Output from fitted after loglm - r

Question

I am trying to reproduce some log-linear modeling analysis from Agresti's Categorical Data Analysis (3rd ed.) [CDA] using the loglm function from the MASS package:
library(MASS)
# read in the data: http://www.stat.ufl.edu/~aa/cda/data.html
dfX = read.table(textConnection('a c m r g count
1 1 1 1 1 405
1 1 1 2 1 23
1 2 1 1 1 13
1 2 1 2 1 2
2 1 1 1 1 1
2 1 1 2 1 0
2 2 1 1 1 1
2 2 1 2 1 0
1 1 2 1 1 268
1 1 2 2 1 23
1 2 2 1 1 218
1 2 2 2 1 19
2 1 2 1 1 17
2 1 2 2 1 1
2 2 2 1 1 117
2 2 2 2 1 12
1 1 1 1 2 453
1 1 1 2 2 30
1 2 1 1 2 28
1 2 1 2 2 1
2 1 1 1 2 1
2 1 1 2 2 1
2 2 1 1 2 1
2 2 1 2 2 0
1 1 2 1 2 228
1 1 2 2 2 19
1 2 2 1 2 201
1 2 2 2 2 18
2 1 2 1 2 17
2 1 2 2 2 8
2 2 2 1 2 133
2 2 2 2 2 17'), header = TRUE)
llACM = loglm(count ~ c + a + m, data = dfX)
summary(llACM)
fitted(llACM)
But I am having difficulty understanding what the .Within. = means, and how I can get a predicted contingency table as given in CDA on page 323.

Answer 1

Really a long comment:
I ran your example, and took a look at things like fitted.loglm which consists of the following code:
{
if (!is.null(object$fit))
return(unclass(object$fit))
cat("Re-fitting to get fitted values\n")
unclass(update(object, fitted = TRUE, keep.frequencies = FALSE)$fitted)
}
update generates all the info that goes into object$fitted . I compared all sorts of data in the MASS example
minn38a <- xtabs(f ~ ., minn38)
fm <- loglm(~ 1 + 2 + 3 + 4, minn38a)
And about the only guess I can make is that your data is not properly dimensioned--or the fitting model adds a needed dimension, and this extra dim is given the default name .Within. . My suggestion would be to read the MASS book, or dig up info on glm fitting models. I agree that the explanations of the dimensions defined in the fitted dataset are somewhat lacking.