You are on page 1of 17

Research Project - I

Graph Theory [MA 2053]

Group 03

Ridwan (110532R) Sudharma (110557X)

Tharindu (110375L) Charith (110044X)

Kulakshi (110161D)

Table of Contents
I.

Related Definitions ......................................................................................................... 2


Graph Connected Graph Adjacent Loop Walk Closed Walk Open Walk Trail Incident[1] Degree[1] Euler Trail Euler Circuit

II. III.

Eulers Theorem .............................................................................................................. 5


Theorem

Mathematical Proof ......................................................................................................... 5


Necessary condition for Part A Necessary condition for part B Sufficient condition for part B Sufficient condition of Part A

IV. V. VI. VII.

Research Problem Description ...................................................................................... 11


Brief Introduction to History

Solution ...........................................................................................................................12 Eulers Solution ..............................................................................................................12 Problems .........................................................................................................................15

VIII. References ......................................................................................................................16

Related Definitions
The following mathematical definitions are included in our solution for the research problem. Therefor having a good understanding of these definitions will set the background knowledge for you to help understand the latter part of this document.
Graph

Graph is a representation of a set of objects where some pairs of objects are connected by links. The interconnected objects are known as Vertices and the links that connect some pairs of vertices are known as Edges. [2]
Connected Graph

A graph is said to be a connected graph if and only if there exists a path between any two vertices in the graph. [2]
A A B D B D

Graph Not Connected


Adjacent

Connected Graph

Two vertices are called adjacent, if they are attached directly by at least one edge. [3]
Loop

Loop is an edge that connects a vertex to itself. [1]


Loop

RESEARCH PROJECT - I - JULY 2013

Walk

A sequence of alternating (adjacent) vertices and edges, in which the order of edges used between adjacent vertices are specified. [2] Ex: a walk in a graph (with vertices v0 to vn where vi-1 and vi is connected with the edge ei), from v0 to vn is denoted by v0e1v1e2envn
A

Walk from B to C: B->A->D->A->C

Closed Walk

A walk is closed if its first and last vertices are the same. [2] v0e1v1e1envo
Open Walk

A walk is Open if its first and last vertices are different. [2] v0e1v1e1envn
Trail

Trail is a walk where no edge of the graph is repeated. [2]


A

Trail from B to C: B->A->D->C


RESEARCH PROJECT - I - JULY 2013 3

Incident

If the vertex v is an endpoint of the edge e, then we say that e is incident with v.[7]
Degree

The degree of a vertex v is the number of edges incident with a vertex (or simply the number of times an edge touches the vertex). We denote the degree of v by d(v). [7] In the following graph; d(A)=3 d(B)=1 d(C)=2 d(D)=4
Euler Trail
C B A

A trail is an Euler trail if it uses all the edges of the graph. [2]

Euler Circuit

A closed Euler trail is known as an Euler Circuit. [2]

RESEARCH PROJECT - I - JULY 2013

Eulers Theorem [5]


Our research problem was solved by the famous mathematician Leonard Euler in early 1980s. He solved the problem using new methods (which are now considered as rules of Graph Theory) and then generalized the solution. Using several hypothesis he proved some valuable theorems. We will consider 2 parts of his theorem to solve our problem.
Theorem Part A

A finite graph G contains an Euler circuit if and only if G is connected and it contains no vertices of odd degree.
Part B

A finite graph G contains an Euler path if and only if G is connected and it contains exactly two vertices of odd degree.

Mathematical Proof [4]


Necessary condition for Part A Suppose there exists an Euler Trail T on Graph G. The figure below shows a graph which has an Euler trail. To visualize the functionality of Euler graphs we would use 3 special vertices. - Starting Vertex There is only one Starting vertex for a given Euler trail. This is where the journey begins. - Middle Vertices There can be any number of middle vertices. These are visited through the journey. - End vertex There is only one end vertex for a given Euler trail. This is where the journey ends.
1 5 2 8

6 4 3

9 10

This is a sample graph with an Euler trail. There can be more Euler Trails for this graph.

RESEARCH PROJECT - I - JULY 2013

Middle vertices: When we consider a middle vertex we can see that a walk does not begin or cease there; when one enters the vertex, always he/she has to leave the vertex. Therefore if there is an incoming edge to enter the vertex there must be an outgoing edge to leave as well. Therefore, Number of times one enters a middle vertex = number of times one leaves that vertex If there are p number of incoming edges, there should be p number of outgoing edges for that vertex. Total number of edges connected to that vertex= 2p Degree of a middle vertex is always EVEN. You can see that each of the middle vertices in our example graph has even degree (2 or 4).

Starting vertex: When we consider the starting vertex, we do not enter to the vertex first but stay there before the walk begins. Walk is started by leaving the starting vertex. During the journey this node may act as same as a middle node. But when the journey begins there must be an additional leaving step (colored in black) taken place in the walk. Because of that, in a starting vertex there is always one more outgoing edge than incoming edges. So that, Number of times one leaves the starting vertex = number of times one enters + 1 If there are n number of incoming edges, there should be (n+1) number of outgoing edges for that vertex. Total number of edges connected to that vertex= 2n+1 Degree of the starting vertex is ODD in an Euler trail. You can see the degree of the starting vertex is odd (3).

RESEARCH PROJECT - I - JULY 2013

End vertex: For the End vertex we can see that finally we do not leave the vertex but stay there after the walk stops; there is no any outgoing from the end vertex afterwards. During the journey this node may act as same as a middle node. But when the journey ends there must be an additional incoming step (colored in yellow) taken place in the end vertex. Because of that, in an end vertex there is always one more incoming edge than outgoing edges. Number of times one enters the end vertex = number of times one leaves + 1 If there are k number of outgoing edges, there should be (k+1) number of incoming edges for that vertex. Total number of edges connected to that vertex = 2k + 1 Degree of the end vertex is ODD in an Euler trail. You can see the degree of the end vertex is odd (3).

In Addition to above logical reasoning, we have to prove that the graph should be connected. Suppose that the graph is disconnected, then it is obvious that one cannot find an Euler trail in the graph. Since none can cover all the edges in one visit. If the Graph is disconnected then Euler Trail does not exist. Using Contrapositive proof we know that; if an Euler trail exist then the graph should be connected.

In Summery

If there is an Euler trail in the graph; The graph is connected and it contains only 2 odd degree vertices. The trail should begin from one of the odd degree vertex and end from the other.

So we have proven the necessary condition of Part A

RESEARCH PROJECT - I - JULY 2013

Necessary condition for part B By definition, Euler Circuit is an Euler trail in which the end vertex and start vertex are the same. According to the necessary condition of part A, we found that in an Euler trail, Start vertex, - should be odd degree -if there are n number of incoming edges, there should be n+1 outgoing edges End Vertex, - should be odd Degree -if there are k number of outgoing edges, there should be k+1 incoming edges. To become an Euler Circuit, End vertex = Start vertex both should be either odd degree or Even degree Suppose they are in odd degree (as in the case of Euler trail) In-degree of end vertex (k+1) = In-degree of start vertex (n) k+1=n Out-degree of end vertex (k) = out-degree of start vertex (n+1) k=n+1 from the equations above n+1+1=n 2=0 therefore this is a contradiction. End vertex and start vertex cannot be odd degree. Start vertex, and end vertex is same and it should have even degree. We have already proven that, each middle vertex has even degree. So we can come to the conclusion, If there is an Euler circuit in a graph, the graph is connected and all vertices are even degree. Up to this point we have proved the necessary condition of part B.

RESEARCH PROJECT - I - JULY 2013

Sufficient condition for part B Following steps are based on the number of edges, to prove that if graph G is connected and each node is even degree, then there is an Euler circuit in G by mathematical induction. Statement: If graph G is connected and each node is even degree, then there is an Euler circuit in G. Base case: Consider a graph G with two vertices and two edges between them. This graph obviously has an Euler circuit.

Assumption: Now suppose that a graph with any k (k > 2) number of edges satisfies the above statement. Lets take the graph with m (m>k>2) number of edges: We start at an arbitrary vertex v and follow edges, arbitrarily selecting one after another until we return to v, call this trail W. We know that we will return to v eventually because every time we encounter a vertex other than v we are listing one edge adjacent to it. There is an even number of edges adjacent to every vertex, so that there will always be a suitable unused edge to list next. So this process will always lead us back to v.

a e d c b

W is denoted by the brown colored line in the figure. Let E be the edges of W. E={a,b,c,d,e} The graph ( G E ) has components C1,C2 Ck
RESEARCH PROJECT - I - JULY 2013 9

(G E) Graph: C1

C2

C3

Each circuit satisfies the induction hypothesis: connected, less than m edges and degree of each vertex is even. We know that every degree is even in (G E ), because when we removed W, we removed an even number of edges from those vertices which are connected by E. By induction, each circuit in ( G E ) has an Euler circuit; call them E1; E2..Ek. Since G is connected, there is a vertex ai in each component Ci on both W and Ei

A E1 V

E2

D
E3

Assume that as we follow W, the vertices A,B,C,D are encountered in that order. We find an Euler circuit in G by starting at V following W until reaching A, walking through the entire E2 and ending back at A, then follow W until reaching C, follow the entire E3 (if exists) and ending back at C and so on. The walk can be terminated at V after going through each vertex by following W until reaching ak, following the entire Ek, ending back at ak , then finishing W in the same way. If graph G is connected and each node is even degree, there should be an Euler Circuit in the graph So we have proven the sufficient condition of Part B. Up to now we have proven Part B completely.

RESEARCH PROJECT - I - JULY 2013

10

Sufficient condition of Part A This can be easily proven using Part B Suppose u and v are vertices of odd degree in graph G (G = (E, V)). Consider E U (u,v).This graph has all even degrees. By Part B, G has an Euler circuit. This circuit uses the edge (u,v). Thus we have an Euler trail in G when we omit the edge (u,v). If a graph is connected and has 2 Odd degree vertices, then it should have an Euler trail. Now we have completely proven both part A and part B with the necessary and sufficient conditions.

Research Problem Description


The Pregel River flowed through the town of Konigsberg, (present day Kaliningrad in Russia). Two islands protruded from the river. On either side of the mainland, two bridges joined one side of the mainland with one island and a third bridge joined the same side of the mainland with the other island. A bridge connected the two islands. In total, seven bridges thus connected the two islands with both sides of the mainland. Is it possible to cross each bridge exactly once during a single walk? Brief Introduction to History [6] According to lore of Konigsberg, the residents of Konigsberg used to spend Sunday afternoons walking around the city. While walking, they decided to create a game to devise a way in which they could walk around the city, crossing each of the seven bridges only once. They could not either to find a route or to prove that it was impossible. Leonard Euler, the famous mathematician who published over 500 books and papers in his lifetime, with another 400 posthumously from St. Petersburg where not too far from Konigsberg, was requested to solve Konigsberg problem by its citizens. First, he neglected the problem due to its trivial nature. But on August 26, 1735, Euler presented a paper containing the solution to the Konigsberg bridge problem. He addresses both this specific problem, as well as a general solution with any number of landmasses and any number of bridges. By solving the problem the way he did, Euler invented the subject of graph theory.

RESEARCH PROJECT - I - JULY 2013

11

Solution
To answer this problem we have to transform this scenario into mathematical model. To make the analysis easy we model this scenario as a Graph in which land masses are denoted by vertex V={A,B,C,D}and bridges E={a,b,c,d,e,f,g} are denoted by edges of the graph. 3

Now our original problem is transformed into a mathematical problem. Here we have to argue for the existence of a path which covers all edges in the graph once and only once. There are two ways we can accomplish such a single walk. One way is to start from one vertex and end at another or start and end at the same vertex, meaning there should be an Euler trail or an Euler circuit for the existence of such a walk. From the theory proved above we know that this graph doesnt have an Euler trail nor an Euler circuit since it has four odd degree vertices.

Eulers Solution [3]


In his Strategy he introduces a notation to describe a journey. If a traveler goes from A to B over bridge, Euler wrote this as AB where the first letter refers to the area the traveler is leaving, and the second refers to the area he arrives at after crossing the bridge. If 2 steps AB and BD combined he denoted by the three letters ABD. ABD = A->B, B->D. Like so ABCD = A->B, B->C, C->D
RESEARCH PROJECT - I - JULY 2013 12

It is clear with this notation, number of letters needed to represent the path which covers all bridges one and only once is, one more than the number of bridges. So we can write a General Expression: Number of characters needed to denote the path (Eulers trail) = number of bridges+1 For the Konigsberg problem we have to use 8 characters to denote the path (since there are 7 bridges) So the problem is reduced to finding a sequence of letters whose length is (bridges+1) number of letters. Number of occurrences of each character, depends on the number of edges connected with that vertex. Lets take a single landmass (vertex), with different number of edges (or bridges). If a traveler crosses this bridge, he must either have been in A before crossing, or have come into A after crossing, so that in either case the letter A will occur once in the representation described above (AX or XA). Like that, we can write a generalized expression for the number of occurrences for an odd degree vertex. Number of occurrence in the path for an Odd degree vertex= (Degree +1)/2

But when the number of bridges tends to a landmass is even, then the results depends on the starting position. If journey starts at A, then the number of occurrences will be 2; AXXXXXXAXX if the journey starts somewhere else, then the number of occurrences will be 1; XXXXAXXXX When we consider the other even degree cases, we can write a generalized expression for the number of occurrences for even degree vertices. If journey starts from A Number of occurrence for an Even Degree vertex= (Degree/2) + 1 If journey starts somewhere else Number of occurrence for an Even Degree vertex= (degree/2)

RESEARCH PROJECT - I - JULY 2013

13

So the existence of a path which cover all the bridges one and only once can be determined by the following manner. S=Total number of occurrences for each character B=Total number of Bridges If S=B+1 OR S=B then an Euler trail may exist, else there doesnt exist an Euler trail.

Applying this method to the original problem: Vertex A B C D Sum Degree 5 3 3 3 14 Number of Occurrence 3 2 2 2 9

B = 7 and S = 9 Therefor it is clear that such a walk does not exist in the Konigsberg problem.

RESEARCH PROJECT - I - JULY 2013

14

Problems
1. A postman has to deliver mails to 6 different places. The map of the area consisting of 6 houses is given. The postman needs to find a route to deliver all the mails for the given houses by walking through all the roads with no repetitions and returning to his place. Is there such a route for the postman? Justify your answer.

2. Determine if following statements are true or false. Give reasons to your answer. I. II. If each bridge must be used exactly twice in a walk, Konigsberg problem can be solved. A graph can be an Eulerian Circuit (Cycle) if and only if there doesnt exist any Loops.

RESEARCH PROJECT - I - JULY 2013

15

References
1. 2. 3. 4. 5. 6. www.wolframalpha.com www.wikipedia.org www.mathsdl.org http://mcis.jsu.edu www.maa.org www.cs.berkerley.edu

7. Daniel Eric Smith (March 9, 2010). Graph Theory: Denitions, Theorems, and Algorithms [Online]. Available: http://mcis.jsu.edu/faculty/desmith/ms108/graph-defn-algor.pdf

RESEARCH PROJECT - I - JULY 2013

16

You might also like