networkBMA (version 1.14.0)

gControl: Control parameters for using Zellner's g-prior in ScanBMA

Description

Assigns default control parameters for the use of Zellner's g-prior in ScanBMA, and allows setting control parameter values.

Usage

gControl( optimize = TRUE, optMethod = "perTarget", g0 = NULL, iterlim = 100, epsilon = 0.1 )

Arguments

optimize

A logical value indicating whether to optimze g using an iterative EM algorithm or use a fixed value of g.

optMethod

A character string indicating how to optimize g. Currently, only perTarget is supported, indicating that g should be optimized individually for each target.

g0

An initial value of g to use if optimize is TRUE, or the fixed value to use without optimization.

iterlim

If optimize is TRUE, the maximum number of iterations of the EM algorithm to use. Ignored otherwise.

epsilon

If optimize is TRUE, the precision with which to find g using the EM algorithm. Ignored otherwise.

Value

A list of values for the named control parameters to be passed to ScanBMAcontrol and ScanBMA.

References

A. Zellner (1986), On assessing prior distributions and Bayesian regression analysis with g-prior distributions, Bayesian inference and decision techniques: Essays in Honor of Bruno De Finetti, 6:233-243. M. Clyde and E.I. George (2004), Model Uncertainty, Statistical Science, 81-94.

See Also

ScanBMAcontrol, ScanBMA, networkBMA

Examples

data(dream4)

network <- 1

nTimePoints <- length(unique(dream4ts10[[network]]$time))

edges1ts10 <- networkBMA( data = dream4ts10[[network]][,-(1:2)], 
                          nTimePoints = nTimePoints,
                          control = ScanBMAcontrol(gCtrl =
                          gControl(optimize = TRUE)) )