simulate.tsissm.estimate: Model Simulation

Description

Simulation function for class “tsissm.estimate”.

Usage

# S3 method for tsissm.estimate
simulate(
  object,
  nsim = 1,
  seed = NULL,
  h = 1,
  newxreg = NULL,
  sim_dates = NULL,
  bootstrap = FALSE,
  innov = NULL,
  innov_type = "q",
  pars = coef(object),
  init_states = tail(object$model$states, 1),
  init_res = NULL,
  init_sigma = NULL,
  ...
)
# S3 method for tsissm.selection
simulate(
  object,
  nsim = 1,
  seed = NULL,
  h = 1,
  newxreg = NULL,
  sim_dates = NULL,
  bootstrap = FALSE,
  pars = coef(object),
  init_states = tail(object$model$states, 1),
  init_res = NULL,
  init_sigma = NULL,
  ...
)

Value

An object of class “tsissm.simulate” with slots for the simulated series and states.

Arguments

object: an object of class “tsissm.estimate”.
nsim: the number of paths per complete set of time steps (h).
seed: a value specifying if and how the random number generator should be initialized (‘seeded’). Either NULL or an integer that will be used in a call to set.seed before simulating the response vectors.
h: the number of time steps to simulate paths for. If this is NULL, it will use the same number of periods as in the original series.
newxreg: an optional matrix of regressors to use for the simulation if xreg was used in the estimation. If NULL and the estimated object had regressors, and h was also set to NULL, then the original regressors will be used.
sim_dates: an optional vector of simulation dates equal to h. If NULL will use the implied periodicity of the data to generate a regular sequence of dates after the first available date in the data.
bootstrap: whether to bootstrap the innovations from the estimated object by re-sampling from the empirical distribution.
innov: an optional vector of innovations (see innov_type). The length of this vector should be equal to nsim x horizon.
innov_type: if ‘innov’ is not NULL, then this denotes the type of values passed, with “q” denoting quantile probabilities (default and backwards compatible) and “z” for standardized errors.
pars: an optional named vector of model coefficients which override the estimated coefficients. No checking is currently performed on the adequacy of these coefficients.
init_states: An optional vector of states to initialize the forecast. If NULL, will use the first available states from the estimated model.
init_res: For a dynamic variance model, the initialization for the ARCH recursion of length equal to max(p,q).
init_sigma: For a dynamic variance model, the standard deviation initialization for the GARCH recursion of length equal to max(p,q).
...: not currently used.