| Date: | Mon, 12 Aug 1996 12:10:43 -0400 |
| Reply-To: | "Dr. Alan Cantor" <alan@DAISY.MOFFITT.USF.EDU> |
| Sender: | "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU> |
| From: | "Dr. Alan Cantor" <alan@DAISY.MOFFITT.USF.EDU> |
| Organization: | H. Lee Moffitt Cancer Center |
| Subject: | Re: SAS IML Code for Maximum Likelihood Esti |
|
Hu, Raymond wrote:
>
> Does anyone have a SAS IML code for maximum likelihood estimates (since my
> data are already in SAS files) I can share with? Any help would be greatly
> appreciated.
>
> Raymond Hu, Ph.D.
> Social and Scientific Systems
> Statistician/Programmer Analyst
> 7101 Wisconsin Avenue, Suite 1300
> Bethesda, MD 20814
> Office: 301-986-4870 ext 279
> Fax: 301-986-8051
> E-mail: RCH@S-3.COM
There are several functions available in PROC
IML (version 6.11) that can find the maximum value
of a function. They are described in a manual of
PROC IML enhancements through SAS 6.11. All you
need to do is define the likelihood (or log
likelihood) to be maximized and call one of these
functions. This manual also gives an example
finding the MLEs of Weibull parameters.
--
Alan B. Cantor, Ph.D.
Program Leader, Biostatistics Core Program
Division of Cancer Control
H. Lee Moffitt Cancer Center & Research Institute
(813) 632-1327 (Voice) (813) 979-6729 (FAX)
alan@daisy.moffitt.usf.edu
|