findedge

Locate edge in graph

Syntax

[sOut,tOut] = findedge(G)

[sOut,tOut] = findedge(G,idx)

idxOut = findedge(G,s,t)

[idxOut,m] = findedge(G,s,t)

Description

[sOut,tOut] = findedge(G) returns the source and target node IDs, sOut and tOut, for all of the edges in graph G.

[sOut,tOut] = findedge(G,idx) finds the source and target nodes of the edges specified by idx.

idxOut = findedge(G,s,t) returns the numeric edge indices, idxOut, for the edges specified by the source and target node pairs s and t. The edge indices correspond to the rows G.Edges.Edge(idxOut,:) in the G.Edges table of the graph. If there are multiple edges between s and t, then all their indices are returned. An edge index of 0 indicates an edge that is not in the graph.

[idxOut,m] = findedge(G,s,t) additionally returns a vector m indicating which node pair (s,t) is associated with each edge index in idxOut. This is useful when there are multiple edges between the same two nodes.

Examples

Locate Edges with Specified End Nodes

Create a graph, and then determine the edge index for the (1,2) and (3,5) edges.

s = [1 1 2 2 2 3 3 3];
t = [2 3 3 4 5 6 7 5];
G = graph(s,t)

G = 
  graph with properties:

    Edges: [8x1 table]
    Nodes: [7x0 table]

idxOut = findedge(G,[1 3],[2 5])

idxOut = 2×1

     1
     6

idxOut contains the row index into G.Edges.EndNodes for each specified edge.

End Nodes of All Graph Edges

Create a graph, and then determine the end nodes of all edges in the graph.

s = {'a' 'a' 'b' 'b' 'c' 'c'};
t = {'b' 'c' 'd' 'e' 'f' 'g'};
G = graph(s,t);
G.Edges

ans=6×1 table
       EndNodes   
    ______________

    {'a'}    {'b'}
    {'a'}    {'c'}
    {'b'}    {'d'}
    {'b'}    {'e'}
    {'c'}    {'f'}
    {'c'}    {'g'}

[sOut,tOut] = findedge(G)

sOut = 6×1

     1
     1
     2
     2
     3
     3

tOut = 6×1

     2
     3
     4
     5
     6
     7

Locate Edges with Specified Edge Indices

Create a graph, and then determine the end nodes for the edges whose indices are 3 and 7.

s = [1 1 1 1 2 2 3 3 4 4];
t = [2 3 4 5 6 7 8 9 10 11];
G = digraph(s,t)

G = 
  digraph with properties:

    Edges: [10x1 table]
    Nodes: [11x0 table]

[sOut,tOut] = findedge(G,[3 7])

sOut = 2×1

     1
     3

tOut = 2×1

     4
     8

Determine Weight of Specified Edge

Create a graph.

s = [1 1 2 3];
t = [2 3 3 4];
weights = [10 20 30 40];
G = graph(s,t,weights)

G = 
  graph with properties:

    Edges: [4x2 table]
    Nodes: [4x0 table]

Find the weight of the (1,3) edge, using findedge to retrieve the index.

G.Edges.Weight(findedge(G,1,3))

ans = 
20

Change Weights of Multigraph Edges

Use findedge to change the weights of several multigraph edges.

Create and plot a multigraph. This graph has two edges between node 2 and node 4.

s = [1 1 2 3 2 2];
t = [2 3 3 4 4 4];
weights = [10 20 30 40 10 10];
G = graph(s,t,weights);
plot(G,'EdgeLabel',G.Edges.Weight)

Change the weights of the edges between nodes (3,2) and (2,4). Specify two outputs to findedge to get the end-node indices, m. This output is useful when there are multiple edges between two nodes, since idxOut can have more elements than the number of node pairs in s and t. The edge idxOut(1) = 3 connects the node pair (s(1),t(1)) = (3,2), and the edges idxOut(2) = 4 and idxOut(3) = 5 connect the edge (s(2),t(2)) = (2,4).

s = [3 2];
t = [2 4];
w = [1 4];
[idxOut, m] = findedge(G, s, t)

idxOut = 3×1

     3
     4
     5

m = 3×1

     1
     2
     2

G.Edges.Weight(idxOut) = w(m);
plot(G,'EdgeLabel',G.Edges.Weight)

Input Arguments

G — Input graph

graph object | digraph object

Input graph, specified as either a graph or digraph object. Use graph to create an undirected graph or digraph to create a directed graph.

Example: G = graph(1,2)

Example: G = digraph([1 2],[2 3])

s,t — Node pairs (as separate arguments)

node indices | node names

Node pairs, specified as separate arguments of node indices or node names. Similarly located elements in s and t specify the source and target nodes for edges in the graph.

This table shows the different ways to refer to one or more nodes either by their numeric node indices or by their node names.

Form	Single Node	Multiple Nodes
Node index	Scalar Example: 1	Vector Example: [1 2 3]
Node name	Character vector Example: 'A'	Cell array of character vectors Example: {'A' 'B' 'C'}
	String scalar Example: "A"	String array Example: ["A" "B" "C"]
	Categorical array Example: categorical("A")	Categorical array Example: categorical(["A" "B" "C"])

Example: G = findedge(G,[1 2],[3 4])

Example: G = findedge(G,{'a' 'a'; 'b' 'c'},{'b' 'c'; 'c' 'e'})

idx — Edge indices

scalar | vector

Edge indices, specified as a scalar or vector of positive integers. The edge index corresponds to a row in the G.Edges table of the graph, G.Edges(idx,:).

Output Arguments

idxOut — Edge indices

scalar | vector

Edge indices, returned as a scalar or vector of nonnegative integers. The edge indices correspond to rows in the G.Edges table of the graph, G.Edges(idxOut,:). An edge index of 0 indicates an edge that is not in the graph.

The length of idxOut corresponds to the number of node pairs in the input, unless the input graph is a multigraph.

m — End node indices

vector

End node indices, returned as a vector. The values in m connect the edge indices in idxOut to the input node pairs (s,t). The edge idxOut(j) connects the node pair with index m(j).

sOut,tOut — Node IDs

scalars | vectors

Node IDs, returned as separate scalars or vectors of positive integers. Similarly located elements in sOut and tOut specify the source and target nodes that form the edges G.Edges(idx,:).

Extended Capabilities

Thread-Based Environment
Run code in the background using MATLAB® backgroundPool or accelerate code with Parallel Computing Toolbox™ ThreadPool.

Version History

Introduced in R2015b