boot.ges: Greedy Equivalence Search Algorithm (GES) With Bootstrapping

In rlebron-bioinfo/gnlearn: Genetic Network Learning

Description

This function allows you to learn a directed graph from a dataset using the Greedy Equivalence Search (GES) algorithm of Chickering (2002).

Usage

boot.ges(
  df,
  blacklist = NULL,
  adaptive = c("none", "vstructures", "triples"),
  maxDegree = integer(0),
  R = 200,
  m = NULL,
  threshold = 0.5,
  to = c("igraph", "adjacency", "edges", "graph", "bnlearn"),
  cluster = parallel::detectCores(),
  seed = sample(1:10^6, 1)
)

Arguments

df	Dataset.
blacklist	A data frame with two columns, containing a set of arcs not to be included in the graph (optional).
adaptive	Whether constraints should be adapted to newly detected v-structures or unshielded triples: 'none', 'vstructures', or 'triples'. Default: 'none'
maxDegree	Parameter used to limit the vertex degree of the estimated graph. Default: integer(0)
R	Number of bootstrap replicates (optional). Default: 200
m	Size of training set (optional). Default: nrow(df)/2
threshold	Minimum strength required for a coefficient to be included in the average adjacency matrix (optional). Default: 0.5
to	Output format ('adjacency', 'edges', 'graph', 'igraph', or 'bnlearn') (optional).
cluster	A cluster object from package parallel or the number of cores to be used (optional). Default: parallel::detectCores()
seed	Seed used for random selection. Default: NULL

Examples

obj <- boot.ges(df)
avg.g <- obj$average
g.rep <- obj$replicates

rlebron-bioinfo/gnlearn documentation built on July 25, 2020, 12:38 p.m.