ACDfit-class: class: ACD Fit Class
In racd: Autoregressive Conditional Density Models

Description Objects from the Class Slots Methods Note Author(s) References Examples

S4 class containing the estimated ACD object.

Objects can be created by calling the acdfit function.

fit:: Object of class "vector" estimated model.
model:: Object of class "vector" model details.

getspec: signature(object = "ACDfit"): returns the specification used to estimate the model.
coef: signature(object = "ACDfit"): estimated coefficients.
convergence: signature(object = "ACDfit"): solver convergence code (0==converged).
infocriteria: signature(object = "ACDfit"): the information criteria for the ACD and restricted GARCH model.
likelihood: signature(object = "ACDfit"): log-likelihood of both the ACD and restricted GARCH model.
show: signature(object = "ACDfit"): summary method.
plot: signature(x = "ACDfit", y = "missing"): plots.
fitted: signature(object = "ACDfit"): conditional mean.
residuals: signature(object = "ACDfit"): the model residuals (with logical option standardize set to FALSE as default).
sigma: signature(object = "ACDfit"): conditional sigma.
skew: signature(object = "ACDfit"): conditional skew.
shape: signature(object = "ACDfit"): conditional shape.
skewness: signature(object = "ACDfit"): conditional skewness.
kurtosis: signature(object = "ACDfit"): conditional excess kurtosis.
quantile: signature(x = "ACDfit"): the conditional quantiles given a vector of probabilities.
pit: signature(x = "ACDfit"): the probability integral transform of the data given the conditional density.

If the skew0 and shape0 were both passed to the acdfit function, then no GARCH model will be estimated to obtain starting parameters, meaning that the ‘infocriteria’ and ‘likelihood’ methods will not have this information for comparison.
The ‘skew’ and ‘shape’ methods return the time varying skew and shape parameters, and have an additional logical option transformed (set to TRUE) which returns the logistic transformed parameters.
The ‘quantile’ method takes an additional vector probs as in the S3 stats package method.
The ‘residuals’ method takes an additional option standardize (default is FALSE) which returns the standardized version of the same.

Alexios Ghalanos

Hansen, B. E. 1994, Autoregressive conditional density estimation, International Economic Review, 35(3), 705–730.
Ghalanos, A., Rossi E., and Urga G. 2013, Independent Factor Autoregressive Conditional Density Model, Econometric Reviews, forthcoming.

## Not run: 
data(sp500ret)
spec = acdspec(variance.model=list(variance.targeting = TRUE),
mean.model=list(armaOrder=c(1,1)),distribution.model=list(model = "jsu",
skewOrder=c(1,1,0), shapeOrder=c(1,1,0)))
fit = acdfit(spec, sp500ret)
head(kurtosis(fit))
head(skewness(fit))
head(sigma(fit))
head(quantile(fit, probs=c(0.01)))

## End(Not run)