garma (version 0.9.24)

predict.garma_model: Predict future values.

Description

Predict ahead using algorithm of Godet (2009).

Usage

# S3 method for garma_model
predict(object, n.ahead = 1, newdata = NULL, ...)

Value

A "ts" object containing the requested forecasts.

Arguments

object

(garma_model) The garma_model from which to predict the values. This should have been generated by the [garma()] function.

n.ahead

(int) The number of time periods to predict ahead. Default: 1

newdata

(real vector or matrix) If the original model was fitted with the 'xreg=' option then this will provide the xreg values for predictions. If this is a vector then its length should be 'n.ahead'; if it is a matrix then it should have 'n.ahead' rows.

It should have columns with the same names as the original xreg matrix.

...

Other parameters. Ignored.

References

Godet, F. Linear prediction of long-range dependent time series, ESAIM: PS (2009) 13 115-134. DOI: https://doi.org/10.1051/ps:2008015

Examples

data(AirPassengers)
ap <- as.numeric(diff(AirPassengers, 12))
mdl <- garma(ap, order = c(9, 1, 0), k = 0, method = "CSS", include.mean = FALSE)
predict(mdl, n.ahead = 12)