network: Bayesian network data structure

Description

A Bayesian network is represented as an object of class network. Methods for printing and plotting are defined.

Usage

network(df,specifygraph=FALSE,inspectprob=FALSE,equalcases=FALSE,vif=1.0,
        doprob=TRUE,tvar=NA,smalldf=NA,yr=c(0,350),xr=yr) 
## S3 method for class 'network':
print(x,filename=NA,master=FALSE,condposterior=FALSE,
                          condprior=FALSE,...) 
## S3 method for class 'network':
plot (x,scale=10,arrowlength=.25,
                        notext=FALSE,
                        sscale=.7*scale,showban=TRUE,yr=c(0,350),xr=yr,
                        unitscale=20,cexscale=8,...)

prob.network (x,df,equalcases=FALSE,vif=1.0,smalldf=NA)

Arguments

A dataframe, where the columns define the variables. A continuous variable should have type numeric and discrete varibles should have type factor.

specifygraph

Interactively specify a directed acyclic graph and possibly a banlist (see below).

inspectprob

Inspect/change the calculated probability distribution using a graphical interface.

equalcases

When calculating the probability distribution for discrete variables, do not (if TRUE) set the probability equal to the observed frequency. If FALSE, observed frequencies are used.

vif

Variance inflation factor. For continuous nodes, the probability distribution is assumed normal with a mean depending on the parents and a variance for each configuration of the discrete parents. The mean and variance are deduced from data

doprob

If TRUE, do not calculate a probability distribution. Used for example in simulation.

tvar

Vector of indices of time-varying nodes, see timeslice.

smalldf

For use with timevarying networks, see timeslice.

The network.

filename

Print to file.

condprior

Print condprior? TRUE/FALSE

condposterior

Print condposterior? TRUE/FALSE

master

Print master? TRUE/FALSE.

scale

The axes in network plots are from -scale to scale.

sscale

The nodes are placed on a circle with radius sscale.

unitscale

Scale parameter for chopping off arrow heads.

cexscale

The size of the nodes.

arrowlength

The length of the arrow heads.

Range of x-axis.

Range of y-axis.

notext

No (or little) text is displayed on the plot if notext==TRUE.

rotate

The nodes are placed on a circle and this may be rotated by this number (in radians).

showban

If banned arrows should be shown (they are shown in red).

...

Passed to plot.node.

Value

A network object with attributes
nodesA list of nodes. If doprob==TRUE, the nodes are given the attribute prob which is the initial probability distribution used by jointprior. Arguments equalcases and vif are used to calculate prob.
nThe number of nodes
discreteA vector of indices of discrete nodes.
continuousA vector of indices of continuous nodes.
banlistA matrix with two columns. Each row contains the indices i -> j of arrows that may not be allowed in the directed acyclic graph.
scoreAdded by learn and is the log-network score.
relscoreAdded by nwfsort and is the relative log-network score -- compared to the best network in a networkfamily.

References

Further information about Deal can be found at: http://www.math.auc.dk/novo/deal.

Examples

Run this code

A <- factor(rep(c("A1","A2"),50))
B <- factor(rep(rep(c("B1","B2"),25),2))
thisnet <- network( data.frame(A,B) )

Run the code above in your browser using DataLab