baycn: Bayesian Inference for Causal Networks
An approximate Bayesian method for inferring Directed Acyclic Graphs
(DAGs) for continuous, discrete, and mixed data. The algorithm can use the
graph inferred by another more efficient graph inference method as input;
the input graph may contain false edges or undirected edges but can help
reduce the search space to a more manageable size. A Metropolis-Hastings-like
sampling algorithm is then used to infer the posterior probabilities of
edge direction and edge absence.
References:
Martin, Patchigolla and Fu (2026) <doi:10.48550/arXiv.1909.10678>.
Documentation:
Downloads:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=baycn
to link to this page.