[Ipopt] Manually loop the iteration

[Ronald Choe]

I am solving an optimization problem that can be broken down in multiple 
smaller optimization problems, ie distributedly. Each smaller problem 
can be solved individually, however, at each iteration step k, it needs 
the values of the set of optimization variables of (some of) the other 
optimization processes that are running in parallel. The problem is 
formulated as such that each smaller problem should be able to be solved 
using a NLP solver (without altering the actual solver algorithm).

I could supply a callback function at each iteration step where for 
example the constraints are computed, however, I am not familiar with 
the entire code and it consists of quite a number of files. So I was 
wondering which files (and functions) I should zoom in to and narrow my 
focus on.

Thanks!

On 7/8/15 11:07 PM, Greg Horn wrote:
> You can supply a callback, but what do you mean by supply the solver 
> with additional information? If you change the objective or constraint 
> function halfway through, bad things could happen
>
> On Wed, Jul 8, 2015 at 8:56 PM Ronald Choe <ronaldchoe at hotmail.com 
> <mailto:ronaldchoe at hotmail.com>> wrote:
>
>     I would like to manually loop the iteration so that at each
>     iteration I can supply the solver with additional (varying)
>     information (not dependent on the optimization variables), and
>     also extract the values of the optimization variables at the
>     respective time step.
>     Which particular file should I look at in order to achieve this?
>     Thanks!
>
>     Ronald
>     _______________________________________________
>     Ipopt mailing list
>     Ipopt at list.coin-or.org <mailto:Ipopt at list.coin-or.org>
>     http://list.coin-or.org/mailman/listinfo/ipopt
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/ipopt/attachments/20150708/8fd95cb4/attachment.html>