FE: Calculate free energy for the approximate posterior

In mattelisi/bmsR: Bayesian model selection for group studies in R

Description

This function derives the free energy for the current approximate posterior. This routine has been translated from the VBA_groupBMC function of the VBA toolbox http://code.google.com/p/mbb-vb-toolbox/ which was written by Lionel Rigoux and J. Daunizeau. See equation A.20 in Rigoux et al. (should be F1 on LHS)

Usage

FE(m, a, g)

Arguments

m	N by K (N subjects by K models) matrix of log-model evidence.
g	N by K (N subjects by K models) matrix of normalized posterior beliefs.
alpha	Dirichlet parameters (usually the posterior estimate).

Value

Free energy (evidence in favor of different model frequencies).

mattelisi/bmsR documentation built on March 9, 2020, 2:55 p.m.