NEWS.md

In beam: Fast Bayesian Inference in Large Gaussian Graphical Models

This file lists changes in the R package for each release.

beam 2.0.2

• changed contact email address
• seperate centering and scaling c++ functions to compute sample variances

beam 2.0.1

• Small bugs fixed in c++ code

beam 2.0.0

Major changes

• the main code of beam() is now written in C++

• new D argument for beam() to specify the shrinkage target of the linear shrinkage estimator

• new function lightbeam() that is a wrapper for beam() and beam.select(). It is faster and outputs a lighter object (sparse matrix).

• new methods postExpSigma and postExpOmega for beam-class objects that provide the posterior expectation of covariance and inverse-covariance matrices, respectively. These methods also allow the shrinkage estimation of the variances (or not).

Minor changes

• Dependency on Rcpp has been added

• date and description have been added in `DESCRIPTION

• a NEWS file has been added

• paper citations have been updated

beam documentation built on July 1, 2020, 10:23 p.m.