<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE erlref SYSTEM "erlref.dtd">
<erlref>
<header>
<copyright>
<year>2000</year><year>2016</year>
<holder>Ericsson AB. All Rights Reserved.</holder>
</copyright>
<legalnotice>
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
</legalnotice>
<title>digraph_utils</title>
<prepared>Hans Bolinder</prepared>
<responsible>nobody</responsible>
<docno></docno>
<approved>nobody</approved>
<checked>no</checked>
<date>2001-08-27</date>
<rev>PA1</rev>
<file>digraph_utils.sgml</file>
</header>
<module>digraph_utils</module>
<modulesummary>Algorithms for Directed Graphs</modulesummary>
<description>
<p>The <c>digraph_utils</c> module implements some algorithms
based on depth-first traversal of directed graphs. See the
<c>digraph</c> module for basic functions on directed graphs.
</p>
<p>A <marker id="digraph"></marker><em>directed graph</em> (or
just "digraph") is a pair (V, E) of a finite set V of
<marker id="vertex"></marker><em>vertices</em> and a finite set E
of <marker id="edge"></marker><em>directed edges</em> (or just
"edges"). The set of edges E is a subset of V × V
(the Cartesian product of V with itself).
</p>
<p>Digraphs can be annotated with additional information. Such
information may be attached to the vertices and to the edges of
the digraph. A digraph which has been annotated is called a
<em>labeled digraph</em>, and the information attached to a
vertex or an edge is called a <marker id="label"></marker>
<em>label</em>.</p>
<p>An edge e = (v, w) is said
to <marker id="emanate"></marker><em>emanate</em> from vertex v and
to be <marker id="incident"></marker><em>incident</em> on vertex w.
If there is an edge emanating from v and incident on w, then w is
said to be
an <marker id="out_neighbour"></marker><em>out-neighbour</em> of v,
and v is said to be
an <marker id="in_neighbour"></marker><em>in-neighbour</em> of w.
A <marker id="path"></marker><em>path</em> P from v[1] to v[k] in a
digraph (V, E) is a non-empty sequence
v[1], v[2], ..., v[k] of vertices in V such that
there is an edge (v[i],v[i+1]) in E for
1 <= i < k.
The <marker id="length"></marker><em>length</em> of the path P is k-1.
P is a <marker id="cycle"></marker><em>cycle</em> if the length of P
is not zero and v[1] = v[k].
A <marker id="loop"></marker><em>loop</em> is a cycle of length one.
An <marker id="acyclic_digraph"></marker><em>acyclic digraph</em> is
a digraph that has no cycles.
</p>
<p>A <marker id="depth_first_traversal"></marker> <em>depth-first
traversal</em> of a directed digraph can be viewed as a process
that visits all vertices of the digraph. Initially, all vertices
are marked as unvisited. The traversal starts with an
arbitrarily chosen vertex, which is marked as visited, and
follows an edge to an unmarked vertex, marking that vertex. The
search then proceeds from that vertex in the same fashion, until
there is no edge leading to an unvisited vertex. At that point
the process backtracks, and the traversal continues as long as
there are unexamined edges. If there remain unvisited vertices
when all edges from the first vertex have been examined, some
hitherto unvisited vertex is chosen, and the process is
repeated.
</p>
<p>A <marker id="partial_ordering"></marker><em>partial ordering</em> of
a set S is a transitive, antisymmetric and reflexive relation
between the objects of S. The problem
of <marker id="topsort"></marker><em>topological sorting</em> is to
find a total
ordering of S that is a superset of the partial ordering. A
digraph G = (V, E) is equivalent to a relation E
on V (we neglect the fact that the version of directed graphs
implemented in the <c>digraph</c> module allows multiple edges
between vertices). If the digraph has no cycles of length two or
more, then the reflexive and transitive closure of E is a
partial ordering.
</p>
<p>A <marker id="subgraph"></marker><em>subgraph</em> G' of G is a
digraph whose vertices and edges form subsets of the vertices
and edges of G. G' is <em>maximal</em> with respect to a
property P if all other subgraphs that include the vertices of
G' do not have the property P. A <marker
id="strong_components"></marker> <em>strongly connected
component</em> is a maximal subgraph such that there is a path
between each pair of vertices. A <marker
id="components"></marker><em>connected component</em> is a
maximal subgraph such that there is a path between each pair of
vertices, considering all edges undirected. An <marker
id="arborescence"></marker><em>arborescence</em> is an acyclic
digraph with a vertex V, the <marker
id="root"></marker><em>root</em>, such that there is a unique
path from V to every other vertex of G. A <marker
id="tree"></marker><em>tree</em> is an acyclic non-empty digraph
such that there is a unique path between every pair of vertices,
considering all edges undirected.</p>
</description>
<funcs>
<func>
<name name="arborescence_root" arity="1"/>
<fsummary>Check if a digraph is an arborescence.</fsummary>
<desc>
<p>Returns <c>{yes, <anno>Root</anno>}</c> if <c><anno>Root</anno></c> is
the <seealso marker="#root">root</seealso> of the arborescence
<c><anno>Digraph</anno></c>, <c>no</c> otherwise.
</p>
</desc>
</func>
<func>
<name name="components" arity="1"/>
<fsummary>Return the components of a digraph.</fsummary>
<desc>
<p>Returns a list
of <seealso marker="#components">connected components</seealso>.
Each component is represented by its
vertices. The order of the vertices and the order of the
components are arbitrary. Each vertex of the digraph
<c><anno>Digraph</anno></c> occurs in exactly one component.
</p>
</desc>
</func>
<func>
<name name="condensation" arity="1"/>
<fsummary>Return a condensed graph of a digraph.</fsummary>
<desc>
<p>Creates a digraph where the vertices are
the <seealso marker="#strong_components">strongly connected
components</seealso> of <c><anno>Digraph</anno></c> as returned by
<c>strong_components/1</c>. If X and Y are two different strongly
connected components, and there exist vertices x and y in X
and Y respectively such that there is an
edge <seealso marker="#emanate">emanating</seealso> from x
and <seealso marker="#incident">incident</seealso> on y, then
an edge emanating from X and incident on Y is created.
</p>
<p>The created digraph has the same type as <c><anno>Digraph</anno></c>.
All vertices and edges have the
default <seealso marker="#label">label</seealso> <c>[]</c>.
</p>
<p>Each and every <seealso marker="#cycle">cycle</seealso> is
included in some strongly connected component, which implies
that there always exists
a <seealso marker="#topsort">topological ordering</seealso> of the
created digraph.</p>
</desc>
</func>
<func>
<name name="cyclic_strong_components" arity="1"/>
<fsummary>Return the cyclic strong components of a digraph.</fsummary>
<desc>
<p>Returns a list of <seealso marker="#strong_components">strongly
connected components</seealso>.
Each strongly component is represented
by its vertices. The order of the vertices and the order of
the components are arbitrary. Only vertices that are
included in some <seealso marker="#cycle">cycle</seealso> in
<c><anno>Digraph</anno></c> are returned, otherwise the returned list is
equal to that returned by <c>strong_components/1</c>.
</p>
</desc>
</func>
<func>
<name name="is_acyclic" arity="1"/>
<fsummary>Check if a digraph is acyclic.</fsummary>
<desc>
<p>Returns <c>true</c> if and only if the digraph
<c><anno>Digraph</anno></c> is <seealso marker="#acyclic_digraph">acyclic</seealso>.</p>
</desc>
</func>
<func>
<name name="is_arborescence" arity="1"/>
<fsummary>Check if a digraph is an arborescence.</fsummary>
<desc>
<p>Returns <c>true</c> if and only if the digraph
<c><anno>Digraph</anno></c> is
an <seealso marker="#arborescence">arborescence</seealso>.</p>
</desc>
</func>
<func>
<name name="is_tree" arity="1"/>
<fsummary>Check if a digraph is a tree.</fsummary>
<desc>
<p>Returns <c>true</c> if and only if the digraph
<c><anno>Digraph</anno></c> is
a <seealso marker="#tree">tree</seealso>.</p>
</desc>
</func>
<func>
<name name="loop_vertices" arity="1"/>
<fsummary>Return the vertices of a digraph included in some loop.</fsummary>
<desc>
<p>Returns a list of all vertices of <c><anno>Digraph</anno></c> that are
included in some <seealso marker="#loop">loop</seealso>.</p>
</desc>
</func>
<func>
<name name="postorder" arity="1"/>
<fsummary>Return the vertices of a digraph in post-order.</fsummary>
<desc>
<p>Returns all vertices of the digraph <c><anno>Digraph</anno></c>. The
order is given by
a <seealso marker="#depth_first_traversal">depth-first
traversal</seealso> of the digraph, collecting visited
vertices in postorder. More precisely, the vertices visited
while searching from an arbitrarily chosen vertex are
collected in postorder, and all those collected vertices are
placed before the subsequently visited vertices.
</p>
</desc>
</func>
<func>
<name name="preorder" arity="1"/>
<fsummary>Return the vertices of a digraph in pre-order.</fsummary>
<desc>
<p>Returns all vertices of the digraph <c><anno>Digraph</anno></c>. The
order is given by
a <seealso marker="#depth_first_traversal">depth-first
traversal</seealso> of the digraph, collecting visited
vertices in pre-order.</p>
</desc>
</func>
<func>
<name name="reachable" arity="2"/>
<fsummary>Return the vertices reachable from some vertices of a digraph.</fsummary>
<desc>
<p>Returns an unsorted list of digraph vertices such that for
each vertex in the list, there is
a <seealso marker="#path">path</seealso> in <c><anno>Digraph</anno></c> from some
vertex of <c><anno>Vertices</anno></c> to the vertex. In particular,
since paths may have length zero, the vertices of
<c><anno>Vertices</anno></c> are included in the returned list.
</p>
</desc>
</func>
<func>
<name name="reachable_neighbours" arity="2"/>
<fsummary>Return the neighbours reachable from some vertices of a digraph.</fsummary>
<desc>
<p>Returns an unsorted list of digraph vertices such that for
each vertex in the list, there is
a <seealso marker="#path">path</seealso> in <c><anno>Digraph</anno></c> of length
one or more from some vertex of <c><anno>Vertices</anno></c> to the
vertex. As a consequence, only those vertices
of <c><anno>Vertices</anno></c> that are included in
some <seealso marker="#cycle">cycle</seealso> are returned.
</p>
</desc>
</func>
<func>
<name name="reaching" arity="2"/>
<fsummary>Return the vertices that reach some vertices of a digraph.</fsummary>
<desc>
<p>Returns an unsorted list of digraph vertices such that for
each vertex in the list, there is
a <seealso marker="#path">path</seealso> from the vertex to some
vertex of <c><anno>Vertices</anno></c>. In particular, since paths may have
length zero, the vertices of <c><anno>Vertices</anno></c> are included in
the returned list.
</p>
</desc>
</func>
<func>
<name name="reaching_neighbours" arity="2"/>
<fsummary>Return the neighbours that reach some vertices of a digraph.</fsummary>
<desc>
<p>Returns an unsorted list of digraph vertices such that for
each vertex in the list, there is
a <seealso marker="#path">path</seealso> of length one or more
from the vertex to some vertex of <c><anno>Vertices</anno></c>. As a consequence,
only those vertices of <c><anno>Vertices</anno></c> that are included in
some <seealso marker="#cycle">cycle</seealso> are returned.
</p>
</desc>
</func>
<func>
<name name="strong_components" arity="1"/>
<fsummary>Return the strong components of a digraph.</fsummary>
<desc>
<p>Returns a list of <seealso marker="#strong_components">strongly
connected components</seealso>.
Each strongly component is represented
by its vertices. The order of the vertices and the order of
the components are arbitrary. Each vertex of the digraph
<c><anno>Digraph</anno></c> occurs in exactly one strong component.
</p>
</desc>
</func>
<func>
<name name="subgraph" arity="2"/>
<name name="subgraph" arity="3"/>
<fsummary>Return a subgraph of a digraph.</fsummary>
<desc>
<p>Creates a maximal <seealso marker="#subgraph">subgraph</seealso> of <c>Digraph</c> having
as vertices those vertices of <c><anno>Digraph</anno></c> that are
mentioned in <c><anno>Vertices</anno></c>.
</p>
<p>If the value of the option <c>type</c> is <c>inherit</c>,
which is the default, then the type of <c><anno>Digraph</anno></c> is used
for the subgraph as well. Otherwise the option value of <c>type</c>
is used as argument to <c>digraph:new/1</c>.
</p>
<p>If the value of the option <c>keep_labels</c> is <c>true</c>,
which is the default, then
the <seealso marker="#label">labels</seealso> of vertices and edges
of <c><anno>Digraph</anno></c> are used for the subgraph as well. If the value
is <c>false</c>, then the default label, <c>[]</c>, is used
for the subgraph's vertices and edges.
</p>
<p><c>subgraph(<anno>Digraph</anno>, <anno>Vertices</anno>)</c> is equivalent to
<c>subgraph(<anno>Digraph</anno>, <anno>Vertices</anno>, [])</c>.
</p>
<p>There will be a <c>badarg</c> exception if any of the arguments
are invalid.
</p>
</desc>
</func>
<func>
<name name="topsort" arity="1"/>
<fsummary>Return a topological sorting of the vertices of a digraph.</fsummary>
<desc>
<p>Returns a <seealso marker="#topsort">topological
ordering</seealso> of the vertices of the digraph
<c><anno>Digraph</anno></c> if such an ordering exists, <c>false</c>
otherwise. For each vertex in the returned list, there are
no <seealso marker="#out_neighbour">out-neighbours</seealso>
that occur earlier in the list.</p>
</desc>
</func>
</funcs>
<section>
<title>See Also</title>
<p><seealso marker="digraph">digraph(3)</seealso></p>
</section>
</erlref>