Dear all,
I am doing something wrong.
I am trying to apply a formula for sample size calculation as in the book
"Design and Analysis of Clinical Trials", from Chow et all.
There a suggested sample size strategy uses the formula
0.30/0.45=F(0.80,2n,2n)/F(0.025,2n,2n)
which gives n=96, as in the book.
(here F(alfa,k,n) is the upper (alfa)th quantile of an F distribution with
k,n degrees of freedom)
I have been trying to get n=96 using the following code in R.
val<-rep(NA,200)
for (n in 1:200) {
val[n]<- qf(0.8, 2*n, 2*n,lower.tail = TRUE, log.p = FALSE)/qf(0.025,
2*n, 2*n,lower.tail = TRUE, log.p = FALSE);
}
but val doesnot get any close to 0.30/0.45.
Thank you,
Manuel
ManuelPerera-Chang at fmc-ag.com wrote:> Dear all, > > I am doing something wrong. > > I am trying to apply a formula for sample size calculation as in the book > "Design and Analysis of Clinical Trials", from Chow et all. > > There a suggested sample size strategy uses the formula > > 0.30/0.45=F(0.80,2n,2n)/F(0.025,2n,2n) > > which gives n=96, as in the book. > > (here F(alfa,k,n) is the upper (alfa)th quantile of an F distribution with > k,n degrees of freedom) > > I have been trying to get n=96 using the following code in R. > > val<-rep(NA,200) > for (n in 1:200) { > val[n]<- qf(0.8, 2*n, 2*n,lower.tail = TRUE, log.p = FALSE)/qf(0.025, > 2*n, 2*n,lower.tail = TRUE, log.p = FALSE); > }Don't you want 'lower.tail = FALSE'? Peter Ehlers> > but val doesnot get any close to 0.30/0.45. > > Thank you, > > Manuel > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html