Displaying 1 result from an estimated 1 matches for "gsvr".
Did you mean:
gsr
2004 Feb 03
1
Error in f(x, ...) : subscript out of bounds
...is the case. All I change in the code is a constant, that is
hard-wired in (not data dependent and not parameter dependent),
furthermore, the constant is not used in any subscripting.
Sorry I cannot provide a toy example, but the likelihood function looks
like this:
lnL <- function(theta, gsvr, gsvR) {
# theta[1]= mu, theta[2]=gamma, theta[3]=kappa1,
theta[4]=kappa2
nn <- 252
d <- 0:nn
wd <- exp((theta[3] * d + theta[4] * d^2))/(sum(exp(theta[3] *
d + theta[4] * d^2)))
sigsq <- numeric(length(gsvR$ret))...