Easy proof about spanning trees (for you)

  • Thread starter Thread starter TheMathNoob
  • Start date Start date
  • Tags Tags
    Proof Trees
AI Thread Summary
The discussion revolves around proving properties of spanning trees in undirected graphs. It establishes that adding an edge not in the spanning tree creates a cycle, demonstrating this through the construction of subgraph G1. Removing any edge from this cycle results in subgraph G2, which retains the spanning tree properties by maintaining connectivity and acyclic nature. The conversation also clarifies notation and definitions, ensuring consistency in variables used throughout the proofs. Overall, the thread emphasizes understanding the relationship between spanning trees and cycles in graph theory.
TheMathNoob
Messages
189
Reaction score
4

Homework Statement


Let T be any spanning tree of an undirected graph G. Suppose that uv is any edge in G that is not in T. The following proofs are easy by using the definitions of undirected tree, spanning tree and cycle

a)Let G1 be the subgraph that results from adding uv to T. Show that G1 has a cycle involving uv, say (w1,w2,...wp,w1, where p>=3, u =w1 and v=wp

b) Suppose anyone edge wi,wi+1, is removed from the cycle created in part(a), creating a subgraph G2(which depends on i). Show that G2 is a spanning tree for G

Homework Equations


I think that this is relevant in this proof, any free tree with N vertices has N-1 edges. "The book did not give any hint".

The Attempt at a Solution


a)Let n be the number of vertices of an undirected graph. Therefore, its minimum spanning tree will have n-1 edges. A spanning tree in this case is also a connected acyclic graph which has the same property. If we add 1 more edge from G that is not in the spanning tree, we get n-1+1=n. This violates the property of CAG which implies that the graph has a cycle.

b)by part a, we have n vertices and n different edges, so if we take away 1, we will have n-1 different edges which implies that the graph is acyclic.
 
Physics news on Phys.org
Let G' be G with the edge wi,wi+1 removed.
Pick any vertex x in G. Divide the vertices of G into subsets A(x) and B(x) where vertices in A(x) can be reached from x via G' and vertices in B(x) cannot.
Pick a vertex y in B(x). Now all you need to do to prove (b) is to demonstrate a route from x to y in G2.
 
Last edited:
andrewkirk said:
Let G' be G with the edge wi,wi+1 removed.
Pick any vertex x in G. Divide the vertices of G into subsets A(x) and B(x) where vertices in A(x) can be reached from v via G' and vertices in B(x) cannot.
Pick a vertex y in B(x). Now all you need to do to prove (b) is to demonstrate a route from x to y in G2.
Hi andrewirk, thank you so much for helping me out. What is v?
 
TheMathNoob said:
What is v?
That should be x. I started with u and v, then changed them to x and y when I saw that you'd already used those letters for edges. It looks like I missed one when I was changing over.
 
andrewkirk said:
That should be x. I started with u and v, then changed them to x and y when I saw that you'd already used those letters for edges. It looks like I missed one when I was changing over.
wh
andrewkirk said:
Let G' be G with the edge wi,wi+1 removed.
Pick any vertex x in G. Divide the vertices of G into subsets A(x) and B(x) where vertices in A(x) can be reached from x via G' and vertices in B(x) cannot.
Pick a vertex y in B(x). Now all you need to do to prove (b) is to demonstrate a route from x to y in G2.
what is G2?
 
TheMathNoob said:
what is G2?
It's what you defined it as in (b) of the OP
 
Back
Top