Dear R-Users,
the problem I was facing it's solved. Actually, the code was ok from the
first time.
You define it as:
lw <- function(h, x, im)
{
peri1 <- per(x)
len <- length(x)
m <- len/im
peri <- peri1[2:(m+1)]
z <- c(1:m)
freq <- ((2*pi)/len) * z
result <- log(sum(f
I made some experiments and I conclude that if I change the interval (e.g.
say c(-100,100)!!!) then the minimum value changes in respect to the
simulated data.
So, probably the problem lies in the "optimal" values of the interval?
On Sat, Jul 26, 2008 at 1:00 PM, Fotis Papailias <[EMAIL PROTECTE
Hi,
thanks for your message.
You mean to rewrite the function like that:
lw <- function(d, x, im)
{
peri1 <- per(x)
len <- length(x)
m <- len/im
peri <- peri1[2:(m+1)]
z <- c(1:m)
freq <- ((2*pi)/len) * z
result <- log(sum(freq^(2*d-1)*peri))-(2*d)/m * sum(log(freq))
On 26/07/2008 7:40 AM, Fotis Papailias wrote:
Dear R-users,
I have sent another mail some hour ago about a matlab Code I was trying to
translate in R.
Actually I have found a simpler code originally written in S-PLUS for the
same function.
Author's page -> http://math.bu.edu/people/murad/method
Dear R-users,
I have sent another mail some hour ago about a matlab Code I was trying to
translate in R.
Actually I have found a simpler code originally written in S-PLUS for the
same function.
Author's page -> http://math.bu.edu/people/murad/methods/locwhitt/
===
5 matches
Mail list logo
