Fit mixture vector autoregressive models

Provides EM-estimation of mixture autoregressive models for multivariate time series

Usage

mixVARfit(y, model, fix = FALSE, tol = 10^-6, verbose = FALSE)

Arguments

y: a data matrix.
model: an object of class "MixVAR" with initial values of parameter for EM estimation.
tol: Threshold for convergence criterion.
fix: if TRUE, fix the shift parameters.
verbose: if TRUE print information during the optimisation.

Details

Estimation is done under the assumption of multivariate Gaussian innovations.

Value

An object of class MixVARGaussian with EM estimates of model parameters.

References

Fong PW, Li WK, Yau CW, Wong CS (2007). “On a Mixture Vector Autoregressive Model.” The Canadian Journal of Statistics / La Revue Canadienne de Statistique, 35(1), 135--150. ISSN 03195724, doi:10.1002/cjs.5550350112 .

Author

Davide Ravagli