PnodeLabels: Lists or changes the labels associated with a parameterize...

In ralmond/Peanut: Parameterized Bayesian Networks, Abstract Classes

Lists or changes the labels associated with a parameterize node.

Description

A label is a character identifier associated with a node which provides information about its role in the models. This function returns or sets the labels associated with a node.

Usage

PnodeLabels(node)
PnodeLabels(node) <- value

Arguments

node	A Pnode object.
value	A character vector containing the names of the labels that node should be associated with. These names should follow the variable naming rules.

Details

Netica node sets are a collection of string labels that can be associated with various nodes in a network. These have proved to be very useful on writing code as often it is useful to perform some operation on only a certain kind of nodes. One purpose of node sets is to label a set of nodes that play a similar role in the model. For example, "ReportingVariable" or "Observable".

The PnodeLabels function is an attempt to generalize that mechanism. The expression PnodeLabels(node) returns the labels currently associated with node, thus provides a general mechanism for identifying the roles that a node might play.

The expression PnodeLabels(node)<-value removes any labels previously associated with node and adds the new labels named in value. The elements of value need not correspond to existing labels, new node sets will be created for new values. (Warning: this implies that if the name of the node set is spelled incorrectly in one of the calls, this will create a new node set. For example, "Observable" and "Observables" would be two distinct labels.)

Two labels have special meaning in the Peanut package. The function BuildAllTables(net) rebuilds the tables for nodes which are labeled “pnode” (i.e., parameterized nodes). The function GEMfit attempts to fit the parameters for nodes labeled “pnodes”, and associates values in the cases argument with the nodes labeled “onodes”.

Value

A character vector giving the names of the labels node is associated with. The setter form returns node.

Author(s)

Russell Almond

See Also

Pnode, BuildAllTables, GEMfit, PnetPnodes

Examples

## Not run: 
library(PNetica)##Requires PNetica
sess <- NeticaSession()
startSession(sess)

nsnet <- CreateNetwork("NodeSetExample", session=sess)

Ability <- NewDiscreteNode(nsnet,"Ability",c("High","Med","Low"))

EssayScore <- NewDiscreteNode(nsnet,"EssayScore",paste("level",5:0,sep="_"))

stopifnot(
  length(PnodeLabels(Ability)) == 0L ## Nothing set yet
)

PnodeLabels(Ability) <- "ReportingVariable"
stopifnot(
  PnodeLabels(Ability) == "ReportingVariable"
)
PnodeLabels(EssayScore) <- "Observable"
stopifnot(
  PnodeLabels(EssayScore) == "Observable"
)
## Make EssayScore a reporting variable, too
PnodeLabels(EssayScore) <- c("ReportingVariable",PnodeLabels(EssayScore))
stopifnot(
  setequal(PnodeLabels(EssayScore),c("Observable","ReportingVariable"))
)

## Clear out the node set
PnodeLabels(Ability) <- character()
stopifnot(
  length(PnodeLabels(Ability)) == 0L
)

DeleteNetwork(nsnet)
stopSession(sess)

## End(Not run)

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