Displaying 1 result from an estimated 1 matches for "hof3".
Did you mean:
bof3
2011 Oct 18
2
Non-linear maximization function in R
...,
0.88888889, 0.89444444, 0.94444444, 0.95555556)
# What I want to do is a log-likelihood model selection process where i
check if a Gaussian fit or a skewed unimodal fit (among others) is better.
# The first thing I do is use this function written by Oksanen
http://cc.oulu.fi/~jarioksa/softhelp/hof3.pdf
HOF.start<-function(y,x,model=5) {
if (model >=4) {
k<-glm(y ~ x + I(x^2),family=binomial)
k1<-k$coefficients[1]
k2<-k$coefficients[2]
k3<-k$coefficients[3]
u<-(-k2/2/k3)
h<-plogis(k1-k2^2/4/k3)
r<-log(1/h*(-2*h+2*sqrt(h))/2)
b<-5.07-0.227*k3
a&...