simuHMM_cpp function

Simulate Hidden Markov model with normally distributed errors

This function simulates a Hidden Markov Model process.

simuHMM_cpp(mdl_h0, burnin = 100L, exog = FALSE)

Arguments

• mdl_h0: List containing the following DGP parameters

  • n: Length of series.
  • k: Number of regimes.
  • mu: A (k x q) vector of means.
  • sigma: A (q x q) covariance matrix.
  • q: Number of series.
  • P: A (k x k) transition matrix (columns must sum to one).
  • eps: An optional (T+burnin x q) matrix with standard normal errors to be used. Errors will be generated if not provided.
  • Z: A (T x qz) matrix with exogenous regressors (Optional) and where qz is the number of exogenous variables.
  • betaZ: A (qz x q) matrix true coefficients on exogenous regressors (Optional) and where qz is the number of exogenous variables.

• burnin: Number of simulated observations to remove from beginning. Default is 100.

• exog: bool determining if there are exogenous variables (true) or not (false). Default is false.

Returns

List with simulated series and its DGP parameters.