As an aside, if anyone is lookng to fit a non-symmetric distribution that
has support over all the reals (as opposed to a translated Gamma), I have a
program to do skew-normal regression. But it is not in a nice user-friendly
mode with a help file, syntax parsing for subsets, etc.
If anyone out there is interested in dressing it up as a new command with a
help file, etc I would be happy to supply the code. For those not familiar
with the skew-normal distribution, try this link:
http://azzalini.stat.unipd.it/SN/Intro/intro.html
Al Feiveson
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of n j cox
Sent: Wednesday, December 07, 2005 10:07 AM
To: [email protected]
Subject: Re: st: extended generalized Gamma distribution
Yes, but also be careful.
There are various three-parameter versions of the gamma. Whether -streg-
offers the version you want, or even one compatible with it by an easy
reparameterisation, will not be clear until you examine the details.
Also, I recollect -streg- has a built-in disposition to deal with responses
that are zero or positive only: do correct me if I am wrong. That makes
sense for survival analysis, but some writers on gamma distributions are
quite happy to define them with different support.
Again, this might be a matter merely of reparameterisation, or reexpression,
but watch out.
>>> Jamie Griffin
-streg- with distribution gamma fits the extended generalized gamma
distribution.
[email protected]
is anyone aware of a Stata program allowing for estimating the three
parameters of the extended generalized Gamma distribution?
*
* 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/
*
* 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/
