To the R time series wizards,
could you help me with an error message that I do not know how to
interpret.
I just want to try out the kalman filter implemented in the package
SSPIR on a random walk model:
> tmp <- ts(cumsum(rnorm(200)), start = c(2007,2), frequency=12)
> plot(tmp)
> m1 <- SS(tmp,m0=matrix(1),C0=matrix(2))
> kfilter(m1);
The error I get is:
"Error in matrix(NA, ss$n, ss$p) : non-numeric matrix extent"
What am I doing incorrectly ? It is OK to set phi to null, is it not ?
Thanks!!
Markus