Estimates the autoregressive-moving average (ARMA) model of a univariate or multivariate (vector) time series according to the method you specify. Wire data to the Xt input to determine the polymorphic instance to use or manually select the instance.


icon

TSA ARMA Modeling Details

This VI estimates the autoregressive (AR) model according to the following equation:

Xt + a1Xt-1 + ,…, + aNaXt-Na = et + b1et-1 + ,…, + bNbet-Nb

where Na is the AR order, Nb is the MA order, Xt is a univariate or multivariate (vector) time series, and et is a Gaussian white noise series with a mean of zero.

For univariate time series, AR coefficients is a 1D array [1, a1, a2,…, aNa], where each coefficient ai is a real number. MA coefficients is a 1D array [1, b1, b2,…, bNb], where each coefficient bi also is a real number.

For multivariate time series, AR coefficients is a 1D array [I, a1, a2,…, aNa], where each coefficient ai is a 2D array cluster. MA coefficients is a 1D array [I, b1, b2,…, bNb], where each coefficient bi also is a 2D array cluster.

The minimum length requirement for the input time series differs for each method you use:

  • Yule-Walker method: minimum length >= AR order + MA order
  • High-Order AR method: minimum length >= 5 x MA order
  • Polynomial method: minimum length >= 5 x (AR order + MA order)

Examples

Refer to the following VIs for examples of using the TSA ARMA Modeling VI:

  • Engine Knocking Monitor VI: labview\examples\Time Series Analysis\TSAApplications
  • ARMA Prediction VI: labview\examples\Time Series Analysis\TSAGettingStarted
  • ARMA Model Estimation VI: labview\examples\Time Series Analysis\TSAGettingStarted