R help - Jul 2009 - package to do inverse probability weighting in longitudinal data

Hi there,

I have a dataset from a longitudinal study with a lot of drop-out. I 
want to implement the inverse probability weighting method by Robins 
1995 JASA paper "Analysis of semiparametric regression models for 
repeated outcomes in the presence of missing data". Does anyone know 
if there is a package to do it in R (or other software)? Thanks a lot!

Lei

Hi all,

I am trying to put the summary output of vglm{VGAM} into a Latex table using
mtable(Memisc}. I think I solved the problem regarding to the fact that vglm
produces a "vglm" object which is not accepted by mtable by default
defining a getSummary.vglm function.

However summary.vglm adds ":" to the end of the coefficient names
followed by the factor level while using the multinomial model. So the
coefficient names look like "(Intercept):1","(Intercept):2"
etc. However this creates a problem in mtable:

"Error in strsplit(coefnames, ":", fixed = TRUE) : non-character
argument"

Any suggestions ? 

Or, are there any other general suggestions about putting vglm summary output
into a Latex table using another method ?

All help is greatly appreciated.

Ugur   

Microsoft gives you windows, Linux gives you the whole house.