s-news
[Top] [All Lists]

Beta function, amoeba, numerical methods.

To: <s-news@lists.biostat.wustl.edu>
Subject: Beta function, amoeba, numerical methods.
From: "Steve Su" <s.su@qut.edu.au>
Date: Wed, 19 Feb 2003 17:17:55 +1000
This is a multi-part message in MIME format.
Dear All,
 
Sometime ago I asked about the beta function which will become part of my equations which I need to solve to obtain parameters for generalised lambda distribution. I am currently aware of Starship methods proposed by King and MacGillivray(1999) in Australia and New Zealand Journal of Statistics as well as method of moments by Karian and Dudewicz(2000) as well as least square methods by Ozturk and Dale (1995), however I would like to examine some alternative methods in obtaining these parameters.
 
One of the fundamental difficulty I faced was with the calculation of the beta function through a numerical method. S+ NuOpt does not currently support gamma or lgamma functions, and
 
 
 
 
 
 
**************************************************************************************
 
 Steve Su (s.su@qut.edu.au)   
 PhD student. 
 
 School of Accountancy
 School of Mathematical Sciences
 Queensland University of Technology 
 
 Postal Address: Steve Su, School of Accountancy, QUT, PO Box 2434, Brisbane,
 Queensland, Australia, 4000.  
 

 Phone:  +61 7 3864 2017
 Fax:    +61 7 3864 1812     
 Mobile: 0421  840  586 
     .                              
   _--_|\                 
  /      QUT                                  
  \_.--._/                                                                          
        v                                              
                                                                     
**************************************************************************************                                                                  

Attachment: neldmead.txt
Description: Text document

<Prev in Thread] Current Thread [Next in Thread>
  • Beta function, amoeba, numerical methods., Steve Su <=