MVB (version 1.1)

stepfit: step-wisd multivariate model fitting

Description

stepwise fit multivariate log-linear Bernoulli model using Newton-Raphson algorithm.

Usage

stepfit(x, y, maxOrder = 2, output = 0, direction = c("backward", "forward"), tune = c("AIC", "BIC", "GACV", "BGACV"), start = NULL)

Arguments

x

input design matrix.

y

output binary matrix with number of columns equal to the number of outcomes per observation.

maxOrder

maximum order of interactions to be considered in outcomes.

output

with values 0 or 1, indicating whether the fitting process is muted or not.

direction

the mode of stepwise search and default is backward.

tune

tuning approach, available methods including AIC, BIC, GACV, BGACV.

start

starting object of type mvbfit.

Value

An object of class mvbfit, for which some methods are available.

Details

The stepfit utilize the class structure of the underlying C++ code and stepwisd fitted the model with Newton-Raphson algorithm.

See Also

mvblps, unifit, stepfit, mvb.simu

Examples

# fit a simple MVB log-linear model
n <- 1000
p <- 5
kk <- 2
tt <- NULL
alter <- 1
for (i in 1:kk) {
  vec <- rep(0, p)
  vec[i] <- alter
  alter <- alter * (-1)
  tt <- cbind(tt, vec)
}
tt <- 1.5 * tt
tt <- cbind(tt, c(rep(0, p - 1), 1))

x <- matrix(rnorm(n * p, 0, 4), n, p)
res <- mvb.simu(tt, x, K = kk, rep(.5, 2))
fitMVB <- mvbfit(x, res$response, output = 1)