Skip to content

nested ANOVA using lme

2 messages · David Semmens, Dimitris Rizopoulos

#
I have an ANOVA model with 2 factors "Environment" and "Site", 
"Diameter" is the response variable. Site should be nested within 
Environment. Site is also a random factor while Environment is fixed. I 
can do this analysis using the "aov" function by using these commands:

 >model<-aov(Diam~Env+Error(Env%in%Site),data=environ)
 >summary(model)
 >summary(aov(Diam~Env/Site,data=environ))

But the model is unbalanced and I want to calculate estimates of the 
variance components. In this case the REML approach should give more 
reliable estimates than the method I have used. But I am unable to work 
out how to specify a nested factor using the "lme" function. I have 
tried:

 >firstgo<-lme(Diam~Env+Env:Site,data=environ)
 >secondgo<-lme(Diam~Env+Env%in%Site,data=environ)
 >thirdgo<-lme(Diam~Env+Env/Site,data=environ)

But I keep getting this error:

Error in getGroups.data.frame(dataMix, groups) :
	Invalid formula for groups

If anyone could help me with code for specifying a random nested factor 
using lme that would be great.

Thanks,
David Semmens.
#
in lme() you need to specify the random-effects with a formula passed 
to the 'random' argument, e.g., something like

fit <- lme(Diam ~ Env + Site, random = ~ 1 | Env / Site, data = 
environ)
summary(fit)


I hope it helps.

Best,
Dimitris

----
Dimitris Rizopoulos
Ph.D. Student
Biostatistical Centre
School of Public Health
Catholic University of Leuven

Address: Kapucijnenvoer 35, Leuven, Belgium
Tel: +32/(0)16/336899
Fax: +32/(0)16/337015
Web: http://med.kuleuven.be/biostat/
     http://www.student.kuleuven.be/~m0390867/dimitris.htm


----- Original Message ----- 
From: "David Semmens" <d.semmens at pgrad.unimelb.edu.au>
To: <r-help at stat.math.ethz.ch>
Sent: Wednesday, August 09, 2006 11:03 AM
Subject: [R] nested ANOVA using lme
Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm