sysid (version 1.0.4)

predict.estpoly: Predictions of identified model

Description

Predicts the output of an identified model (estpoly) object K steps ahead.

Usage

"predict"(object, newdata = NULL, nahead = 1, ...)

Arguments

object

estpoly object containing the identified model

newdata

optional dataset to be used for predictions. If not supplied, predictions are made on the training set.

nahead

number of steps ahead at which to predict (Default:1). For infinite- step ahead predictions or pure simulation, supply Inf.

...

other arguments

Value

Time-series containing the predictions

References

Arun K. Tangirala (2015), Principles of System Identification: Theory and Practice, CRC Press, Boca Raton. Chapter 18

Examples

data(arxsim)
mod1 <- oe(arxsim,c(2,1,1))
Yhat <- predict(mod1,arxsim) #  1-step ahead predictions 
Yhat_2 <- predict(mod1,arxsim,nahead=2) # 2-step ahead predictions
Yhat_inf <- predict(mod1,arxsim,nahead=Inf) # Infinite-step ahead predictions