Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: RE: large numbers in comb(n,k) function: no success


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: RE: large numbers in comb(n,k) function: no success
Date   Thu, 5 Feb 2009 13:32:50 +0000 (GMT)

--- On Thu, 5/2/09, Inna Becher wrote:
> > I have to calculate  probabilities for observations to
> > be sampled. That's why I use exp with lnfactorial...

--- On Thu, 5/2/09, Maarten buis wrote:
> Looks like you want to calculate a likelihood function. Is
> that right?
> If so, then you can use the fact that you almost always
> maximize the log likelihood, which will alow you to avoid
> taking the exp of lnfactorial.  

Some time ago I wrote a short text on how to move from the
probability density function of the beta distribution to
the log-likelihood function in Stata. This contains a 
similar problem except that the relevant function in the
beta is the Gamma function instead of the factorial function.
Stata has a -lngamma()- but no -gamma()- function, and this
text shows that for the computation of the likelihood a 
-lngamma()- function is much more convenient than a -gamma()-
function. You can find this text at:
http://home.fsw.vu.nl/m.buis/software/betafit.html

-- Maarten

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------






      

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index