[R-sig-ME] Factor analysis with lme?

[Kevin Wright]

Just to follow up, I received no information regarding the factor
analytic model in lme, but learned that it can be done with asreml-r,
e.g.

d1=d1[order(d1$B, d1$A),]
m1 = asreml(y ~ A, data=d1, rcov=~B:facv(A, 1))

K Wright



On Jan 30, 2008 2:17 PM, Kevin Wright <kw.statr at gmail.com> wrote:
> I'm looking for an example of how (if?) lme can be used with factor
> analysis.  By that, I mean modeling the residual variance/covariance
> matrix as
>
> Sigma = (Gamma)(Gamma)' + D
>
> where Gamma is a vector (or matrix) corresponding to latent variables
> and D is a diagonal matrix.
>
> I have looked around but have not found any examples.  Pointers,
> comments, and code are all welcome.
>
> K Wright
>