Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | "Feiveson, Alan H. (JSC-SK311)" <alan.h.feiveson@nasa.gov> |
| To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
| Subject | RE: st: RE: loglink and normality, mixed model |
| Date | Mon, 24 Feb 2014 15:08:02 +0000 |
Hi Ruud - OK - I agree, that log(EY) is what you say and that depending on e[i,j], Y =can be non-positive. I assume that Y is regarded as continuous if Y > 0. But you also said you had a number of cases with Y = 0. This can't happen unless e[i,j] is a mixture of a discrete and continuous distribution, In that case, I would think glm would have problems because it (glm) is designed for distributions from an exponential family. Al -----Original Message----- From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Ruud van Lieshout Sent: Monday, February 24, 2014 8:34 AM To: statalist@hsphsun2.harvard.edu Subject: Re: st: RE: loglink and normality, mixed model Dear Al, thank you for your response. I do indeed mean an additive eps. The expectation of Y is therefore equal to a*t^b and the log of the expectation is then equal to log(E(Y))=log(a)+b*log(t). Im NOT transforming the observation itself, just its expectation. regards, Ruud * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/