I've been following this tutorial for pre-whitening time series in order to apply CCF.

In the example, the authors use an ARIMA(1,1,0) model, which is then defined as presenting the following polynomial expression:

$(1 - (1 + \phi_1)B + \phi_1B^2)y_t$

This translates into the following filter in R:

filter(y, filter=c(1,-(1+phi1),phi1), sides=1)

On the other hand, ARIMA models with only the AR component present the following polynomial representation:

$1 - \sum_{j=1}^{p}{\phi_jB^j}$

which implies a filter like this in R:

filter(y, filter=c(1, -arima_model$coeffs), sides=1)

How would it be a filter for an ARIMA(2,1,0) and for series including the MA component?

Sorry if the question looks awkward, I'm trying to implement the pre-whitening in collections of rasters, which may have different AR, I, and MA values.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.