grash-operations {gRain} | R Documentation |
Make operations on undirected and directed acyclic graphs.
moralize(dag) mcs(ug,amat=as.adjmat(ug), vn=colnames(amat), root=NULL, index=FALSE) triangulate(ug,method="standard", nLevels, matrix=FALSE) ripOrder(ug,root=NULL,nLevels=NULL) jTree(ug, method="standard", vn=nodes(ug),nLevels, control=list())
dag |
A directed graph (of class "dagsh") |
ug |
An undirected graph (of class "ugsh") |
method |
The triangulation method, "standard" is a C
implementation of a minimum clique weight heuristic,
"mcwh" is a Fortran implementation of the same (kindly
provided by Peter Green) and "R" is a corresponding R
implementation. |
matrix |
Should the triangulated graph be returned as an adjacency matrix |
amat |
The graph represented as an adjancecy matrix |
vn |
The node names as they appear in the adjancecy matrix |
root |
Variables which must be in the first clique in the rip-ordering |
nLevels |
Number of levels of the nodes in the graph |
index |
If TRUE, then a permutation is returned |
control |
Currently not used |
... |
Currently not used. |
The jTree (for "junction tree") is just a wrapper for a call of
triangulate
followed by a call of ripOrder
.
moralize
and triangulate
returns objects of class
"graphsh".
mcs
returns a vector or NULL.
ripOrder
returns a list with cliques, separators etc.
Søren Højsgaard, sorenh@agrsci.dk
newugsh
, newug
,
newdagsh
, newdag
## Undirected graphs ## ug <- newug(~me+ve,~me+al,~ve+al,~al+an,~al+st,~an+st) edges(ug) nodes(ug) mcs(ug) ripOrder(ug) ## Directed graphs ## dag <- newdag(~me+ve,~me+al,~ve+al,~al+an,~al+st,~an+st) edges(dag) nodes(dag) moralize(dag)