It's here (search the sources for the appropriately named qbeta.c) but
I can't guarantee you'll understand it easily: Martin Maechler (and
others) works hard to make these the best in the business, but that
performance comes at a price, here paid in opacity:
http://svn.r-project.org/R/trunk/src/nm
Hi David:
Thanks, I actually realized it was error on my part. I didnt think there
was a problem with R-code, just wanted to get a better understanding of the
function qbeta.
Anamika
On Fri, Mar 9, 2012 at 3:04 PM, David Winsemius wrote:
>
> On Mar 9, 2012, at 2:48 PM, Anamika Chaudhuri wrote:
Take a look at n-x+1, the second parameter to the beta distribution:
> n <- c(10, 45, 38)
> x <- rbind(c( 7, 45, 31),
+c(10, 40, 35),
+c( 9, 44, 33),
+c( 8, 44, 31),
+c( 8, 45, 36))
> n - x + 1
[,1] [,2] [,3]
[1,]4 -6 15
[2,] 36 -29
On Mar 9, 2012, at 2:48 PM, Anamika Chaudhuri wrote:
HI All:
Does anyone know the code behind the qbeta function in R?
Well, yes, but don't you think it would be wise to question whether
your code might be the problem rather than the R code?
I am using it to calculate exact confidence i
4 matches
Mail list logo
