I'm not given a data set but rather just the model: (1 + .2B − .35B^2)(1 − 0.8B^12)Xt = Zt (Zt is a white noise RV). I'm trying to find SARIMA(p,d,q)x(P,D,Q)s for this model, but I'm not sure if I have enough information. Should I be looking at the ACF and PACF of this model in R? I know that I'm supposed to find d,D to make the model stationary, so I guess I can use R to find the roots and see whether this model is stationary. As for p,P,q,Q, I'm completely lost. When I multiplied everything out, I got an AR(14) model, but I don't think that's related to p,q,P,Q. Can anyone point me in the right direction?
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
2
By inspection the SARIMA model is (2,0,0)(1,0,0)12 because there is 1 ar polynomial with 2 coefficients with 0 differencing and 0 ma polyNomials THUS from left to right we have 2,0,0
Since there is 1 seasonal ar structure AND no seasonal differencing and no seasonal ma structure we have from left to right (1,0,0) 12
-
$\begingroup$ Thank you very much for responding. Could you please elaborate on why there's no differencing, or rather what that would look like? $\endgroup$ Commented Nov 5, 2017 at 19:48
-
$\begingroup$ if there was a regular differencing it would look like (1 − 1.0B^1)Xt or simply (1 −B^1)Xt . Seasonal differencing if included with look like (1 − B^12)Xt $\endgroup$ Commented Nov 5, 2017 at 21:28