bsm: Basic Structural Time Series models

In gallegoj/tfarima: Transfer Function and ARIMA Models

Basic Structural Time Series models

Description

bsm creates/estimates basic structural models for seasonal time series.

Usage

bsm(
  y,
  bc = FALSE,
  seas = c("hd", "ht", "hs"),
  par = c(irr = 0.75, lvl = 1, slp = 0.05, seas = 0.075),
  fixed = c(lvl = TRUE),
  xreg = NULL,
  fit = TRUE,
  updmdl = NULL,
  ...
)

Arguments

y	an object of class ts, with frequency 4 or 12.
bc	logical. If TRUE logs are taken.
seas	character, type of seasonality (Harvey-Durbin (hd), Harvey-Todd (ht), Harrison-Steven (ht))
par	real vector with the error variances of each unobserved component.
fixed	logical vector to fix parameters.
xreg	matrix of regressors.
fit	logical. If TRUE, model is fitted.
updmdl	function to update the parameters of the BSM.
...	additional arguments.

Value

An object of class stsm.

References

Durbin, J. and Koopman, S.J. (2012) Time Series Analysis by State Space Methods, 2nd ed., Oxford University Press, Oxford.

Examples

bsm1 <- bsm(AirPassengers, bc = TRUE)

gallegoj/tfarima documentation built on March 31, 2024, 10:32 a.m.