Date: Mon, 10 Aug 1998 15:52:17 GMT
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
Organization: Deja News - The Leader in Internet Discussion
Subject: Re: How to maximize a likelihood function?
Proc NLP is the perfect proceedure to implement custom MLEs. A NLP solver is
also available within IML. If you have SAS/OR loaded, type 'help nlp' at the
command prompt for syntax.
In article <3Q0MVU$zbR@bbs.agec.ntu.edu.tw>,
firstname.lastname@example.org ($j@Y) wrote:
> !0 $^-email@example.com (Larry Van Horn)!n$';J(%!G
> : Folks,
> : I would like to give sas a specific objective function and perform a
> : maximum likelihood estimation. SAS suggested that I use PROC NLP from
> : the OR package. I couldn't find it in the OR manual. Is this the
> : correct procedure? Is there IML code I should be using? Any
> : direction here would be helpful. Thanks in advance,
> : Larry
> the question what you asked that I meet before. I have asked the SAS
> compapy,but I do not have the answer. So what I suggest is that:
> 1.You sould be sure what problem you have,and you must know the different
> between OR and IML procedure.
> 2.If you do not have a linear programing problem,In my opion,NLP will give
> no help.You should use IML procedure,but it is not easy to learn.You must
> write it step by step,and you must be sure the matrix calculate is ok!
> 3.the another way you can do is enter the MACRO for SAS,and try to make it
> useful,but you must have a good computer skill.
> anyway,SAS is too easy for users,and you can not easy to do any change,in
> another word,SAS is not flexible for maximum lilelihood estimation. I suggest
> you can try the TSP . TSP is what I use to do the maximum lilelihood estima.
> ps: I have a poor English,so God help you that you can understand what I say
> well.I wish it can be helpful for you.
> !0 Origin: %x$j9A8g !; From: 18.104.22.168
> !9 Origin: %x$j9A8gBBS/8 (bbs.agec.ntu.edu.tw) !9
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum