Re: [R] Linear Trend in Residiuals From lme

From: Rick Bilonick <>
Date: Thu 10 Aug 2006 - 06:32:51 EST

On Wed, 2006-08-09 at 15:04 -0500, Douglas Bates wrote:
> On 8/9/06, Rick Bilonick <> wrote:
> > I'm fitting a mixed effects model:
> >
> > fit.1 <- lme(y~x,random=~1|id,data=df)
> >
> > There are two different observations for each id for both x and y. When
> > I use plot(fit.1), there is a strong increasing linear trend in the
> > residuals versus the fitted values (with no outliers). This also happens
> > if I use random=~x|id. Am I specifying something incorrectly?
> Could you provide a reproducible example please?
> I suspect that the problem comes from having only two observations per
> level of id. When you have very few observations per group the roles
> of the random effect and the per-observation noise term in explaining
> the variation become confounded. However, I can't check if this is
> the case without looking at some data and model fits.

Unfortunately, I can't send the actual data. I did make a simple intercepts-only example with two observations per group but it does not exhibit the linear trend.


x <- rnorm(20,5,1)
id <- factor(rep(1:20,each=2))

y <- as.vector(sapply(x,rnorm,n=2,sd=0.2))

df <- data.frame(id,y) <- groupedData(y~x|id,data=df)

summary(lme.1 <- lme(y~1,random=~1|id, plot(lme.1)

If I fit an intercepts-only model to the actual data, I still see the trend in the residuals.

What other analysis would you suggest?

Rick B. mailing list PLEASE do read the posting guide and provide commented, minimal, self-contained, reproducible code. Received on Thu Aug 10 06:37:59 2006

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.1.8, at Thu 10 Aug 2006 - 08:20:35 EST.

Mailing list information is available at Please read the posting guide before posting to the list.