Presample Data for regARIMA Model Simulation
When simulating realizations from a regression model with ARIMA errors, the software requires presample unconditional disturbances and innovations to initialize the error process. The
regARIMA model property
P stores the number of presample unconditional disturbances that you need to initialize the simulation. The property
Q stores the number of presample innovations that you need to initialize the simulation.
You can specify your own presample data, or let
simulate generate presample data. If you let
simulate generate default presample data, then
simulate sets the required number of presample unconditional disturbances and presample innovations to 0.
simulate only accepts presample data for the error process, even if the response and predictors are time series.
- Simulate Regression Models with ARMA Errors
- Simulate Regression Models with Nonstationary Errors
- Simulate Regression Models with Multiplicative Seasonal Errors