gera.bn.structure {bnpa}

Learn the Bayesian Network structure from data and build a PA model

Description

This function receives a data set, a list of parameters to learn the BN structure based on this data set. Then with the BN ready it will build a PA model if required. The process will then save the graphs of BN and PA and PA parameters.

Usage

gera.bn.structure(data.to.work, white.list = "", black.list = "",
  nreplicates = 1000, cb.algorithms = c("gs", "iamb", "fast.iamb",
  "inter.iamb", "mmpc", "si.hiton.pc"), sb.algorithms = c("hc", "tabu"),
  cb.tests = "", sb.tests = "", optimized.option = "FALSE",
  outcome.var, build.pa)

Arguments

data.to.work

is a data from which the BN structure will be learned.

white.list

is a list of mandatory connections of BN structure to be created.

black.list

is a list of forbiden connections of BN structure to be created.

nreplicates

is how many times the boostrap will run.

cb.algorithms

the name of constrained-based algorithms.

sb.algorithms

the name of score-based algorithms.

cb.tests

the name of tests for constrained-based algorithms.

sb.tests

the name of network scores for score-based algorithms.

optimized.option

a paremeter of bnlearn package to optmize the BN learn structre learning.

outcome.var

is the outcome (dependent) variable.

build.pa

indicates if the process will bulld a PA model or not.

Author(s)

Elias Carvalho

References

Scutari M (2017). Bayesian Network Constraint-Based Structure Learning Algorithms: Parallel and Optimized Implementations in the bnlearn R Package. Journal of Statistical Software, 77(2), 1-20.

Examples

## Not run: 
# Clean environment
closeAllConnections()
rm(list=ls())
# Set environment
# setwd("To your working directory")
# Load packages
library(bnpa)
# Load Data
data(dataQualiN)
# Set variables to work
nreplicates = 1000
white.list <- NULL
black.list <- "L-T"
cb.algorithms = c("gs")
sb.algorithms = c("hc")
cb.tests = "jt"
sb.tests = "aic"
optimized.option="FALSE"
outcome.var = "E"
build.pa = 0
# Learn the BN from data and save results (data & images)
gera.bn.structure(dataQualiN, white.list, black.list, nreplicates, cb.algorithms,sb.algorithms,
                 cb.tests, sb.tests, optimized.option, outcome.var, build.pa)

## End(Not run)