I doubt it -- though perhaps you'll hear that someone has done this.
As an alternative you can put a penalty in the objective function for
breaking the constraints.
The more it breaks the constraints the higher the penalty should be.
Then you can use nlminb
or whatever. If you need more speed than that, S Poetry has a section
on putting the optimization
down into C. (In putting the penalty on, you need to not have a jump at
the boundary of the
constraint, or the derivative estimator will get very confused.)
Good luck,
Patrick Burns
pburns@pburns.seanet.com
http://www.seanet.com/~pburns/Spoetry
http://www.i-way.co.uk/~pat/Spoetry
Sofia Ziagka wrote:
>Dear all,
>
>Does anyone know if there is a library in S-plus that supports non-linear
>optimization with non- linear constraints ?
>That is, if there are any codes - tools in a S-Plus library for the
above
>analysis?
>
>Thank you all in advance,
>Sofia Ziagka
>
>
>----------------------------------------------------
>Sofia Ziagka
>I - Consulting
>
>Tel : +30 10 67 98 883
>Fax: +30 10 67 73 589
>sofia.ziagka@initiativemedia.com
>
>
>--------------------------------------------------------------------
>This message was distributed by s-news@lists.biostat.wustl.edu. To
>unsubscribe send e-mail to s-news-request@lists.biostat.wustl.edu with
>the BODY of the message: unsubscribe s-news
>
>
|