bssm (version 1.0.0)

particle_smoother: Particle Smoothing

Description

Function particle_smoother performs filter-smoother or forward-backward smoother, using a either bootstrap filtering or psi-auxiliary filter with stratification resampling.

Usage

particle_smoother(model, nsim, ...)
# S3 method for nongaussian
particle_smoother(
  model,
  nsim,
  method = "psi",
  seed = sample(.Machine$integer.max, size = 1),
  max_iter = 100,
  conv_tol = 1e-08,
  ...
)
# S3 method for ssm_nlg
particle_smoother(
  model,
  nsim,
  method = "psi",
  seed = sample(.Machine$integer.max, size = 1),
  max_iter = 100,
  conv_tol = 1e-08,
  iekf_iter = 0,
  ...
)
# S3 method for ssm_sde
particle_smoother(
  model,
  nsim,
  L,
  seed = sample(.Machine$integer.max, size = 1),
  ...
)

Arguments

model

Model.

nsim

Number of samples.

...

Ignored.

method

Choice of particle filter algorithm. For Gaussian and non-Gaussian models with linear dynamics, options are "bsf" (bootstrap particle filter) and "psi" ($\psi$-APF, the default), and for non-linear models options "ekf" (extended Kalman particle filter) is also available.

seed

Seed for RNG.

max_iter

Maximum number of iterations used in Gaussian approximation. Used $\psi$-APF.

conv_tol

Tolerance parameter used in Gaussian approximation. Used $\psi$-APF.

iekf_iter

If zero (default), first approximation for non-linear Gaussian models is obtained from extended Kalman filter. If iekf_iter > 0, iterated extended Kalman filter is used with iekf_iter iterations.

L

Integer defining the discretization level.