Thank you for the information. I was wondering about the censoring. That helps.
Thank you.
------Original Message------
From: Verkuilen, Jay
Sender: [email protected]
To: '[email protected]'
ReplyTo: [email protected]
Subject: st: RE: Generalized Gamma
Sent: Jan 22, 2010 5:53 PM
Lori M. Ward wrote:
>>I am interested in estimating a two part generalized gamma model. My response
variable is cost, and I hace several predictor variables. I am interested in
the two part generalized gamma distribution because I have been told about its
incorporation of heterogeneity, and that STATA could automatically generate
initial parameter estimates for such a model. Is anyone familiar with running such a model in STATA?<<
Same answer as we gave you to the previous question about something similar:
-help streg-
-streg- will ordinarily assume you have censored data. If you don't have censored data, just make a variable that's the censoring indicator that never indicates a censored observation and it will drive on just fine. Whether the generalized gamma model is what you need, well that's a different question, but -streg- has a lot of other distributions suitable for a variable like cost.
*
* 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/
Lori M. Ward
*
* 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/
