PnetMakeStubNodes: Creates (or removes) references to nodes in a network
In ralmond/Peanut: Parameterized Bayesian Networks, Abstract Classes

PnetMakeStubNodes

R Documentation

Creates (or removes) references to nodes in a network

Description

A stub node is a reference in a spoke network to a node in a hub network. The function PnetMakeStubNodes makes stub nodes in the spoke network. The function RemoveStubNodes removes them.

Usage

PnetMakeStubNodes(net, nodes)
PnetRemoveStubNodes(net, nodes)

Arguments

`net`	A `Pnet` object in which the stub nodes will be created or removed. This is generally a spoke (evidence model) network.
`nodes`	A list of `Pnode` objects. In the case of `PnetMakeStubNodes` these are the nodes in the hub model which are to be copied. In the case of `PnetRemoveStubNodes` these are the stub nodes to be removed.

Details

In the hub-and-spoke model, spoke models (evidence models) reference nodes in the central hub model (proficiency model in educational applications). The stub node is a node (or pseudo-node) in the spoke model which is actually a reference to a node in the hub model. In the operation PnetAdjoin when the spoke model is combined with the hub model, the stubs are replaced with the actual nodes they represent.

The pair of functions PnetMokeStubNodes and PnetRemoveStubNodes are used inside of Qmat2Pnet to create the necessary references to the proficiency nodes (in the columns of the Q-matrix) while building the conditional probability tables for the observable nodes (the rows of the Q-matrix). The function PnetMakeStubNodes gets called before the conditional probability tables are built, and the function PnetRemoveStubNodes gets called after all conditional probability tables are built.

Value

The function PnetMakeStubNodes returns a list of the newly created stub nodes.

The return of the function PnetRemoveStubNodes is implementation dependent, and is called mainly for its side effects.

Both functions destructively modify the net argument.

Note

The behavior of these functions will depend a lot on the underlying implementation, and they should be thought of as a pair. The function PnetMakeStubNodes gets called before constructing the conditional probability tables, and PnetRemoveStubNodes. For example, this could be used to give the nodes the official hub node name while constructing the conditional probability tables and then rename them to something else.

In the PNetica-package implementation, the function PnetMakeStubNodes copies the nodes from the hub to the spoke, and the function PnetRemoveStubNodes deletes them (which if they are attached as a parent, automatically creates a stub node in Netica).

Author(s)

Russell Almond

References

Almond, R. G. (presented 2017, August). Tabular views of Bayesian networks. In John-Mark Agosta and Tomas Singlair (Chair), Bayeisan Modeling Application Workshop 2017. Symposium conducted at the meeting of Association for Uncertainty in Artificial Intelligence, Sydney, Australia. (International) Retrieved from http://bmaw2017.azurewebsites.net/

Examples

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

PM <- ReadNetworks(system.file("testnets", "miniPP-CM.dne",
                               package="PNetica"), session=sess)
EM1 <- ReadNetworks(system.file("testnets", "PPcompEM.dne",
                                package="PNetica"), session=sess)

## Find the target node and its parents.
obs <- PnetFindNode(EM1,"CompensatoryObs")
pars <- PnetFindNode(PM,c("NTL","POfMom"))

## Make stub nodes for the parents
stubs <- PnetMakeStubNodes(EM1,pars)
## Set them as the parents
PnodeParents(obs) <- stubs

## Build the CPT
PnodeLink(obs) <- "partialCredit"
PnodeRules(obs) <- "Compensatory"
PnodeAlphas(obs) <- c(NTL=0.9,POfMom=1.1)
PnodeBetas(obs) <- 0.3
PnodeQ(obs) <- TRUE
BuildTable(obs)

##Done, now remove the stubs
PnetRemoveStubNodes(EM1,stubs)

DeleteNetwork(list(PM,EM1))
stopSession(sess)

## End(Not run)

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