[R] Question on bivariate GEE fit

From: souvik banerjee <bansouvik_at_gmail.com>
Date: Tue, 08 May 2007 21:04:34 +0530

 I have a bivariate longitudinal dataset. As an example say, i have the data frame with column names

var1 var2 Unit time trt

(trt represents the treatment)

 Now suppose I want to fit a joint model of the form for the *i* th unit  var1jk = alpha1 + beta1*timejk + gamma1* trtjk + delta1* timejk:trtjk + error1jk

 var2 = alpha2 + beta2*timejk + gamma2* trtjk + delta2* timejk:trtjk + error2jk

where j index time and k index the treatment received

Using indicator variables I have been able to fit and run the code for a bivariate model using unstructured covariance matrix. However, I want to fit a model for a structured variance covariance matrix. The error structure for the grouping unit is as follows

sigma = ( sigma1      sigma12 )
             ( sigma12       sigma2)

sigma1, sigma2 and sigma12 are matrices with where
sigma1 = sig1 * AR1(rho1)
sigma2 = sig2* AR1(rho2)
sigma12 = sig12 * AR1(rho12)

My question is whether there is any method to fit such data using packages like gee or geepack (or may be any other package ) in R. The function genZcor() of geepack can be used to construct correlation but I have been unable to use it in the present context. Any help is greatly appreciated.

Souvik Banerjee
Department of statistics
Memari College

        [[alternative HTML version deleted]]

R-help_at_stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. Received on Tue 08 May 2007 - 16:06:57 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Tue 08 May 2007 - 16:31:25 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.