Yesterday I wrote a message asking for help extracting information from
an MI list object that had been subjected to some sort of statistical
analysis. I received one reply [shown below] from Tim Hesterberg,
Insightful corp which did exactly what I wanted.
#--------snip---------------------
For the most part, you can just wrap those commands inside a call to
miEval(). For example, if you would normally do:
summary(glmmPQL(y ~ time, random = ~ 1|id, fam = binomial,
data = data))$tTable)
where `data' has no imputations, then try
miEval(
summary(glmmPQL(y ~ time, random = ~ 1|id, fam = binomial,
data = data.imp))$tTable)
)
miEval() also works for complicated expressions, including compound
expressions inside {}, e.g.
{
line 1
line 2
etc
}
Respectfully,
Frank R. Lawrence
|