# Guides to VARMA modelling in R

I'm looking at using a VARMA model to both determine the driver so value in some advertising campaigns and also to forecast future activity. I'm looking at the paper by Takada and Bass as a reference point and I'm looking through the documentation for the MTS package by Tsay (the accompanying book is on order).

I've noticed that there does not seem to be much online for getting started with fitting VARMA moels, are there any open source docs that anyone knows of on how to get started? (While I wait for the book to arrive)

• MTS may be a slow package to work with. I have the impression that it was written as an illustration for the textbook but was not optimized in terms of speed. You may want to check out dse package or perhaps FKF, KFAS, dlm, MARSS and documentation for them. It helps if you understand the relation between state space models and VARMA models before you start working with dse; this may or may not be a hefty investment, but it will help your understanding of VARMA in addition to allowing you to use dse more easily. – Richard Hardy Nov 24 '14 at 9:40