[Coin-ipopt] Cox proportional-hazards regression

[yi mu]

Hi, All
 
I try to use to Ipopt to fit my survey data to the cox proportional hazard regression model, i know this can be easily done by using newton raphson iteration, but I try to learn ipopt and also want to see whether it is more efficient than newton raphson.  I can't get any solution, the derivative checker said that my hessian matrix is not correct, but the hessian matrix produced by ipopt is not symmetric. Can any one tell me why? since I don't have any constraints, should I return false for the gradient and Jacobin function or return [] like someone suggest in this forum? thanks. 
 
hellen
_________________________________________________________________
Peek-a-boo FREE Tricks & Treats for You!
http://www.reallivemoms.com?ocid=TXT_TAGHM&loc=us
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.coin-or.org/pipermail/ipopt/attachments/20071013/ace78a6a/attachment.html