graph-create: Create undirected and directed graphs

In DataSciBurgoon/gRbase: A Package for Graphical Modelling in R

Description

These functions are wrappers for creation of graphs as implemented by graphNEL objects in the graph package.

Usage

ug(..., result = "graphNEL")

ugList(x, result = "graphNEL")

dag(..., result = "graphNEL", forceCheck = FALSE)

dagList(x, result = "graphNEL", forceCheck = FALSE)

Arguments

...	A generating class for a graph, see examples below
result	The format of the graph. The possible choices are "graphNEL" (for a 'graphNEL' object), "igraph" (for an 'igraph' object), "matrix" (for an adjacency matrix), "dgCMatrix" (for a sparse matrix).
x	A list or individual components from which a graph can be created.
forceCheck	Logical determining if it should be checked if the graph is acyclical. Yes, one can specify graphs with cycles using the dag() function.

Value

Functions ug(), and dag() can return a graphNEL object, an 'igraph' object, a sparse or a dense adjacency matrix.

Author(s)

Søren Højsgaard, [e-mail]

Examples

## The following specifications of undirected graphs are equivalent:
uG1 <- ug(~ a:b:c + c:d)
uG2 <- ug(c("a", "b", "c"), c("c", "d"))
uG3 <- ug(c("a", "b"), c("a", "c"), c("b", "c"), c("c", "d"))

graph::edges(uG1)
graph::nodes(uG1)

## The following specifications of directed acyclig graphs are equivalent:
daG1 <- dag(~ a:b:c + b:c + c:d)
daG2 <- dag(c("a", "b", "c"), c("b", "c"), c("c", "d"))

graph::edges(daG1)
graph::nodes(daG2)

## dag() allows to specify directed graphs with cycles:
daG4 <- dag(~ a:b + b:c + c:a) # A directed graph but with cycles

## A check for acyclicity can be done with
## daG5 <- dag(~ a:b + b:c + c:a, forceCheck=TRUE) 

## A check for acyclicity is provided by topoSort
topo_sort( daG2 )
topo_sort( daG4 )

## Different representations
uG6 <- ug(~a:b:c + c:d, result="graphNEL")  # default
uG7 <- ug(~a:b:c + c:d, result="igraph")    # igraph
uG8 <- ug(~a:b:c + c:d, result="matrix")    # dense matrix
uG9 <- ug(~a:b:c + c:d, result="dgCMatrix") # sparse matrix

DataSciBurgoon/gRbase documentation built on March 25, 2020, 12:03 a.m.