From: "Nick Cox" <[email protected]>
Reply-To: [email protected]
To: <[email protected]>
Subject: st: RE: generate a positive-valued random variable
Date: Sat, 30 Apr 2005 12:53:11 +0100
I understand an inverse gamma distribution to
be defined by taking the reciprocals of
a gamma distribution. It would normally have
either two or three parameters.
For two parameters, the Hilbe-Linde-Zwirble
program -rndgamma- offers one solution. Bobby
Gutierrez's program -gengamma- is more general
yet, covering both two and three parameter cases.
Naturally I can't tell you whether this
particular distribution, out of many defined
to be entirely positive, will do what you
want.
Nick
[email protected]
Jun Xu
> I am still using Stata 8. I am trying to generate a
> positive-valued random
> variable, and I think I should turn to inverse gamma
> distribution. I am not
> sure if this is right (with its shape parameter equal to 6):
>
> gen positivevar = invgamma(6, uniform())
>
> I tried searching Stata listserv and FAQ and I might have
> missed it. But I
> suggest Stata could post some codes for creating commonly used random
> variables in its online FAQ section. I think that will be
> great for a lot of intermediate users.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
_________________________________________________________________
