Hello everybody,
I am trying to compute the following double summation and am encountering
problems; I wonder if you can help? Basically I have a matrix B of data
points (data frame?) and want to create a matrix g such that its (i,t) entry
is defined as follows:
g(i,t) = sum_{s = 1}^T
(Indicator function(n[t] = n[s]))
sum_{j = 1}^{n[t]}
1/w*K( (B[i,t] - B[j,s]) / w)
where K is the kernel estimator, w is the bandwidth, and n is another
vector. How would you go about this? I tried something like
g <- matrix(0,N,T)
for (t in 1:T) {
for (i in 1:N) {
for (s in 1:T) {
kern <- density(x = B[1:n[t],s], bw = "nrd0", kernel =
"gaussian", weights
= NULL, n = T)
g[i,t] <- ghat[i,t] + (n[s] == n[t])*(kern$y[i])
}
}
}
but R just freezes completely...
Sorry for lack of clarity; I'm very new to R...
Thank you very much!
--ML
[[alternative HTML version deleted]]