A path in a graph is a subgraph of
that is a path graph (West 2000, p. 20). The length of a path is the number of edges it contains.
In most contexts, a path must contain at least one edge, though in some applications (e.g., defining the path covering number), "degenerate" paths of length 0 consisting of a single vertex are allowed (Boesch et al. 1974).
An -path is a path whose endpoints (vertices of degree 1) are the vertices with distinct indices
and
. (The symbols
and
are also commonly used.) A single
-path may be found in the Wolfram Language using FindPath[g, s, t], while FindPath[g, s, t, kspec, n] finds at most
paths of length kspec (where kspec may be Infinity and
may be All).
For a simple graph, a path is equivalent to a trail and is completely specified by an ordered sequence of vertices. For a simple graph , a Hamiltonian path is a path that includes all vertices of
(and whose endpoints are not adjacent).
The number of (undirected) -walks from vertex
to vertex
in a graph with adjacency matrix
is given by the
element of
(Festinger 1949). In order to compute the number
of graph paths, all closed
-walks that are not paths must be subtracted.
The first few matrices of -paths
can be given in closed form by
(1) | |||
(2) | |||
(3) |
(Luce and Perry 1949, Katz 1950, Ross and Harary 1952, Perepechko and Voropaev), where is the matrix formed from the diagonal elements of
and
denotes matrix element-wise multiplication.
Furthermore, the number of -cycles is related to
by
(4) |
where denotes the trace.
Giscard et al. (2016) gave the formula for the path matrix giving the number of -paths from
to
as
(5) |
where the sum is over connected induced subgraphs of
containing both
and
,
denotes the number of neighbors of
in
(namely vertices
of
which are not in
and such that there exists at least one edge from
to a vertex of
),
denotes the matrix trace, and
is the
th element of the
th matrix power of the adjacency matrix of
restricted to the connected induced subgraph
, namely
(6) |
with .