binve: Inverts a marginal log-linear parametrization

In ggm: Graphical Markov Models with Mixed Graphs

Inverts a marginal log-linear parametrization

Description

Inverts a marginal log-linear parametrization.

Usage

binve(eta, C, M, G, maxit = 500, print = FALSE, tol = 1e-10)

Arguments

eta	a vector of dimension t-1 where t is the number of cells of a contingency table.
C	A contrast matrix.
M	A marginalization matrix.
G	G is the model matrix of the loglinear parameterization with no constant term.
maxit	an integer, specifying the maximum number of iterations. Default 500.
print	a logical value: if TRUE, prints the criterion after each cycle.
tol	A small value specifying the tolerance for the convergence criterion. Default: 1e-10.

Details

A marginal log-linear link is defined by \eta = C (M \log p). See Bartolucci et al. (2007).

Value

A vector of probabilities p.

Note

From a Matlab function by A. Forcina, University of Perugia, Italy.

Author(s)

Antonio Forcina, Giovanni M. Marchetti

References

Bartolucci, F., Colombi, R. and Forcina, A. (2007). An extended class of marginal link functions for modelling contingency tables by equality and inequality constraints. Statist. Sinica 17, 691-711.

See Also

mat.mlogit

ggm documentation built on May 29, 2024, 7:27 a.m.