[R] correaltion equal 1

[Douglas Bates]

On 11/3/06, Doran, Harold <HDoran at air.org> wrote:
> or, it could mean you need to recenter your time variable.
>
>
> -----Original Message-----
> From: r-help-bounces at stat.math.ethz.ch on behalf of Marc Bernard
> Sent: Fri 11/3/2006 7:24 AM
> To: r-help at stat.math.ethz.ch
> Subject: [R] correaltion equal 1
>
> Dear All,
>
>   I wonder if this is a technical or an interpretation problem. I fitted an lme model including a random intercept  and a random slope. The estimated correlation, from the lme,  between the random intercept and the random slope is equal to 1. Does it mean that I should suppress  one of the random effect from the model?

It probably indicates that the estimate of the variance-covariance
matrix for the random effects is singular.  If you refit the model
using lmer from the lme4 package and the estimate of the
variance-covariance matrix is singular, you will get a warning to that
effect.

As Harold mentioned, you may be able to avoid the problem by
re-centering the covariate.  However, this may also be a
characteristic of your data.