Displaying 1 result from an estimated 1 matches for "modbis".
Did you mean:
mobil
2011 Jul 20
0
The C function getQ0 returns a non-positive covariance matrix and causes errors in arima()
...y small, and then, the likelihood
computed by KalmanLike is a number (and not NaN). Here is a function
allowing to compare the three methods
test <- function(phi,theta){
out <- makeARIMA(phi,theta,NULL)
Q0bis <- Q0bis(phi,theta)
Q0ter <- Q0ter(phi,theta)
mod <- out
modbis <- out
modter <- out
modbis$Pn <- Q0bis
modter$Pn <- Q0ter
set.seed(1)
x <- arima.sim(100,model=list(ar=phi,ma=theta))
s <- KalmanLike(x,mod=mod,fast=FALSE)
sbis <- KalmanLike(x,modbis)
ster <- KalmanLike(x,modter)
test12 <- all.equal(out$Pn,Q0b...