new8noralk: very simplified version of new8alk

In rebastaba: Handling Bayesian networks

Description

(ba) In this simplified version of new8alk, the user has no more than giving the name and the parents of the node and the regression coefficients. The distribution will be automatically put to 'normal' and the parameter 'mu' to the linear combinations of the parents.

Usage

new8noralk(des, parent=character(0), coef=0:length(parent), sigma=1)

Arguments

des	either the name or a description ('ds' object).
parent	(character(0)) Character vector with the names of the parents. character(0) means that there is no parents.
coef	coefficient to define mu. If there are two parents, then mu will be 'coef[1]+coef[2]*{{parent[1]}}+coef[3]*{{parent[2]}}'.
sigma	standard deviation.

Details

length(coef) must equals 1 + length(parent); sigma must be numeric.

Value

an object of class "bn" with an additional node

Examples

 rebastaba3k("RESET"); # needed only for R checking, to be forgotten
 print(new8noralk(char2des("some /alk/"), parent=c("A", "AA")));

rebastaba documentation built on May 2, 2019, 5:24 p.m.