The TimeSeriesOrders function automatically determines the following for nonseasonal,
univariate time series:
- The orders (p, d, q)
For descriptions of these orders, see the information about nonseasonal ARIMA models in https://otexts.com/fpp2/non-seasonal-arima.html.
- Whether to include a nonzero constant term (mean) and an explicit dependence on time (a drift term)
If a time series has a seasonal component, the function reports p, d, and q, but does not calculate or report seasonal orders sp, sd, and sq. (For descriptions of the seasonal orders, see the information about seasonal ARIMA models in https://otexts.com/fpp2/seasonal-arima.html.)
You can use the TimeSeriesOrders output as the orders table of the ARIMA (ML Engine) or VARMAX (ML Engine) function.