ARMA.hstep: Computes h-step-ahead predictions from an ARMA(p,q) model

In gregorkb/tscourse: Contains several functions for time series analysis.

Computes h-step-ahead predictions from an ARMA(p,q) model

Description

Computes h-step-ahead predictions from an ARMA(p,q) model

Usage

ARMA.hstep(X, h, phi, theta, sigma)

Arguments

X	a vector containing time series data.
h	the number of steps ahead for which to make predictions.
phi	a vector with autoregressive coefficients.
theta	a vector the moving average coefficients.
sigma	the white noise variance.

Value

a list containing the predicted values as well as the MSPEs of the predictions and the AIC and BIC. This function builds a matrix of autocovariances for the ARMA(p,q) model using the MA(inf) representation of the process. It then runs the innovations algorithm on this matrix of autocovariances.

gregorkb/tscourse documentation built on Oct. 3, 2022, 5:31 p.m.