I want to calculate the
sum_{i}^{n} x_{i}^2 * lambda^(n-i) / sum_{i}^{n} lambda^(n-i)
Doing so in R i manage to do the following.But zeros occured.What am i doing wrong?
n = 10
x = seq(1,n,1);x
lambda = 0.99
mat = 0
for (i in 2:n) {
mat = (lambda^(n-i)*x[i-1]^2) / (lambda^(n-i))
}
mat
the result must be the variance of each entry x_{i}
CodePudding user response:
Very well, I will try to see if this is what you need @The Red:
n <- 10
x <- seq(1,n,1);x
lambda <- 0.99
mat <- 0
vect_mat <- rep(0, n-1); vect_mat
for (i in 2:n) {
mat <- mat (lambda^(n-i)*x[i-1]^2) / (lambda^(n-i))
vect_mat[i-1] <- mat
}
mat
vect_mat
After running it, it results in:
> mat
[1] 285
> vect_mat
[1] 1 5 14 30 55 91 140 204 285
Please accept the answer if this is what you need.