Displaying 2 results from an estimated 2 matches for "lme0".
Did you mean:
lme
2009 Apr 01
3
How to prevent inclusion of intercept in lme with interaction
...pe*grd$t+rnorm(nrow(grd),0,0.2)
xyplot(newbone~t|treat,groups=subject,data=grd,
type="l",xlim=c(0,20),ylim=c(0,3))
# With intercept
grd.lme1 = lme(newbone~t*treat,data=grd,random=~1|subject)
grd$pred1 = predict(grd.lme1,level=0)
summary(grd.lme1)
# How go force intercept = 0 ???
grd.lme0 = lme(newbone~t*treat-1,data=grd,random=~1|subject)
grd$pred0 = predict(grd.lme0,level=0)
summary(grd.lme0)
# Gives true,
all.equal(grd$pred1,grd$pred0)
# Everything as expected without treat
grd.lme2 = lme(newbone~t,data=grd,random=~1|subject)
grd$pred2 = predict(grd.lme2,level=0)
summary(grd.lme...
2006 Aug 23
0
Random structure of nested design in lme
...h level of 'soiltype' occurs only once within each
level of 'destination' in the self-contained example you provided below.
To confirm this, I deleted 'soiltype' from this model:
fit.lme <- lme(response~soiltype*habitat, random=~1|destination/origin)
fit.lme0 <- lme(response~soiltype*habitat, random=~1|destination)
The answers seemed to be identical except for one thing:
> VarCorr(fit.lme)
Variance StdDev
destination = pdLogChol(1)
(Intercept) 0.004149471 0.06441639
origin = pdLogChol(1)
(Intercept)...