logo

Tree Graph 📂Graph Theory

Tree Graph

Definition 1

A tree is a connected graph that does not contain any cycles.

Description

The concept of trees is commonly seen in areas such as data structures in computer science. For those in STEM fields who have even a cursory involvement with computers, they may have heard of heap sorting. The heap mentioned here is indeed a type of tree. The union of trees is intuitively called a Forest. In the case of a directed graph, a node with an input degree of 00 is called the Root, and a node with an output degree of 00 is called the Leaf. Trees may seem like a familiar and easily applicable concept, however, in pure mathematics, they reveal their complex true selves.

Theorem

The following are equivalent conditions for a graph to be a tree.

Let TT be a graph with nn vertices. The following statements are equivalent:

  • (i): TT is a tree.
  • (ii): TT does not have cycles and has n1n-1 edges.
  • (iii): TT is a connected graph and has n1n-1 edges.
  • (iv): Every edge in TT is a bridge.
  • (v): There is only one path connecting any two vertices in TT.
  • (vi): TT does not have cycles but adding an edge would result in exactly one cycle forming.

  • An edge bGb \in G is called a bridge if its removal would disconnect the graph.

Proof

(i)    (ii)(i) \implies (ii)

Since TT does not contain any cycles, removing any edge results in two trees. Inductively repeating this process shows that each tree will have exactly one fewer edge than the number of vertices. Conversely, the tree TT must have n1n-1 edges.


(ii)    (iii)(ii) \implies (iii)

Suppose TT is not connected. Each component of TT does not have cycles and has 11 fewer edges than vertices. However, this contradicts the premise that there are n1n-1 edges, thus TT must be connected.


(iii)    (iv)(iii) \implies (iv)

Edges of a disconnected graph: Let a simple graph GG have nn vertices. If GG has kk vertices, then the number of edges mm satisfies the following: nkm(nk)(nk+1)/2 n-k \le m \le (n-k)(n-k+1)/2

Removing any edge from TT leaves the number of vertices as nn and the number of edges as n2n-2, thus nkn2n-k \le n-2, indicating that the number of components increases with the removal of any edge, meaning all edges in TT are bridges.


(iv)    (v)(iv) \implies (v)

As TT is a connected graph, any pair of vertices belong to at least one path. If a pair of vertices belongs to two paths, this would form a cycle, contradicting the premise that all edges are bridges.


(v)    (vi)(v) \implies (vi)

If TT already included a cycle, then any two vertices on that cycle would belong to at least two paths, contradicting the premise. If an edge ee is added to TT, the vertices connected by ee are already connected, thus forming a cycle. Therefore, adding an edge to TT results in exactly one cycle forming.


(vi)    (i)(vi) \implies (i)

If TT is not a connected graph, adding edges to connect each component would not produce a cycle. However, this contradicts the premise, so TT must be connected and, given it does not contain cycles, TT is a tree.


  1. Wilson. (1970). Introduction to Graph Theory: p43. ↩︎