bgms 0.2.0.0

This release rebuilds most of the package on top of 0.1.6.3. bgm() now fits Gaussian graphical models and mixed discrete-continuous networks alongside ordinal ones; priors are supplied as objects instead of loose scalars; inclusion inference is Rao-Blackwellized throughout; bgmCompare() has been reworked; and a set of checking tools (verdicts(), calibration_check(), prior_sensitivity_check(), network and edge plots) is new. Several defaults changed, some arguments were removed, and pairwise parameters are reported on a different scale, so a script written against 0.1.6.3 can run and give different numbers. Those changes are listed first.

Throughout this file and in the package’s output, “log” means the natural logarithm.

Breaking changes

Pairwise parameters are reported on the association scale.

Changed defaults.

Removed and reshaped arguments.

Changed output and semantics.

New features

New model classes.

Prior specification.

Inclusion inference.

Checking a fitted model.

Plots and centrality.

New extractors.

Prior-only sampling.

Edge priors on continuous and mixed models.

The hierarchical precision-graph prior (continuous and mixed models).

Other changes

Bug fixes

Deprecated

bgms 0.1.6.3

New features

Bug fixes

Other changes

Deprecated

bgms 0.1.6.2

New features

Other changes

Bug fixes

bgms 0.1.6.1

Other changes

Bug fixes

bgms 0.1.6.0

New features

Other changes

Bug fixes

Deprecated

bgms 0.1.5.0 (GitHub only)

New features

Other changes

Bug fixes

bgms 0.1.4.2

Bug fixes

Other changes

bgms 0.1.4.1

This is a minor release that adds some documentation and output bug fixes.

bgms 0.1.4

New features

Other changes

bgms 0.1.3

New features

User level changes

bgms 0.1.2

This is a minor release that adds some bug fixes.

bgms 0.1.1

This is a minor release adding some new features and fixing some minor bugs.

New features

User level changes