anova.islasso {islasso}R Documentation

General Linear Hypotheses method for islasso objects

Description

General linear hypotheses for linear combinantions of the regression coefficients in islasso fits

Usage

## S3 method for class 'islasso'
anova(object, A, b, ...)

Arguments

object

a fitted model object of class "islasso".

A

a specification of the linear hypotheses to be tested. Linear functions can be specified by either a single vector (of length p) or by a matrix (of dimension k x p) of one or more linear hypotheses.

b

an optional numeric vector specifying the right hand side of the hypothesis. Can be a scalar.

...

not used.

Details

For the islasso regression model with coefficients beta, the null hypothesis is

H_0: A β=b

where A and b are known matrix and vector. A can be a vector and b can be a scalar

Examples

## Not run: 
set.seed(1)
n <- 100
p <- 100
p1 <- 10  #number of nonzero coefficients
coef.true <- sort(round(c(seq(.5, 3, l=p1/2), seq(-1, -2, l=p1/2)), 2))
sigma <- 1

coef <- c(coef.true, rep(0, p-p1))

X <- matrix(rnorm(n*p), n, p)
eta <- drop(X %*% coef)
mu <- eta
y <- mu + rnorm(n, 0, sigma)

o <- islasso(y~-1+X, family=gaussian)
A <- rbind(rep(c(1,0), c(10, p-10)),
           rep(c(0,1), c(10, p-10)))
anova(o, A)

A <- cbind(diag(10), matrix(0, 10, p-10))
b <- coef.true
anova(o, A, b)

## End(Not run)

[Package islasso version 1.1.0 Index]