# Step-by-step example of predicting time series with ARIMAX or ARMAX model?

Could someone give me a step-by-step example of time series prediction using ARIMAX or ARMAX model? The example doesn't need to be long or complicated. It could be for example forecasting temperature with past data of only ten values (e.g [15, 16, 17, 15, ..., 12])

You don't have to calculate autocorrelation etc. step-by-step but I need it to be clear A) What are you doing and why, B) Which data you're using to get what etc..

For example, I would like to see the procedure of finding AR-, MA-, and X- lags, possible integrations, finding the error terms $y_i - \widehat{y}_i = e_i$, solving the coefficients for AR-, MA- and X-parts and making the predictions.

• I am no time series expert, by any means, but there are some good examples in the SAS documentation (especially helpful if you are using SAS, of course, but pretty nice even if you aren't). – Peter Flom - Reinstate Monica Apr 8 '13 at 10:28
• Thank you for your help =) I would prefer nonetheless example here =) I don't want examples involving functions, I would love an example done by hand (to some level) =). Like you would be defining an algorithm and what it does at each step =) – jjepsuomi Apr 8 '13 at 10:35