emgllmfitter {gllm}

Generalized log-linear modelling by EM and iterative proportional fitting

Description

Fits log-linear models for incomplete contingency tables, via an EM approach.

Usage

emgllmfitter(y,s,X,maxit,tol) 

Arguments

y

is the observed contingency table.

s

is a vector of indices, one for each cell of the full (unobserved) contingency table, representing the appropriate cell of y

X

is the design matrix.

maxit

is the number of EM iterations.

tol

is the convergence criterion for the LR criterion.

Details

The call to Andreas Borg's C code that fits the model by EM/IPF. The algorithm follows the approach in Haber (1984).

Value

A list with components:

y

the observed table

ji

s, the scatter vector

c

the design matrix

istop

maximum EM iterations

conv

the convergence tolerance

e

expected counts for the full (unobserved) table

ni

nrow(X)

nj

length(y)

nk

ncol(X)-1

f

expected counts

References

Haber M (1984). AS207: Fitting a general log-linear model. Appl Statist 33:358-362.