Vector time series analysis involves more than one dependent time series variable, with possible interrelations or feedback between the dependent variables.
The VARMASIM subroutine generates various time series from the underlying VARMA models. Simulations of time series with known VARMA structure offer learning and developing vector time series analysis skills.
The VARMACOV subroutine provides the pattern of the autocovariance function of VARMA models and helps to identify and fit a proper model.
The VARMALIK subroutine provides the log-likelihood of a VARMA model and helps to obtain estimates of the parameters of a regression model.
The following subroutines are supported:
computes the theoretical cross covariances for a multivariate ARMA model
evaluates the log-likelihood function for a multivariate ARMA model
generates a multivariate ARMA time series
generates a multivariate normal random series
computes the characteristic roots of a multivariate ARMA model