i´m really sorry, once again. ok i will try to explain what i have to
programm. i want to programm a powerfunction.
i have to research if the correlations in a bivariate random sample are
homogeneous. for that i saperate the random sample in m blocks and
calculate the correlation of each block(partial sample). than i look at the
random variable p=cor(x1,y1)^2/sum(cor(x,y)^2) which is a probability. my
null hypothesis is that all correlations are homogeneous. if we have this
situation the entropy take the value log(m).
my test based on the entropy. so my teststatic is H=log(m)-sum(p*log(p)).
the following programm was actually, which i have worked the most of the
time. i hope that i don´t confuse u too much.
i of course i hope u understand my problem and my theme.
n=1000
m=2
k=n/m
N=100
myfun <- function(n, m, alpha = .05, seeder = 1000) {
l=matrix(0,nrow=m,ncol=N)
for(i in 1:N){
set.seed(i)
for(j in 1:m){
x=rnorm(n,0,0.5)
y=rnorm(n,0,0.8)
l[j,i]=cor((x[(((j-1)*k)+1):(((j-1)*k)+k)]),
(y[(((j-1)*k)+1):(((j-1)*k)+k)]))
}
}
gute <- function(x,m) {
q_1 <- qnorm(alpha, 0, 0.05)
q_2 <- qnorm(1 - alpha, 0, 0.05)
p=matrix(0,nrow=m,ncol=N)
H=matrix(0,nrow=N,ncol=1)
for(i in 1:N){
for (j in 1:m){
p[j,i]=x[j]^2/sum(x[,i]^2)
}
H[i]=log(m)-sum(p[,i]*log(p[,i]))
}
1 - mean(q_1 <= H & H <= q_2)
}
output <- gute(x = l,m=m)
return(output)
}
--
View this message in context:
http://r.789695.n4.nabble.com/the-function-doesn-t-work-tp2714105p2714137.html
Sent from the R help mailing list archive at Nabble.com.
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
