Replication (version 0.1.2)

llratio.f: Likelihood ratio statistic for constrained versus unconstrained model

Description

The function first uses solve.QP to find the best fitting estimates under the imposed constraints. Subsequently, it calculates the likelihood ratio for the constrained versus unconstrained model.

Usage

llratio.f(BKcov, Q, R, r = NULL, E = 0L)

Arguments

BKcov

BKcov is the variance-covariance matrix of the model parameters.

Q

Q is a vector with the (unconstrained) model estimates.

R

A matrix defining the constraints under which we want to minimize the quadratic function.

r

vector holding the minimimum sum for each row in R.

E

Numeric. The first E constraints are treated as equality constraints, all further as inequality constraints. Default value = 0.

Value

llratio

scalar, the value of the resulting likelihood ratio.

See Also

solve.QP

Examples

# NOT RUN {
data <- data.frame(y=ChickWeight$weight,g=ChickWeight$Diet)
aggregate(data$y,by=list(data$g),mean)
reg <- lm(data$y~data$g)
BKcov <- vcov(reg)
Q <- reg$coefficients

#b1=102, b3-b2>20, b3-b4>10, b4>30
R <- rbind(c(1,0,0,0),c(0,-1,1,0),c(0,0,1,-1),c(0,0,0,1))
r = c(102,20,10,30)
E = 1

llratio.f(BKcov=BKcov, Q=Q, R=R, r=r, E = E)
# }