GAS (version 0.3.3)

UniGASSpec: Univariate GAS specification

Description

Specify the conditional distribution, scaling mechanism and time--varying parameters for univariate GAS models.

Usage

UniGASSpec(Dist = "norm", ScalingType = "Identity",
           GASPar = list(location = FALSE, scale = TRUE,
                         skewness = FALSE, shape = FALSE, shape2 = FALSE))

Arguments

Dist

character Indicating the label of the conditional distribution. Available distribution can be displayed using the function DistInfo. Default valueDist = "norm".

ScalingType

character Indicating the scaling mechanism for the conditional score. Possible choices are "Identity", "Inv", "InvSqrt". Note that, for some distribution only ScalingType = "Identity" is supported, see the function DistInfo. When ScalingType = "InvSqrt" the inverse of the cholesky decomposition of the information matrix is used. Default value ScalingType = "Identity".

GASPar

list Containing information about which parameters of the conditional distribution have to be time-varying. location = TRUE refers to the location parameter, scale = TRUE refers to the scale parameter, skewness = TRUE refers to the parameter controlling the skewness, shape = TRUE refers to the shape parameter (e.g. the degree of freedom of the Student-t distribution), shape2 = TRUE refers to the second shape parameter. If the distribution specified in the Dist argument does not have, say, a shape parameter, the condition shape = TRUE or shape = FALSE is ignored. Note that, for some distributions, these labels are not strictly related to their literal statistical meaning. Indeed, for the Exponential distribution exp, the term location indicates the usual intensity rate parameter. See the DistInfo function for more details.

Value

An object of the class uGASSpec.

Details

All the information regarding the supported univariate conditional distributions can be investigated using the DistInfo function.

References

Ardia D, Boudt K and Catania L (2016). "Generalized Autoregressive Score Models in R: The GAS Package." http://ssrn.com/abstract=2825380.

Creal D, Koopman SJ, Lucas A (2013). "Generalized Autoregressive Score Models with Applications." Journal of Applied Econometrics, 28(5), 777-795. 10.1002/jae.1279.

Harvey AC (2013). Dynamic Models for Volatility and Heavy Tails: With Applications to Financial and Economic Time Series. Cambridge University Press.

Examples

# NOT RUN {
# Specify an univariate GAS model with Student-t
# conditional distribution and time-varying location, scale and shape parameter
library("GAS")

GASSpec = UniGASSpec(Dist = "std", ScalingType = "Identity",
                     GASPar = list(location = TRUE,
                                   scale = TRUE, shape = TRUE))

GASSpec
# }