Michael,
Thanks for bringing this our attention. The most recent version of Amelia
II (posted Friday) fixes this bug. The program now saves the output before
removing it from memory.
Thanks again,
matt blackwell.
On Tue, 26 Jun 2007, Michael Richard Kellermann wrote:
Hi -
I ran across this problem when helping someone attempting to do imputation on
a large dataset. To conserve memory, she would like to write each imputed
dataset to a file and then discard it before moving on to the next
imputation. I think that should be possible using the keep.data=F and
write.out = T arguments. Setting these options appears to erase the imputed
datasets before they are sent to a file, and return NAs instead:
if (keep.data) {
impdata[[i]] <- impfill(x.orig = data, x.imp = ximp,
noms = prepped$noms, ords = prepped$ords)
names(impdata)[i] <- paste("m", i, sep = "")
}
else {
impdata[[i]] <- NA
}
if (write.out) {
write.csv(impdata[[i]], file = paste(prepped$outname,
i, ".csv", sep = ""))
}
Cheers,
Mike
---
Michael Kellermann
Ph.D Candidate, Department of Government
Harvard University
kellerm(a)fas.harvard.edu
http://people.fas.harvard.edu/~kellerm/
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive:
http://lists.gking.harvard.edu/?info=amelia
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive:
http://lists.gking.harvard.edu/?info=amelia