R: Recipes for Analysis, Visualization and Machine Learning

Building an automated ARIMA model

The forecast package provides the auto.arima function to fit the best ARIMA models for a univariate time series.

Getting ready

If you have not already downloaded the files for this chapter, do it now and place them in your R working directory. Install and load the forecast package.

How to do it...

To build an automated ARIMA model, follow these steps:

1. Read the data. The file has monthly stock prices from Yahoo! Finance for Infosys between March 1999 and January 2015:

   > infy <- read.csv("infy-monthly.csv")

2. Create the time series object:

   > infy.ts <- ts(infy$Adj.Close, start = c(1999,3), frequency = 12)

3. Run the ARIMA model:

   > infy.arima <- auto.arima(infy.ts)

4. Generate the forecast using the ARIMA model:

   > infy.forecast <- forecast(infy.arima, ...