Bmggm: Bayesian multiple Gaussian graphical models by MCMC.

Description Usage Arguments Value

View source: R/mcmc_sampler.R

Description

Bayesian multiple Gaussian graphical models by MCMC.

Usage

1
Bmggm(dat, options, PriorPar, InitVal)

Arguments

dat

a list of objets: n: number of observations. p: dimension of each pathway. K: number of pathways. z_P: indicator vector of genes membership P: dimension of the data.

options

a list of objets: burnin: number of MCMC iterations before burnin. nmc: number of MCMC iterations after burnin.

PriorPar

a list of objets: a: shape1 parameter for Theta for off-digonal block. b: shape2 parameter for Theta for off-digonal block. a0: shape1 parameter for Theta for digonal block. b0: shape2 parameter for Theta for digonal block. eps: rate parameter for v0^2. delta: shape parameter for v0^2. c: the parameter for decision boundary of spike-and-slab. Theta: a K x K initial graph PPI matrix.

InitVal

a list of objets: mu: intercept term. sigma2: overall noise level, same across groups. Beta: a P x P initial coefficient matrix. adj: a P x P initial adjacency matrix.

Value

a list of objets: Beta_save: p x p x K x nmc sample of coefficient matrix adj_save: p x p x K x nmc sample of adjacency matrix Theta_save: K x K x nmc sample of graph similarity matrix


jlin-vt/BMGGM documentation built on Dec. 30, 2019, 11:43 p.m.