logLik-AIC-methods: Extract Log-Likelihood and Akaike's Information Criterion

In ghyp: Generalized Hyperbolic Distribution and Its Special Cases

Extract Log-Likelihood and Akaike's Information Criterion

Description

The functions logLik and AIC extract the Log-Likelihood and the Akaike's Information Criterion from fitted generalized hyperbolic distribution objects. The Akaike information criterion is calculated according to the formula -2 \cdot \mbox{log-likelihood} + k \cdot n_{par}, where n_{par} represents the number of parameters in the fitted model, and k = 2 for the usual AIC.

Usage

## S4 method for signature 'mle.ghyp'
logLik(object, ...)

## S4 method for signature 'mle.ghyp'
AIC(object, ..., k = 2)

Arguments

object	An object of class mle.ghyp.
k	The “penalty” per parameter to be used; the default k = 2 is the classical AIC.
...	An arbitrary number of objects of class mle.ghyp.

Value

Either the Log-Likelihood or the Akaike's Information Criterion.

Note

The Log-Likelihood as well as the Akaike's Information Criterion can be obtained from the function ghyp.fit.info. However, the benefit of logLik and AIC is that these functions allow a call with an arbitrary number of objects and are better known because they are generic.

Author(s)

David Luethi

See Also

fit.ghypuv, fit.ghypmv, lik.ratio.test, ghyp.fit.info, mle.ghyp-class

Examples

  data(smi.stocks)

  ## Multivariate fit
  fit.mv <- fit.hypmv(smi.stocks, nit = 10)
  AIC(fit.mv)
  logLik(fit.mv)

  ## Univariate fit
  fit.uv <- fit.tuv(smi.stocks[, "CS"], control = list(maxit = 10))
  AIC(fit.uv)
  logLik(fit.uv)

  # Both together
  AIC(fit.uv, fit.mv)
  logLik(fit.uv, fit.mv)

ghyp documentation built on Aug. 21, 2023, 5:12 p.m.