causal.discovery: Causal discovery algorithms

In bnlearn: Bayesian Network Structure Learning, Parameter Learning and Inference

Causal discovery algorithms

Description

Learn the structure of a (causal) Bayesian network with DirectLiNGAM.

Usage

direct.lingam(x, cluster, whitelist, blacklist, mi, maximize = "alasso",
  maximize.args = list(), debug = FALSE)

Arguments

x	a data frame containing the variables in the model.
cluster	an optional cluster object from package parallel.
whitelist	a data frame with two columns (optionally labelled "from" and "to"), containing a set of arcs to be included in the graph.
blacklist	a data frame with two columns (optionally labelled "from" and "to"), containing a set of arcs not to be included in the graph.
mi	a character string, the label of the mutual information measure used to identify the topological ordering of the nodes. It can take values "pwling" (the default) or "gkernel".
maximize	a character value, the label of the method used to identify the parents of each node given the causal ordering. The only possible value is "alasso" (adaptive LASSO).
maximize.args	a list of arguments to be passed to the method specified by maximize. See below for details.
debug	a boolean value. If TRUE, a lot of debugging output is printed. Otherwise, the function is completely silent.

Details

The "alasso" method has the following optional arguments:

• gamma: the coefficient of the weights used in the adaptive LASSO. The default value is 1.

• lambda.min.ratio and pmax: identical to the arguments of the same names in glmnet(), with the same default values and interpretation.

• k: the coefficient of the penalised likelihood score used to choose the optimal shrinkage in both ridge and LASSO models in the adaptive LASSO. It defaults to the BIC penalty.

Value

An object of class bn. See bn-class for details.

Note

See structure learning for a complete list of structure learning algorithms with the respective references.

Author(s)

Marco Scutari

See Also

local discovery algorithms, score-based algorithms, constraint-based algorithms, hybrid algorithms.

bnlearn documentation built on July 13, 2026, 5:08 p.m.