[19040] in s-news-athena
[S] GLM, quasi likelihood , restriction to dispersion > 1
daemon@ATHENA.MIT.EDU (Spyros Vrontos)
Wed Aug 25 11:55:59 1999
Message-Id: <3.0.3.32.19990825184323.006e9538@stat-athens.aueb.gr>
Date: Wed, 25 Aug 1999 18:43:23 +0300
To: s-news@wubios.wustl.edu
From: Spyros Vrontos <svrontos@stat-athens.stats.aueb.gr>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Hi S+ users.
I fit the following model
test <- glm(formula=v1 ~ v2*v3,
family=quasi(link="log", var="mu^2"),
data=test.df, maxit = 50)
If I restrict dispersion=1 as following,
summary(test, dispersion = 1)
it works.
If I restrict dispersion=1.1 it works.
But if I put
summary(test, dispersion > 1)
it does not work.
Is there a way to restrict dispersion > 1 ?
Thanks a lot.
Spyros Vrontos.
Athens University of Economics and Business.
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
