dlmodeler.build.dseasonal: Build a "dummy seasonal" model

In dlmodeler: Generalized Dynamic Linear Modeler

Description

Builds an univariate "dummy seasonal" DLM of the specified order.

Usage

dlmodeler.dseasonal(ord, sigmaH = NA, sigmaQ = 0,
                    name = "dseasonal")

deterministic.season(ord, name="deterministic season")
stochastic.season(ord, name="stochastic season")

# old function name
dlmodeler.build.dseasonal(ord, sigmaH = NA, sigmaQ = 0,
                          name = "dseasonal")

Arguments

ord	period of the seasonal pattern.
sigmaH	std dev of the observation disturbance (if unknown, set to NA and use dlmodeler.fit to estimate it). Default = NA.
sigmaQ	std dev of the state disturbance (if unknown, set to NA and use dlmodeler.fit to estimate it). Default = 0.
name	an optional name to be given to the resulting DLM.

Details

The seasonal pattern is represented by ord seasonal indices a[1], a[2], ..., a[ord]. The indices are constrained such that their sum equals 0, with a[ord] = -a[1] - a[2] - a[3] ... -a[ord-1]. This only requires ord-1 state variables.

The initial value P0inf is parametered to use exact diffuse initialisation (if supported by the back-end).

The deterministic season model, is a special case of the dseasonal model, where sigmaH=0 and sigmaQ=0.

The stochastic season model, is a special case of the dseasonal model, where sigmaH=0 and sigmaQ=NA.

Value

An object of class dlmodeler representing the dummy seasonal model.

Note

State representations are not unique, so other forms could be used to achieve the same goals.

Author(s)

Cyrille Szymanski <[e-mail]>

References

Durbin, and Koopman, Time Series Analysis by State Space Methods, Oxford University Press (2001), pages 38-45.

See Also

dlmodeler, dlmodeler.build, dlmodeler.build.polynomial, dlmodeler.build.tseasonal, dlmodeler.build.structural, dlmodeler.build.arima, dlmodeler.build.regression

Examples

## Not run: 
require(dlmodeler)

# generate some quarterly data
n <- 80
level <- 12
sigma <- .75
season <- c(5,6,8,2)
y <- level + rep(season,n/4) + rnorm(n, mean=0, sd=sigma)

# deterministic level + quarterly seasonal + disturbance
mod <- dlmodeler.build.polynomial(0,sigmaH=sigma) +
       dlmodeler.build.dseasonal(4,sigmaH=0)
f <- dlmodeler.filter(y, mod)

# show the one step ahead forecasts
plot(y,type='l')
lines(f$f[1,],col='light blue')

## End(Not run)

Example output

dlmodeler documentation built on May 29, 2017, 11:33 a.m.