loglm
This function provides a frontend to the standard function,
loglin
, to allow loglinear models to be specified and fitted
in a manner similar to that of other fitting functions, such as
glm
.
Usage
loglm(formula, data, subset, na.action, …)
Arguments
 formula

A linear model formula specifying the loglinear model.
If the lefthand side is empty, the
data
argument is required and must be a (complete) array of frequencies. In this case the variables on the righthand side may be the names of thedimnames
attribute of the frequency array, or may be the positive integers: 1, 2, 3, … used as alternative names for the 1st, 2nd, 3rd, … dimension (classifying factor). If the lefthand side is not empty it specifies a vector of frequencies. In this case the data argument, if present, must be a data frame from which the lefthand side vector and the classifying factors on the righthand side are (preferentially) obtained. The usual abbreviation of a.
to stand for ‘all other variables in the data frame’ is allowed. Any nonfactors on the righthand side of the formula are coerced to factor.  data

Numeric array or data frame. In the first case it specifies the
array of frequencies; in then second it provides the data frame
from which the variables occurring in the formula are
preferentially obtained in the usual way.
This argument may be the result of a call to
xtabs
.  subset
 Specifies a subset of the rows in the data frame to be used. The default is to take all rows.
 na.action
 Specifies a method for handling missing observations. The default is to fail if missing values are present.
 …

May supply other arguments to the function
loglm1
.
Details
If the lefthand side of the formula is empty the data
argument
supplies the frequency array and the righthand side of the
formula is used to construct the list of fixed faces as required
by loglin
. Structural zeros may be specified by giving a
start
argument with those entries set to zero, as described in
the help information for loglin
. If the lefthand side is not empty, all variables on the
righthand side are regarded as classifying factors and an array
of frequencies is constructed. If some cells in the complete
array are not specified they are treated as structural zeros.
The righthand side of the formula is again used to construct the
list of faces on which the observed and fitted totals must agree,
as required by loglin
. Hence terms such as
a:b
, a*b
and a/b
are all equivalent.
Value
An object of class "loglm"
conveying the results of the fitted
loglinear model. Methods exist for the generic functions print
,
summary
, deviance
, fitted
, coef
,
resid
, anova
and update
, which perform the expected
tasks. Only loglikelihood ratio tests are allowed using anova
. The deviance is simply an alternative name for the loglikelihood
ratio statistic for testing the current model within a saturated
model, in accordance with standard usage in generalized linear
models.
Warning
If structural zeros are present, the calculation of degrees of
freedom may not be correct. loglin
itself takes no action to
allow for structural zeros. loglm
deducts one degree of
freedom for each structural zero, but cannot make allowance for
gains in error degrees of freedom due to loss of dimension in the
model space. (This would require checking the rank of the
model matrix, but since iterative proportional scaling methods
are developed largely to avoid constructing the model matrix
explicitly, the computation is at least difficult.) When structural zeros (or zero fitted values) are present the
estimated coefficients will not be available due to infinite
estimates. The deviances will normally continue to be correct, though.
References
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
See Also
Examples
library(MASS)
# The data frames Cars93, minn38 and quine are available
# in the MASS package.
# Case 1: frequencies specified as an array.
sapply(minn38, function(x) length(levels(x)))
## hs phs fol sex f
## 3 4 7 2 0
##minn38a < array(0, c(3,4,7,2), lapply(minn38[, 5], levels))
##minn38a[data.matrix(minn38[,5])] < minn38$f
## or more simply
minn38a < xtabs(f ~ ., minn38)
fm < loglm(~ 1 + 2 + 3 + 4, minn38a) # numerals as names.
deviance(fm)
## [1] 3711.9
fm1 < update(fm, .~.^2)
fm2 < update(fm, .~.^3, print = TRUE)
## 5 iterations: deviation 0.075
anova(fm, fm1, fm2)
# Case 1. An array generated with xtabs.
loglm(~ Type + Origin, xtabs(~ Type + Origin, Cars93))
# Case 2. Frequencies given as a vector in a data frame
names(quine)
## [1] "Eth" "Sex" "Age" "Lrn" "Days"
fm < loglm(Days ~ .^2, quine)
gm < glm(Days ~ .^2, poisson, quine) # check glm.
c(deviance(fm), deviance(gm)) # deviances agree
## [1] 1368.7 1368.7
c(fm$df, gm$df) # resid df do not!
c(fm$df, gm$df.residual) # resid df do not!
## [1] 127 128
# The loglm residual degrees of freedom is wrong because of
# a nondetectable redundancy in the model matrix.