hillClimb.bnet: Hill-Climbing structure search

In gherardovarando/Rbnet: Define and Manage Bayesian Networks

Description

Function that perform a stocastic hill-climbing search in the space of possible Bayesian networks, danities and conditional densities are estimated with a bmop penalized-logLik search, package Rbmop

Usage

hillClimb.bnet(data, bnet = NULL, whitelist = NULL, blacklist = NULL,
  score = BIC, maxp = dim(data)[2], maxrep = 1000, ...)

Arguments

data	data frame of observation
bnet	optional, a bnet object to start the search from
whitelist	optional, a matrix or data frame with two columns each row indicates an arc
blacklist	optional, not implemented
score	a score function to be minimized in the search, usually BIC orAIC
maxp	positive integer, the maximum number of parents for each node
maxrep	positive integer, the maximum number of iterations or jump in the search space
...	additional parameters to be passed to bmop fitting functions

Value

A bnet object, the bnet among the ones visited that minimized the score.

gherardovarando/Rbnet documentation built on May 17, 2019, 4:18 a.m.