procedure ARMAForecast(const Data: TVec; const P: TVec; const T: TVec; const Residuals: TVec; const n: Integer; const mu: Double; const Forecast: TVec; const fStdDev: TVec);
Forecast time series by using ARMA(p,q) model.
| # | Name | Description |
|---|---|---|
| 1 | Data | The original time series data set. |
| 2 | P | ARIMA Phi (AR) coefficients. Assumes P (AR model) to be without the leading 1.0. |
| 3 | T | ARIMA Theta (MA) coefficients. |
| 4 | Residuals | Residuals as returned by ARMAMLE. |
| 5 | n | Number of samples to forecast. |
| 6 | mu | This modified average value is included in to the optimization process of ARMAMLE, which returns optimal value for it. |
| 7 | Forecast | Results of the forecasting (beyond the last index value of Data starting at Data.Length). |
| 8 | fStdDev | Returns standard deviation of residuals. |
Result: stored in self (calling object)