NEWS.md

In gmgm: Gaussian Mixture Graphical Model Learning and Inference

gmgm 1.1.2

• Formatted the equations in the documentation to avoid HTML5 validation problems.

gmgm 1.1.1

• Fixed formatting and typing errors in the documentation.
• Added slightly more details to the main documentation page.

gmgm 1.1.0

• Created function relevant, which extracts the minimal sub-Gaussian mixture graphical model required to infer a subset of nodes.
• Changed the default values of several arguments to increase the speed of learning functions and make them easier to use:
• regul = NULL → regul = 0.01 (functions AIC, BIC, logLik, em)
• max_iter_em = Inf → max_iter_em = 100 (function em)
• max_iter_smem = Inf → max_iter_smem = 10 (function smem)
• max_iter_step = Inf → max_iter_step = 10 (function stepwise)
• max_iter_pem = Inf → max_iter_pem = 5 (functions param_em, struct_em)
• max_iter_sem = Inf → max_iter_sem = 5 (function struct_em)

gmgm 1.0.2

• Replaced dplyr's scoped verbs (_all, _at, _if), which have been superseded by the use of function across.
• Modified function gmm to coerce the elements of the gmm object to type double.
• Provided a more detailed example in the README file.

gmgm 1.0.1

• Removed a unit test of function em that failed with some BLAS/LAPACK implementations.

gmgm 1.0.0

• Initial release.

gmgm documentation built on Sept. 9, 2022, 1:07 a.m.