Hi
Is there any way to fit a bivaraite longitudinal mixed model using R. I have
a data set with col names
resp1 (Y_ij1), resp2 (Y_ij2), timepts (t_ij), unit(i)
j=1,2,..,m and i=1,2,..n.
I want to fit the following two models
Model 1
Y_ij1, Y_ij2 | U_i = u_i ~ N(alpha + u_i + beta1*t_ij, Sigma)
U_i ~ iid N(0, sigu^2)
Sigma = bivariate AR structure
alpha and beta are vectors of order 2.
Model 2
Y_ij, Y_ij2 | U_i = u_i ~ N(alpha + u_i + beta1*t_ij + U_i2*t_ij, Sigma)
U_i=(U_i1,U_i2) ~ iid N(0, Omega)
Sigma =bivariate AR structure
Here alpha and beta are vectors of order 2.
Moreover can we assume a structure for Omega.
Any help is greatly appreciated.
Souvik Banerjee
Junior Research Fellow
Dept of Statistics
University of calcutta
[[alternative HTML version deleted]]