REMixed (version 1.1.2)

eBIC: eBIC

Description

Computes extended bayesian information criterion as $$ eBIC = -2\mathcal{LL}_{y}(\hat\theta,\hat\alpha)+P\log(N)+2\gamma\log(\binom(k,K))$$ where \(P\) is the total number of parameters estimated, \(N\) the number of subject, \(\mathcal{LL}_{y}(\hat\theta,\hat\alpha)\) the log-likelihood of the model, \(K\) the number of submodel to explore (here the numbre of biomarkers tested) and \(k\) the numbre of biomarkers selected in the model.

Usage

eBIC(object, ...)

Value

eBIC.

Arguments

object

output of remix or cv.remix.

...

opptional additional arguments.

References

Chen, J. and Z. Chen. 2008. Extended Bayesian information criteria for model selection with large model spaces. Biometrika 95 (3): 759-771.

Examples

if (FALSE) {
project <- getMLXdir()

ObsModel.transfo = list(S=list(AB=log10),
                        linkS="yAB",
                        R=rep(list(S=function(x){x}),5),
                        linkR = paste0("yG",1:5))

alpha=list(alpha0=NULL,
           alpha1=setNames(paste0("alpha_1",1:5),paste0("yG",1:5)))

y = c(S=5,AB=1000)
lambda = 1440

res = remix(project = project,
            dynFUN = dynFUN_demo,
            y = y,
            ObsModel.transfo = ObsModel.transfo,
            alpha = alpha,
            selfInit = TRUE,
            eps1=10**(-2),
            eps2=1,
            lambda=lambda)

eBIC(res)
}