StatTimeSerAnalysis.ARMAInnovationsFit Function

Innovations ARMA estimation.

Pascal

procedure ARMAInnovationsFit(const Data: TVec; const Phi: TVec; const Theta: TVec; out Sigma2: double; const PhiSE: TVec = nil; const ThetaSE: TVec = nil; MaxLags: Integer = -1); overload;

File

StatTimeSerAnalysis

Parameters

Parameters	Description
Data	Zero-mean time series. If this is not the case, subtract the mean from data.
Phi	Returns estimates for phi coefficients phi[1]..phi[p]. AR(p) order is determined by Phi length.
Theta	Returns estimates for theta coefficients theta[1]..theta[q]. MA(q) order is determined by Theta length.
Sigma2	Returns estimate for Sigma^2 i.e. MA model variance.
PhiSE	If not nil, returns estimated phi coefficients standard errors.
ThetaSE	If not nil, returns estimated phi coefficients standard errors.
MaxLags	Defines maximum lag used in calculation of ACVF. If MaxLags is -1 then the following formula will be used to automatically set lag number: Ceil(10*Log10(Data.Length)).

Description

Uses innovations algorithm to predict ARMA(p,q) process coefficients.