computation.

Sebastian



More information about the Ipopt mailing list