pointLik: Point likelihood values

In config-i1/greybox: Toolbox for Model Building and Forecasting

Point likelihood values

Description

This function returns a vector of logarithms of likelihoods for each observation

Usage

pointLik(object, log = TRUE, ...)

Arguments

object	Time series model.
log	Whether to take logarithm of likelihoods.
...	Some stuff.

Details

Instead of taking the expected log-likelihood for the whole series, this function calculates the individual value for each separate observation. Note that these values are biased, so you would possibly need to take number of degrees of freedom into account in order to have an unbiased estimator.

This value is based on the general likelihood (not its concentrated version), so the sum of these values may slightly differ from the output of logLik.

Value

This function returns a vector.

Author(s)

Ivan Svetunkov, [e-mail]

See Also

AIC, BIC

Examples

xreg <- cbind(rnorm(100,10,3),rnorm(100,50,5))
xreg <- cbind(100+0.5*xreg[,1]-0.75*xreg[,2]+rnorm(100,0,3),xreg,rnorm(100,300,10))
colnames(xreg) <- c("y","x1","x2","Noise")
ourModel <- alm(y~x1+x2,as.data.frame(xreg))

pointLik(ourModel)

# Bias correction
pointLik(ourModel) - nparam(ourModel)

# Bias correction in AIC style
2*(nparam(ourModel)/nobs(ourModel) - pointLik(ourModel))

# BIC calculation based on pointLik
log(nobs(ourModel))*nparam(ourModel) - 2*sum(pointLik(ourModel))

config-i1/greybox documentation built on Dec. 12, 2024, 1:42 p.m.