The function MC3() searches for log-linear models with the highest posterior probability. The function gibbsSampler() is a blocked Gibbs sampler for sampling from the posterior distribution of the log-linear parameters. The functions findPostMean() and findPostCov() compute the posterior mean and covariance matrix for decomposable models which, for these models, is available in closed form.

Documentation

Manual: bayesloglin.pdf
Vignette: bayesloglin-R-package

Maintainer: Matthew Friedlander <friedla at yorku.ca>

Author(s): Matthew Friedlander

Install package and any missing dependencies by running this line in your R console:

install.packages("bayesloglin")

Depends igraph
Imports
Suggests
Enhances
Linking to
Reverse
depends
Reverse
imports
Reverse
suggests
Reverse
enhances
Reverse
linking to

Package bayesloglin
Materials
URL
Task Views
Version 1.0.1
Published 2016-12-27
License GPL (>= 2)
BugReports
SystemRequirements
NeedsCompilation yes
Citation
CRAN checks bayesloglin check results
Package source bayesloglin_1.0.1.tar.gz