PnodeMin-class: A minimal Pnode class for use in interface construction.

In ralmond/Peanut: Parameterized Bayesian Networks, Abstract Classes

A minimal Pnode class for use in interface construction.

Description

This is an experimental class which contains all of the meta-data of a Pnode, but isn't linked to backing Bayes net software. Its intended use is for assistance in building elicitation user interfaces.

Usage

PnodeMin(name, states = c("Yes", "No"), parents = list(), title = title(), 
         description = character(), labels = character(), continuous = FALSE, 
         tvals = effectiveThetas(length(states)), rules = "Compensatory", 
         link = "partialCredit", alphas = 1, betas = 1, linkScale = NA, 
         QQ = 1, PriorWeight = NA, CPT = NULL)

Arguments

name	Name of the node.
states	Node state names.
parents	List of other nodes to use a parents.
title	Title of the node.
description	Node description.
labels	Node labels.
continuous	Logical flag for continuous nodes.
tvals	Values assigned to node levels.
rules	Combination rules.
link	Link function.
alphas	Slope parameters.
betas	Intercept parameters.
linkScale	Scale parameter
QQ	Inner Q-matrix
PriorWeight	Prior weights
CPT	Conditional probability table.

Details

This class is bascially just a list containing the necessary fields. See the various accessor functions for the interpretation of the fields for more complete defintion.

Value

An Pnode object.

Note

This is an experimental class. It may be removed in future releases.

Author(s)

Russell Almond

See Also

Pnode.

Examples

anode <- PnodeMin("anode")

ralmond/Peanut documentation built on Sept. 19, 2023, 8:27 a.m.