In this video, we demonstrate how to use NumXL to construct and calibrate a seasonal ARIMA (aka SARIMA) model in Excel. We'll also project a forecast using the same model.
Hello and welcome to the NumXL tutorial demo for the seasonal ARIMA aka SARIMA model. In this example we will use the airline passenger monthly log data from between January 1949 and December 1960. To begin first select an empty cell in your worksheet. Next locate the ARMA icon in the NumXL toolbar and select the seasonal ARIMA aka SARIMA option.
The SARIMA wizard pops up note that the output range references the currently selected cell in your worksheet. Let's select the input data. Select the cells range in column C. Note that the different options of the wizard are now enabled. For the model order we will select SARIMA 0 1 1 0 1 1 with seasonal length of 12 months. This is, in essence an airline model but using the SARIMA model representation. For options let's have the wizard search and calibrate the model for optimal values. Now change the number of iterations to 200. When you're finished click OK.
And here you go the SARIMA model table along with the goodness of fit and residuals diagnosis tables are generated and displayed in your worksheet. Examining the residuals diagnosis table we see that the selected model fits the data pretty well as input data satisfies all model assumptions. For instance normality, lack of autocorrelation etc.
Starting with version 1.63 the NumXL wizard generates comments for the different output cells. Note the red arrow tips on the top right corners of the cells. To examine those comments select cell G21 mu and observe the comment bubble. Repeat the same steps for different cells in the output table.
Now that we have a calibrated and valid time series model let's conduct a forecast. Select the cell that heads the model table, in this case cell F19. Now locate the forecast icon in the NumXL toolbar and click it.
The forecast wizard pops up, note that the model input references the current selected cell in your worksheet. Now we need to select the latest observations in the sample data. Next change the number of steps to twelve.
In the output range select the empty cell on your worksheet where you wish the output forecast table to be displayed. When you're done click OK.
The forecast wizard generates the output forecast table in your worksheet. The forecast value along with the confidence interval is shown in the graph.
That's it for now, thank you for watching!
Note: For more information (e.g. example spreadsheet, screenshots and write-up), see SARIMA section in NumXL user's guide.