InformationMatrixARMA: Expected large-sample information matrix for ARMA

In FitARMA: Fit ARMA or ARIMA Using Fast MLE Algorithm

Description

The expected large-sample information matrix per observation for ARMA(p,q) models is computed.

Usage

InformationMatrixARMA(phi = numeric(0), theta = numeric(0))

Arguments

phi	AR coefficients
theta	MA coefficients

Details

The information matrix is derived by Box and Jenkins (1970).

Value

a matrix of order (p+q)

Author(s)

A.I. McLeod

References

Box and Jenkins (1970). Time Series Analysis: Forecasting and Control.

See Also

FitARMA

Examples

#The covariance matrix estimates of the parameters phi and theta in an ARMA(1,1)
#with phi=0.9 and theta=0.5 and n=200 is
v<-solve(InformationMatrixARMA(0.9,0.5))/200
v
#and the standard errors are
sqrt(diag(v))
  

FitARMA documentation built on May 2, 2019, 9:33 a.m.