TransWikia.com

Eacf table interpretation in R

Cross Validated Asked on December 8, 2020

I’m new to time series in R and have an assignment to identify the parameters for the AR and MA processes for a given time series, as well as to use eacf. Here are the results from the three functions:

enter image description here

enter image description here

enter image description here

From the ACF function we can conclude that MA(1) is a possible candidate, from PACF we can say that AR(2) is possible, which tells us that the process could be ARMA(2, 1). But I don’t understand how to interpret the results from the EACF table and how to get other candidate models from it. I can’t seem to find any guidelines either. Thank you in advance.

2 Answers

You are looking for the corners formed by the x's from the bottom right side moving towards the top left. In your case, this would indicate the following possibilities for your ARMA models: Yellow highlighter indicates possible corners

This means you can attempt to fit a ARMA(1,2) and ARMA(3,2) or even an ARMA(4,1). Hope this helps!

Answered by pwjvr on December 8, 2020

ACF and PACF help to identify either AR or MA but not ARMA modeling. They can be hint but nothing sure.

The EACF table is when you got cross, you have non-significant p-value for your order where a circle is the opposite.

But here, since your ARMA(2,1) seems to work for both graphic and eacf table, i'd say it could a good choice.

Like I always say, to achieve the best modeling, you have to have a goal, predictive and/or explanatory ?

If it's predictive, achieving a minimization of RMSE for example will be the best modeling while when it's explanatory, significance of your orders and good estimates will help you with a good hint based on the plots.

Answered by josef_joestarr on December 8, 2020

Add your own answers!

Ask a Question

Get help from others!

© 2024 TransWikia.com. All rights reserved. Sites we Love: PCI Database, UKBizDB, Menu Kuliner, Sharing RPP