IPOPT restoration

[Sahar]

Hello,

I wrote my optimization problem using Ampl and IPOPT as a solver. The
problem is when I slightly change my initial conditions, I get a
restoration failed problem or converged to an infeasible point even
though I'm sure that there's a solution. My problem is feasible and I
have no problems in the definition of the constraints since sometimes
it converged to an acceptable level solution.

Could you please help. I do not really understand what is going on in
the restoration phase.

Thank you!

[Robert Fourer]

For information specific to the Ipopt solver -- such as details of the
restoration phase -- it is best to contact the developers of the solver.
Contact information is given on the Ipopt homepage,
https://projects.coin-or.org/Ipopt.

Problems of convergence are often very subtle and model-specific. Thus it
is always helpful to supply model and data files for a simple example that
shows the problem that shows the error that you are seeing.

Bob Fourer
4...@ampl.com

[Sahar El Khoury]

thank you Bob, I indeed contacted the developers and it's working much better!

--
You received this message because you are subscribed to the Google Groups "AMPL Modeling Language" group.
To post to this group, send email to am...@googlegroups.com.
To unsubscribe from this group, send email to ampl+uns...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/ampl?hl=en.