Math 396: Applied Combinatorics and Graph Theory

Lecture Notes
Created by: Dr. Amanda Harsy and Dr. Marie Meyer
June 21, 2019

1
.

2
Contents
1 Syllabus and Schedule 5
1.1 Group Project Description: . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2 Introduction to Graph Theory 19

2.1 Why Study Graphs? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2 ICE 1: Graphical Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.3 Basic Graph Theory Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3.1 Paths, Walks, Trails, and more . . . . . . . . . . . . . . . . . . . . . 33
2.4 ICE 2: Intro to Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.5 Families of Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.5.1 Path Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.5.2 Cycle Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.5.3 Complete Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.5.4 Bipartite Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.5.5 K-partite Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.5.6 Star Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.6 Creating New Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.6.1 Union Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.6.2 Complement Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.6.3 Join Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.6.4 Cartesian Product of Graphs . . . . . . . . . . . . . . . . . . . . . . . 42
2.6.5 Hypercubes Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.7 Digraphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.8 ICE 3: Classes of Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3 Degrees 47
3.1 ICE 4: Degrees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2 Regular Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.3 Degree Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.4 Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

4 Isomorphic Graphs 59
4.1 Useful Theorems to help with the Isomorphism Question . . . . . . . . . . . 62
4.1.1 Degree Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.1.2 Subgraphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.2 Other Theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.3 Ice 5: Isomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

3
5 Trees 69
5.1 Search Trees and Spanning Trees . . . . . . . . . . . . . . . . . . . . . . . . 74
5.1.1 Minimal Spanning Tree Algorithms for Weighted Trees . . . . . . . . 78
5.2 Ice 6- Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

6 Traversability 83
6.1 Hamiltonian Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.2 Activity: Constructing Disjoint Hamiltonian Cycles of K2n+1 . . . . . . . . . 85
6.3 ICE 7: Eulerian and Hamiltonian Graphs . . . . . . . . . . . . . . . . . . . . 88

7 Digraphs 90
7.1 Strong Digraphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
7.2 Tournaments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
7.3 ICE 8: Digraphs and Tournaments . . . . . . . . . . . . . . . . . . . . . . . 97

8 Network Algorithms 99
8.1 Shortest Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
8.2 Network Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
8.3 ICE 9: Network Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

9 Combinatorics 109
9.1 Basic Counting Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
9.1.1 Ice -Counting Principles . . . . . . . . . . . . . . . . . . . . . . . . . 113
9.1.2 Ice -Does Order Matter . . . . . . . . . . . . . . . . . . . . . . . . . . 115
9.2 Permutations and Combinations . . . . . . . . . . . . . . . . . . . . . . . . . 117
9.2.1 Ice -Permutations and Combinations . . . . . . . . . . . . . . . . . . 121
9.3 Pigeon Hole Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
9.4 ICE- More Counting and Pigeon Hole Principle . . . . . . . . . . . . . . . . 125

10 Generating Functions 127

10.1 Generating Function Models . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
10.1.1 ICE -Generating Function Models . . . . . . . . . . . . . . . . . . . . 131
10.2 Calculating Coefficients of Generating Functions . . . . . . . . . . . . . . . . 133
10.2.1 Coefficients of Generating Functions . . . . . . . . . . . . . . . . . . 137

4
1 Syllabus and Schedule
“Combinatorial reasoning underlies all analysis of computer systems. It plays
a similar role in discrete operations research problems and in finite probability.
Two of the most basic mathematical aspects of computer science concern the speed
and logical structure of a computer program. Speed involves enumeration of the
number of times each step in a program can be performed. Logical structure in-
volves flow charts, a form of graphs. Analysis of the speed and logical structure
of operations research algorithms to optimize efficient manufacturing or garbage
collection entails similar combinatorial mathematics. Determining the probability
that one of a certain subset of equally likely outcomes occurs requires counting the
size of the subset. Such combinatorial probability is the basis of many nonpara-
metric statistical tests. Thus, enumeration and graph theory are used pervasively
throughout the mathematical sciences”. -Alan Tucker, Applied Combinatorics

Thanks for taking Applied Combinatorics and Graph Theory with me! I have never taught
or taken a course like this, so I am excited to go through this journey with you!

As this is my first time teaching this course, I have a few goals in mind when putting this
workbook together for you. First I am trying to pick out a selection of topics which are
interesting, mathematically rigorous, but also practical, intuitive, and algorithmic. As such
we will have lots of definitions, theorems, lemmas throughout the course along with prac-
tical examples using these definitions and theorems. We will be doing some proofs, so it is
important to review your Introduction to Proofs materials and/or read through Appendix
1.2 and 3 from our textbook.

During the last third of the course, we will go into combinatorics, the study of counting, which
like graph theory has many applications in a variety of disciplines especially in computer sci-
ence, mathematics, and the physical sciences. Some of the main combinatorial questions one
may ask are, “Is there an existence of an arrangement” (is it possible to arrange these objects
in the following way) and ”what is the enumeration or classification of the arrangement.”1 .
One we know such an arrangement exists, we can then explore it further and try to find the
best or optimal arrangement. A typical solution of a combinatorial problem often includes
the following steps:
1. Set up a mathematical model.
2. Study the model.
3. Do some computation for small cases in order get get some intuition and insight on
the problem.
4. Use careful reasoning, previous results, and theorems to solve the problem.
1
Introductory Combinatorics by Richard Brualdi

5
I am looking forward to teach you these exciting topics which are useful in many areas of
applied mathematics. Specifically, two of my main research areas: Predictive Modeling of
Sports Ranking Systems and Modeling DNA Self-Assembly use graphs, so I am very excited
to go through this material with you! I have worked hard to create this course packet for
you, but it is still a work in progress. Please be understanding of the typos I have not
caught, and politely bring them to my attention so I can fix them for the next time I teach
this course. I look forward to meeting you and guiding you through this course!

Cheers,
Dr. H

Acknowledgments: No math teacher is who she is without a little help. I would like to
thank my own undergraduate professors from Taylor University: Dr. Ken Constantine, Dr.
Matt Delong, and Dr. Jeremy Case for their wonderful example and ideas for structuring
excellent learning environments. I also want to thank Dr. Marie Meyer who will teach this
course for 6 weeks and also has given significant insight for this course. And finally, I would
like to thank you and all the other student for making this job worthwhile and for all the
suggestions and encouragement you have given me over the years to improve.

6
 MATH-396-03 APPLIED COMBINATORICS AND GRAPH THEORY
Spring 2019 SYLLABUS

Instructor: Amanda Harsy Office: AS-124-A

Email: [email protected] Office phone: 815-836-5688
Class Time: TR 12:30-1:45 Class Room: AS-157-A
Office Hours: M 12-2, T 2-3, W 1:30-2:00, R 2:30-3:30, or by appt
Office hours are subject to change, once I find out meeting times for my committees, but I will let you know if they change. Remember you
can ALWAYS make an appointment.

Special Note: I am due to have a baby at the end of February. During this time, we will have class
as normal and Dr. Marie Meyer will be taking over for me. Her office is AS-120-A and her email is
[email protected] . During these 4-6 weeks, I will not be holding normal office hours, but will be
available through email. Dr. Meyer will be available during this time. Her office hours are M 12-1pm and
2-3pm, T 9-10am, W 2-3pm, F 12-1pm, or by appointment.

Course description
This course covers a selection of standard topics from combinatorics and graph theory with an emphasis
on the applications of these topics.

About the Course

Text: A First Course in Graph Theory By Gary Chartrand, Ping Zhang. (ISBN: 0486483681)
Suggested Text: Applied Combinatorics By Alan Tucker (ISBN: 0470458380)
Prerequisites: none.

Course Objectives:
To help each student
• understand and use concepts of Graph Theory and Combinatorics,
• apply graph theory and combinatorial concepts to various applications,
• become better at solving mathematical problems,
• experience doing mathematics cooperatively,
• effectively communicate mathematics, and
• enjoy his/her experience with mathematics.

Resources
Blackboard: Check the Blackboard website regularly (at least twice a week) during the semester for
important information, announcements, and resources. It is also where you will find the course discussion
board. Also, check your Lewis email account every day. I will use email as my primary method of com-
munication outside of office hours.

Help: Don’t wait to get help. Visit me during my office hours, use the discussion forum in Blackboard,
go to the Math Study Tables, find a study partner, get a tutor!
7
Dr. Harsy’s web page: For information on undergraduate research opportunities, about the Lewis
Math Major, or about the process to get a Dr. Harsy letter of recommendation, please visit my website:
http://www.cs.lewisu.edu/~harsyram.
1
2 MATH-396-03 APPLIED COMBINATORICS AND GRAPH THEORY SPRING 2019 SYLLABUS

Course Requirements And Grading Policy

Homework: Almost every week, I will collect a homework assignment. I will post these homework
assignments on Blackboard. You may work with others on the homework, but it must be your own work.
If I catch you copying homework, you will get a 0. Please see the academic honesty section below. Some-
times the homework include a “warm-up” assignment to either supplement what was learned in class or
to introduce what will be covered in the following class.

Graph Theory and Applications Final Project: You, as part of a 3-4 person group, will select a
topic pertaining to the course, read additional material pertaining to the topic, write a short paper summa-
rizing the main ideas and then give a 20 minute presentation to the class on your topic. This presentation
will during our final exam block which is on 5/13 at 10:30am-12:30pm. Rubrics for this project are posted
in Blackboard.

Mastery-Based Testing: This course will use a testing method called, “Mastery-Based Testing.”
There will be three (3) paper-and-pencil, in-class Mastery Exams given periodically throughout the semes-
ter. In mastery-based testing, students receive credit only when they display “”mastery,” but they receive
multiple attempts to do so. The primary source of extra attempts comes from the fact that test questions
appear on every subsequent test. Since this course has never been offered before, the timing and number
of concepts for MBT may change, but for now we expect Test 1 will have 6 questions. Test 2 will have 12
questions -a remixed version of the questions from Test 1 and new questions on the new concepts. Test 3
will have 16 questions- a remixed version of the questions from Test 2 and four new questions. We will also
have three testing weeks. During these weeks, students can use doodle to sign up to retest concepts during
(extended) office hours and math study tables. Students are allowed to test any concept, but cannot retest
that concept the rest of the week. So for example, a student can test concepts 2, 3 and 5 on Monday and
concept 6 on Tuesday, but would not be able to test concept 5 again. These testing weeks are tentatively
set for 3/18-3/22, 4/11-17, and during Finals week. Students will be able to retest the new concepts from
the last exam twice during the final testing week.

Grading of Mastery-Based Tests: The objectives of this course can be broken down into 18 main
concepts/problems. For each sort of problem on the exam, I identify three levels of performance: master
level, journeyman level, and apprentice level. I will record how well the student does on each problem
(an M for master level, a J for journeyman level, a 0 for apprentice level) on each exam. After the Final
testing week, I will make a record of the highest level of performance the student has made on each sort
of problem or project and use this record to determine the student’s total exam grade. Each of the first
8 concept/questions students master will count 9% points towards their exam grade. After that, each
concept/question will be worth 3.5% towards your exam grade. So for example, if you master 10 of the 15
concepts your grade will be a 79%.

This particular way of arriving at the course grade is unusual. It has some advantages. Each of you will
get several chances to display mastery of almost all the problems. Once you have displayed mastery of a
problem there is no need to do problems like it on later exams. So it can certainly happen that if you do
well on the midterms you might only have to do one or two problems on the Final. (A few students may
not even have to take the final.) On the other hand, because earlier weak performances are not averaged
in, students who come into the Final on shaky ground can still manage to get a respectable grade for the
course. This method of grading also stresses working out the problems in a completely correct way, since
accumulating a lot of journeyman level performances only results in a journeyman level performance. So it
pays to do one problem carefully and correctly as opposed to trying get four problems partially correctly.
Finally, this method of grading allows you to see easily
8 which parts of the course you are doing well with,
and which parts deserve more attention. The primary disadvantage of this grading scheme is that it is
complicated. At any time, if you are uncertain about how you are doing in the class I would be more than
glad to clarify the matter. The tentative test dates are below.
 MATH-396-03 APPLIED COMBINATORICS AND GRAPH THEORY Spring 2019 SYLLABUS 3

Exam # Date
1 Thursday, 2/28
2 Thursday, 4/4
3 Tuesday, 4/30
Grade scale
90%-92% A- 93%-96% A > 97% A+
80%-82% B- 83%-86% B 87%-89% B+
70%-72% C- 73%-76% C 77%-79% C+
60%-62% D- 63%-66% D 67%-69% D+
< 60% F
Category Percentage
Mastery Exams 65
Homework 20
Project 15

Course Policies and Procedures

Calculator/Technology/Phone Policy: No phones or other forms of technology can be used on
in-class exams unless specifically noted. A basic scientific or graphing calculator is ok. Please do not have
your phones out unless it is a class activity. If I see a phone out during a quiz or exam, you will receive an
F on that assessment.

Make-Ups: There will be no make-ups for any assignments. If you are late or miss class, your assign-
ment will not be accepted and there will be no make-up offered, except in extenuating and unpredictable
circumstances. If you will miss class for a justifiable and unavoidable reason, you can contact me before
you miss class and it is possible you can have a make-up. If you do not contact me and explain your
absence, you will not be allowed a make-up.

Class Attendance: Students are expected to attend all classes as part of the normal learning process.
In addition, students must be especially consistent in attendance, both on-ground and online, during the
first two weeks of the semester to confirm registration and to be listed on the official course roster. Students
who fail to follow this procedure and who have not received prior approval from the instructor for absences
will be withdrawn from the courses in question by certification of the instructor on the official class lists.
Instructors may publish specific, additional reasonable standards of attendance for their classes in the
course syllabus. Students may receive failing grades if they do not observe attendance requirements. The
Illinois Student Assistance Commission also requires attendance as a demonstration of academic progress
toward a degree as one criterion for retaining financial aid awards. (2015-2016 Undergraduate Catalog, p.
34).

Attendance is critical for success in this course and is required of all students without exception. A student
absent from class is responsible for all material covered that day. I usually post solutions to ICE sheets and
post notes. I will stop doing this if there is a trend of absences. In other words, I will not post lecture notes
unless there is consistent attendance. Students are expected to attend all classes as part of the normal
learning process.

Course Relationship to Mission: Lewis University, guided by its Catholic and Lasallian heritage,
provides to a diverse student population programs for a liberal and professional education grounded in
the interaction of knowledge and fidelity in the search for truth.Lewis promotes the development of the
complete person through the pursuit of wisdom and 9justice. Fundamental to its Mission is a spirit of asso-
ciation, which fosters community in all teaching, learning and service. This course embraces the Mission
of the University by fostering an environment in which each student is respected as an individual within
a community of learners. In the spirit of the vision of Lewis University, the goals and objectives of this
4 MATH-396-03 APPLIED COMBINATORICS AND GRAPH THEORY SPRING 2019 SYLLABUS

course seek to prepare students to be successful, life-long learners who are intellectually engaged, ethically
grounded, socially responsible, and globally aware.

Academic Integrity: Scholastic integrity lies at the heart of Lewis University. Plagiarism, collusion
and other forms of cheating or scholastic dishonesty are incompatible with the principles of the University.
Students engaging in such activities are subject to loss of credit and expulsion from the University. Cases
involving academic dishonesty are initially considered and determined at the instructor level. If the student
is not satisfied with the instructors explanation, the student may appeal at the department/program level.
Appeal of the department /program decision must be made to the Dean of the college/school. The Dean
reviews the appeal and makes the final decision in all cases except those in which suspension or expulsion
is recommended, and in these cases the Provost makes the final decision.

University Copyright and Intellectual Properties Guidelines can be found at https://www.lewisu.edu/

academics/library/pdf/Copyright-Intellectual-Property-Guidelines.pdf

University Student Complaint Policy: The University Student Complaint Policy can be found at lewisu.
edu/studentcomplaints.

University Grade Appeal Policy: The University Grade Appeal Policy can be found at lewisuedu/
studentcomplaints.

Classroom Decorum: In order to maintain an environment conducive to learning and student develop-
ment, it is expected that classroom discourse is respectful and non-disruptive. The primary responsibility
for managing the classroom environment rests with the faculty. Students who engage in any prohibited or
unlawful acts that result in disruption of a class may be directed by the faculty member to leave class for
the remainder of the class period. Students considered to be a disruption or who present a threat of poten-
tial harm to self or others may be referred for action to the Dean of Student Services. 2015-2016 Student
Handbook, p. 14, Lewis University website https://www.lewisu.edu/sdl/pdf/studenthandbook.pdf.

Sanctified Zone: This learning space is an extension of Lewis Universitys Sanctified Zone, a place
where people are committed to working to end racism, bias and prejudice by valuing diversity in a safe
and nurturing environment. This active promotion of diversity and the opposition to all forms of prej-
udice and bias are a powerful and healing expression of our desire to be Signum Fidei, Signs of Faith,
in accordance with the Lewis Mission Statement. To learn more about the Sanctified Zone, please visit:
http://www.lewisu.edu/sanctifiedzone.

Center for Health & Counseling Services

To support student success, all Lewis students are eligible for free health and mental health services on the
Romeoville campus. This includes commuters and those living on campus, part-time and full-time students,
graduate and undergraduate students, and those taking Lewis classes at other locations. For more infor-
mation, visit the Center for Health & Counseling website at www.lewisu.edu/studentservices/health
or call (815)836-5455.

Requests for Reasonable Accommodations Lewis University is committed to providing equal ac-
cess and opportunity for participation in all programs, services and activities. If you are a student with a
disability who would like to request a reasonable accommodation, please speak with the Learning Access
Coordinator at the Center for Academic Success and Enrichment (CASE). Please make an appointment
by calling 815-836-5593 or emailing [email protected]. Since accommodations require early
10
planning and are not provided retroactively, it is recommended that you make your request prior to or
during the first week of class. It is not necessary to disclose the nature of your disability to your instructor.
 MATH-396-03 APPLIED COMBINATORICS AND GRAPH THEORY Spring 2019 SYLLABUS 5

For more information about academic support services, visit the website at: www.lewisu.edu/CASE.

Lewis University has adopted Blackboard Ally providing alternative formats for files uploaded by in-
structors. Students can click the down arrow next to any file, and select Alternative Formats.

Applied Combinatorics and Graph Theory Student Learning Outcomes:

Upon successful completion of this course, students should be able to
(1) Understand and identify the fundamental definitions and concepts of graph theory
(2) Use Graphs to model and solve mathematical problems
(3) Use Combinatorics to solve mathematical problems

Mathematics Program Learning Outcomes:

A Lewis University Mathematics Graduate will be able to:
(1) Model the essential components of a problem mathematically so that it can be solved through
mathematical means.
(2) Solve problems through the appropriate technology.
(3) Write clear and concise formal proofs of mathematical theorems, and evaluate the validity of proofs
written by others.
(4) Solve mathematics problems as a member of a group and in front of a group.
(5) Describe and leverage the interdependency of different areas of mathematics, the connections be-
tween mathematics and other disciplines, and the historical context for the development of math-
ematical ideas.
Of these, this course addresses program learning outcomes 1, 4, and 5, all at a reinforced level.

Lewis University Baccalaureate Characteristics:

A Lewis University undergraduate education is marked by the following characteristics:
(1) Essential Skills: The baccalaureate graduate of Lewis University will read, write, speak, calculate,
and use technology at a demonstrated level of proficiency.
(2) Major Approaches to Knowledge: The baccalaureate graduate of Lewis University will under-
stand the major approaches to knowledge.
(3) Faith, Religion and Spirituality: The baccalaureate graduate of Lewis University will under-
stand the place of faith, religion, and spirituality in the search for truth and meaning.
(4) Moral and Ethical Decision-Making: The baccalaureate graduate of Lewis University will
understand and prepare for moral and ethical decision-making.
(5) Responsible Citizenship: The baccalaureate graduate of Lewis University will become an in-
formed, involved, and responsible citizen of a diverse yet interconnected national and global com-
munity through a grounding in economic, political, social, and historical influences that are inherent
in shaping, developing, and advancing nations and the world.
(6) Critical Thinking: The baccalaureate graduate of Lewis University will think critically and
creatively.
(7) Lifelong Learning: The baccalaureate graduate of Lewis University will possess the knowledge,
skills, and dispositions to enter or advance a career, or to begin graduate study.
This course addresses Baccalaureate Characteristics 1, 6, and 7 all at a reinforced level. It provides
students essential skills in mathematics, abstraction, and algorithmic thinking. It reinforces quantitative
11
and empirical approaches to knowledge. In asking students to reduce problems to their important compo-
nents and charting a path for solving them, it fosters critical thinking. Finally, in casting mathematics as
a dynamic and evolving influencer of our collective future, it encourages curiosity and lifelong learning.
6 MATH-396-03 APPLIED COMBINATORICS AND GRAPH THEORY SPRING 2019 SYLLABUS

Math Baccalaureate
Student Learning Outcome Demonstrated by:
Program Characteristic
homework, exams,
Understand and identify properties of graphs 1 6, 7
class work,
homework, exams,
Use graphs to model and solve mathematical problems 1,4,5 1,6,7 class work,
final group project
homework, exams,
Use combinatorics to solve mathematical problems 1,4,5 1,6,7 class work,
final group project

This, That, Tips, and Suggestions...

Here are some suggestions that may help you in the class.

Situation: I am a student that learn best from writing everything down and I wish the lecture notes weren’t so
typed out.
Suggestion: Writing is the best way for me to remember things too! I want to encourage you to write if it helps
you. Using lecture notes keep me somewhat organized and helps us get through all of the material (of which there
is a ton)! Also, my handwriting is horrible so having some words typed is good for you. My suggestion is to
rewrite your lecture notes into a notebook. Rewriting notes is a wonderful study tool and I encourage you to do
this if you don’t think the partially filled out notes allow you to learn best.

Situation: I can’t see the document camera.

Suggestion: I am used to higher quality document cameras and am waiting to get a better one. I have horrible
board handwriting so I don’t like to write on the board for the majority of class. I do write on the board sometimes.
Also I like the fact that the document camera allows me to face the class, so hopefully I can keep tabs on how the
class is doing. My suggestion for you is to move to a different seat if you can’t see the board. And also please let
me know if you are having issues and I can try to refocus/adjust the document camera.

Practice, Practice, Practice: The most common suggestion from past students is to do ALL the homework.
Learning mathematics is like any activity and requires that you practice in order to become proficient.

Note: This syllabus is subject to change. If there is a change, it will be discussed in class or through an email
and will be updated in Blackboard.

Class Notation
• R represents all real numbers (so no imaginary numbers)
• Z represents integer numbers
• Q represents rational numbers [numbers that can be written as a fraction pq where p,q are integers]
• { } represents a set. For example {x; such that x is an integer }=Z
• ∈ means “contained in”. For example, 5 ∈ <
• ∈
/ means “not contained in”. For example π ∈ / Z.
• s.t. means “such that”
• “|” or “:” mean “such that” in a set setting. For example {x|x ∈ Z}= {x : x ∈ Z} =
{x; such that x is an integer } = Z
• ••• means “therefore”.
• ∃ means “there exists”
• ∀ means “for all”
• ⇒ means “this implies”
• ⇐⇒ or “iff” means “if and only if”
12
• Pf means “Proof”
Applied Combinatorics and Graph Theory Tentative Schedule Spring 2019
This schedule is subject to change, but you will be notified of any changes ahead of time.

Agenda Special Notes

Week 1: 1/14
Intro to Graphs and Graph Models
Week 2: 1/21
Graph Definitions and Properties HW 1 due
Week 3: 1/28
Degrees HW 2 due
Week 4: 2/4
Isomorphisms HW 3 due
Week 5: 2/11
Trees HW 4 due
Week 6: 2/18
Traversability Mastery Exam 1 on Thursday
Mastery Exam 1
Week 7: 2/25
Flex HW 5 due
Week 8: 3/4
Digraphs HW 6 due
Retesting Week 1 Retesting Week 1
Topic for Group Presentation Due Thursday
Spring Break: 3/11
No class this week!

Week 9: 3/18
Network Algorithms HW 7 due

Week 10: 3/25

Additional Graph Theory Topics HW 8 due

Week 11: 4/1 Mastery Exam 2 on Thursday

Additional Graph Theory Topics
Mastery Exam 2
Week 12: 4/8
Introduction to Combinatorics Retesting week 2 Thurs-next Wed
(multiplicative and additive counting Outline of paper & talk due Thursday
principles, permutations and (in class)
combinations, and the pigeonhole
principle)
Week 13: 4/15 No class on Thursday due to
Generating Functions Easter Break
Retesting week 2 Retesting week 2
HW 9 due Tuesday
Week 14: 4/22
Recurrence relations Draft of Paper due Tuesday
HW 10 due
Week 15: 4/29
Flex HW 11 due
Mastery Exam 3 Mastery Exam 3 on Tuesday
Week 16: 5/6 13
Group Presentations during our final Final Testing week
exam block: Thursday, 10:30 a.m.- 12:30 Final Paper due Thursday
Final Testing Week Presentations on Thursday
1.1 Group Project Description:
See next page!

14
Graph Theory and Combinatorics Applications Final Project

Overview: You will select a topic pertaining to the course, read additional material
pertaining to the topic, write a short paper summarizing the main ideas and then give a
20 minute presentation to the class on your topic.

Goals: So far this semester, we’ve seen a number of both theoretical and applied results
from graph theory. This is your chance to either explore one of those topics in more
depth or to explore a more advanced topic. You will also gain experience reading journal
articles from mathematics or another discipline and presenting technical information.

Parameters:
• The choice of topic must be made in consultation with Dr. Harsy or Dr. Meyer,
some ideas are listed below. The topics below are very broad; you’ll want to
quickly narrow to a single idea.

• Your paper and presentation must make significant use of at least one journal
article, must include at least one application to a field outside graph theory, and
must include at least one significant graph theoretic result (though not necessarily
its proof). You do not necessarily need to understand (or even) read the entire
journal article.

• You may consult online sources such as Wikipedia, though the substance of your
paper and presentation must come from other sources.

• You must carefully cite every source you use for your paper or presentation, using
an inline reference to a bibliography/works cited page (you can use MLA, APA,
or the LaTeX bibtex format). I can show you how to easily do this using Google
Scholar and LaTeX. You may also include a “Works Consulted” list for sources
which did not have a direct impact on your final product but which were helpful
in getting you there. A source you glance at only briefly need not be placed in
the Works Consulted list.

Citing our sources is a way of giving thanks to those whose work has influenced
you and enables later scholars to retrace your intellectual journey. Additionally, it
enables you and others to distinguish your contributions (perhaps only
expositional!) from those of others.

• Your paper should be 5 – 10 pages long, though significant use of figures may
make it longer. The target audience can be either your classmates or someone
working in another field who may benefit from graph theoretic concepts. The
grading rubric for the paper is attached.

• Your presentation should be 20 minutes long and is aimed at your classmates. It is

likely a condensed version of part of your paper. You may like to use slides to
enable rapid dissemination of figures15or background ideas. The grading rubric for
the presentation is attached.
 Deadlines:

Thursday, March. 7: Final decision on project topic. Have 2 back-ups in case of multiple
groups wanting the same topic.

Thursday, April 11: Outline of both paper and talk due in class.

Tuesday, April 23. Draft of paper due.

Thursday, May 9: In class presentations (to be scheduled later) 10:30am-12:30pm

Friday, May. 10: Final version of paper due.

Presentation Grading Rubric:

(30%) Significant graph theoretic content beyond what was done in class.
(15%) Effective contextualization of the mathematics. Does the speaker motivate the ideas
introduced? Does the speaker make connections with other subjects?
(25%) Clarity of presentation. Is the presentation understandable by the class? Is all
relevant notation and terminology defined? Is effective use made of examples?
(20%) Preparation. Are visual aids well-designed? Is effective use made of the board? Has
the talk been practiced?
(5%) Group Work -were you a good group member?
(5%) Student Audience Evaluation -was the talk interesting and easy to follow?

Paper Grading Rubric:

(30%) Organization. Is the organization logical? Is there both flow and structure?
(15%) Motivation. Is clear why the reader should be interested in the topic of the paper?
(35%) Depth. Does the paper make significant use of at least one journal article? Does the
paper explain connections to a subject outside of graph theory? Does the paper contain
significant mathematics.
(15%) Mechanics. Are grammar, punctuation, spelling (mostly) correct? Are citations
correct? Are quotation marks used appropriately?
(5%) Group Work -were you a good group member?

16
 Ideas for paper topics
You may suggest your own topic! But here are some ideas.

Note: You may not base your project on any project/paper/assignment/material

from another course, though you are welcome to use the project to delve deeper
into connections with another course.

• Independent Sets and Graph Homomorphisms.

As an undergraduate Yufei Zhao solved a prominent open conjecture in graph
theory concerning the number of independent sets in a d-regular graph. In a
recent Monthly article, Zhao explains his solution and gives lots of references for
understanding the significance of the problem.
• Algorithms. Code some of the more difficult algorithms from the course and
investigate ways they’ve been developed or optimized by others.
• Random Graphs. A random graph is created by choosing a vertex set and then
assigning edges between vertices according to some probability distribution.
Investigate how the randomness interacts with various graph theoretic properties
or explore how random graphs are used in mathematical modelling.
• Colorful Variations. Explore variations of the 4-color problem.
• Game Theory and the Chromatic Number. Coloring the vertices of a graph can
be viewed in game theoretic terms and some of the theorems we’ve learned about
can be reproven from that perspective.
• Graphs and Knot Theory. In knot theory, the Kauffman bracket is an
exceptionally simple idea with profound consequences. It is calculated using
recursion relationships similar to those in graph theory. Similarly, other
numbers associated to graphs can be turned into knot invariants.
• Chromatic Number and Scheduling In this project you’ll investigate scheduling
algorithms related to the chromatic number of a graph.
• Percolation. Percolation is an idea from statistical physics in which quantities at
one vertex are redistributed to other vertices. There are close connections to
physics, chemistry, and probability theory.
• Small-World Networks. Many social networks are modelled using “small-world
networks”. You’ll investigate what these networks are and what their properties
are. You could perhaps begin with an article on the network of interactions
between the characters in Game of Thrones.
• Random Walks on Graphs A random walk in a graph is a created by moving
from one vertex to an adjacent vertex with some probability. One basic question
is “What’s the probability of returning to the original vertex?”
• Graphs and Games: Investigate how graphs are used to study board games
(typically as Markov Chains)
• Constantine’s Problem: A classic problem about how to distribute legions to
maintain effective control over the empire. Fun and accessible.
• Modeling DNA Self-Assembly. Graphs can be used to help design
nanostructures. Joe Ellis-Monaghan has17 some papers which use tile-based
assembly and graph theory results to help model this. Note if you are one of
Dr. Harsy’s researchers, you cannot choose this topic.
• Traveling Salesman Problem: Given a collection of cities and the cost of travel
between each pair of them, find the cheapest way of visiting all of the cities
 and returning to your starting point. Represent the situation as an undirected
weighted graph. This is one of the most intensively studied problems in graph
theory and has yet to be solved; however, there is progress to report as well as
solutions for specific cases.
• Spanning trees or Matrix Representation of a Graph: Spanning Trees:
Compare and contrast algorithms for finding spanning trees of a graph and
present/prove a way to count the spanning trees using the Laplacian matrix of
a graph. Kirchoff's Matrix-Tree Theorem will be useful) Matrix Representation
of a Graph: We use powers of the adjacency matrix of a graph G to compute
the number of walks between pairs of vertices in G. The Laplacian matrix is
another matrix uniquely associated to G that can be used to compute the
number of spanning trees of G. Present a proof of Kirchoff's Matrix-Tree
Theorem and explore the connection with linear algebra.
• Dijkstra's Algorithm: Using this algorithm, one can find the shortest path
between two vertices. Explore applications to different transportation
problems.
• Greedy Algorithms: A greedy algorithm involves making the best optimal
choice at each local stage with the goal of finding a globally optimal solution.
Choose one or two greedy algorithms to study and look into problems which
do have solutions under the given greedy algorithms.
• Have your own idea? Share your idea with Dr. Harsy or Dr. Meyer!

We may add other potential project ideas and will update you via Bb if we do!

18
2 Introduction to Graph Theory
2.1 Why Study Graphs?
Many problems can be modeled and solved with the help of graphs, directed graphs, and
pseudo-graphs. We can model networks (social, website, circuit, etc.), job assignments and
scheduling, chemical storage, traffic flow, spell checkers, athletic events, nanostructure de-
sign, surveillance, and more. In this section, we give a few examples of how we can use
graphs to help with a variety of problems. These problems and examples will continue to
come up throughout the course as we learn new graph or combinatorial ideas.

Example 2.1. You can use graphs to model outcomes of sporting events like this “Poor
sportsmanship” model. How would you rank these teams?
A

B D

Example 2.2. Dr. Harsy uses graphs in her research modeling DNA self-assembly:

Figure 1: Representing the complementary cohesive end types (left), 3 -armed branched
junction molecule (center left) with example tile representation (center right) Complete
Graph(right)

19
Example 2.3. 2 The Fantastic Math Kitties have started their own publishing company,
The Cat’s Meow. They have hired 10 editors in the various areas of Naps, Strings, Bird-
watching, Milk, etc. These editors have been divided into 7 committees to meet each Friday
at 3 potential time periods to discus specific topics of interest to the company, advertising,
securing reviewers, finances, etc. Some pairs of committees cannot meet during the same
period because more than one editor may be on both committees. The editors are split in
the following ways for the committees: c1 = {1, 2, 3}, c2 = {1, 3, 4, 5}, c3 = {2, 5, 6, 7}, c4 =
{4, 5, 8, 9}, c5 = {2, 6, 7}, c6 = {8, 9, 10}, c7 = {1, 3, 9, 10}. Use a picture or a graph to help
organize and model this problem.

2
Example adopted from A first course in Graph Theory

20
Example 2.4. 3 The figure below shows the traffic lanes at the intersection of two busy
streets. When a vehicle approaches this intersection,it could be in one of 9 lanes: L1, L2, ...,
L9. We want to program the traffic light so that they tell the drivers when to proceed through
the intersection in a way that will not result in a wreck. For example cars in L1 and L7
should not drive through the intersection at the same time, but L1 and L5 could! Represent
this as a graph, G, with V (G) = {L1, L2, ..., L9} and having the vertices joined by an edge
if the vehicles in these two lanes cannot safely enter the intersection at the same time.

3
Example from A first course in Graph Theory

21
 4
Example 2.5. A spelling checker looks at each word X (represented in a computer as a binary
number) in a document and tries to match X with some word in its dictionary, which typically
contains close to 100,000 words. To understand how this checking works, we consider the simplified
problem of matching an unknown letter X with one of the 26 letters in the English alphabet. In
the spirit of the strategy humans use to home in on the page in a dictionary where a given word
appears, the computer search procedure would first compare the unknown letter X with M, to deter-
mine whether X ≤ M or X > M . The answer to this comparison locates X in the first 13 letters
of the alphabet or the second 13 letters, thus cutting the number of possible letters for X in half.
This strategy of cutting the possible matches in half can be continued with as many comparisons
as needed to home in on Xs letter. Continuing in this manner we can determine what letter X is.
Model this process using a graph.

For reference: ABCDEFGHIJKLMNOPQRSTUVWXYZ

Note: For our original spelling-checker problem, a word processor would use a similar, but larger,
tree of comparisons. With just 12 rounds of comparisons, it could reduce the number of possible
matches for an unknown word X from 100,000 down to 25, about the number of words in a column
4
Example from Applied Combinatorics

22
of a page in a dictionary. (Once a list was reduced to about 25 possibilities, a computer search for
X would usually run linearly down that list, just as a human would.)

23
Example 2.6. Suppose Figure 2.1 below represents a section of a citys street map. We want to
position police officers at corners (vertices) so that they can keep every block (edge) under surveil-
lance -that is, every edge should have police officers at (at least) one of its end vertices. What is
the smallest number of police officers that can do this job?5

Figure 2: Figure from from Applied Combinatorics

5
Example from Applied Combinatorics

24
Example 2.7. Suppose psychological studies of a group of people determine which members of the
group can influence the thinking of others in the group. We can make a graph with a vertex for
each person and a directed edge (a, b) whenever person a influences b. The figure below models
such a situation. What is a minimal subset of people who can spread an idea through to the whole
group, either directly or by influencing someone who will influence someone else, and so forth. In
graph-theoretic terms, we want a minimal subset of vertices with directed paths to all other vertices.6

Figure 3: Figure from from Applied Combinatorics

6
Example from Applied Combinatorics

25
.

26
2.2 ICE 1: Graphical Models

Figure 4: Figure from Applied Combinatorics

1. Suppose the graph above represents a network of telephone lines. We are interested in the
networks vulnerability to accidental disruption. We want to identify sets of those lines and
switching centers that must stay in service to avoid disconnecting the network.
a) Is there a telephone line (edge) whose removal will disconnect the telephone network
(graph)?

b) Is there a vertex whose removal disconnects the graph?

c) Is there any pair of edges whose removal disconnects the graph?

d) Find a minimal set of edges needed to link together the 11 vertices in Figure 2.2.

2. Answer the question from Ex 2.6.

3. Answer the question from Ex 2.7.

27
.

28
2.3 Basic Graph Theory Terms
Definition 2.1. Considered in an abstract setting, a graph G = (V, E) is defined in terms of a
set of vertices, V , and a set of edges, E, joining different pairs of vertices. We denote the number
of vertices and edges of G respectively by |V (G)| and |E(G)|.
V (G) is the vertex set of G and E(G) is the edge set of G.
Although we could have infinite graphs with infinite vertices, most of this class will focus only on
finite graphs. Sometimes vertices are called points or nodes and edges are called lines.
Example 2.8. What is V (G), |V (G)|, E(G), |E(G)| for the graph below?

a b c

d e f

Figure 5: Example of a graph

Definition 2.2. If two vertices are connected by an edge, we say these vertices are .
So in the example above, a and b are adjacent, but a and f are not.

And if e = uv is an edge of G, vertices u and v are adjacent. In fact we say u and v are neighbors
and are joined by e.
Definition 2.3. If two edges share the same vertex, they are called edges.
Definition 2.4. The of a given vertex is the number of edges incident to that vertex.
So for the graph in Figure 5, vertex a has degree since edges (a, b) and (a, d) are incident
edges.
Definition 2.5. The order of a graph G is given by the number of vertices.
The size of a graph G is given by the number of edges of G.
Note: Some textbooks use size for the order of a graph (as does Dr. H -oops).

Example: The order of the graph in Figure 5 is . The size is .

Example 2.9. The Trivial Graph is a graph with 1 vertex.

So a non-trivial graph must have at least vertices.

29
Example 2.10. Draw a graph with 1 vertex and one edge.

This is called a . Some textbooks allow loops in graphs, but your textbook only allows
them in “pseudo-graphs.” Loops do come up in applications so it is important to know what they
are.

We also have a variation of graphs that our textbook calls, “Multigraphs.”

Definition 2.6. A multigraph, M, consists of a finite, non-empty vertex set and edge set in which
every two vertices of M are joined by a finite number of edges (possibly zero).

This means we can have something called, parallel edges in multigraphs which are multiple edges
between the same vertices.

Pseudo-graphs can have parallel edges and loops (edges joining a vertex to itself).

Figure 6: Example of a graph

Example 2.11. Which of diagrams in Figure 6 are graphs?

(a) M1 & M4

(b) M1, M2, & M4

(c) M4

(d) all of the diagrams

(e) none of the diagrams

30
Example 2.12. Which of diagrams in Figure 6 are multigraphs?

(a) M1 & M4

(b) M1, M2, & M4

(c) M4

(d) all of the diagrams

(e) none of the diagrams

Example 2.13. Which of diagrams in Figure 6 are pseudographs?

(a) M1 & M4

(b) M1, M2, & M4

(c) M4

(d) all of the diagrams

(e) none of the diagrams

Example 2.14. What are the degrees of each of the vertices of the graph below? Are they odd or
even?

31
Example 2.15. The graph below has...

a (a) 6 edges, 4 vertices (exactly 2 of which are

b
odd)

(b) 4 edges, 6 vertices (all of which are odd)

d c (c) 6 edges, 4 vertices (all of which are odd)

(d) 4 edges, 4 vertices (exactly 2 of which are

odd)

Example 2.16. All vertices of a graph could be odd.

(a) True and I am very confident!

(b) True, but I am not very confident.

(c) False and I am very confident!

(d) False, but I am not very confident.

Example 2.17. All vertices of a graph could be even.

(a) True and I am very confident!

(b) True, but I am not very confident.

(c) False and I am very confident!

(d) False, but I am not very confident.

(e) I love math!

32
2.3.1 Paths, Walks, Trails, and more
Often we want to explore various ways in which we can move around in a graph especially if our
vertices are representing locations. These next few definitions introduce common terminology about
moving about a graph.

Definition 2.7. A W is a sequence of vertices with each pair of consecutive vertices

in P joined by an edge. We can denote this as W = (x1 , x2 , ..., xn ) or W = x1 − x2 − x3 − ... − xn .

Definition 2.8. If a walk starts and ends at the same vertex, we say it is a closed walk otherwise
it is an open walk.

The length of a walk is the number of edges in the walk.

Definition 2.9. A P is a sequence of distinct (no repeated) vertices with each pair of
consecutive vertices in P joined by an edge. We can denote path as P = (x1 , x2 , ..., xn ) or sometimes
P = x1 − x2 − x3 − ... − xn .

Definition 2.10. A T is a walk without any repeated edges. We can denote path as
T = (x1 , x2 , ..., xn ) or sometimes T = x1 − x2 − x3 − ... − xn .

Definition 2.11. A circuit in a graph G is a closed trail of length 3 or more.

This means a circuit starts and ends at the same vertex but repeats no edges.

Definition 2.12. A circuit that does not repeat any vertex except the first and last is called a
. A k-cycle is a cycle of length k.

Examples:

a a a
e e e

d d d

b c b c b c

Definition 2.13. We say a graph G is if there is a path between every pair of vertices.

Note: The Trivial graph is connected since each vertex has a path connecting to itself.

We will talk more about connected graphs later.

33
.

34
2.4 ICE 2: Intro to Graphs

Figure 7:

1. Referring to the graph above, give an example of each or explain why such an example doesn’t
exist.
(a) an x-y walk of length 6

(b) an x-z path of length 3

(c) a v-w trail that is not a v-w path.

2. Draw a graph with 4 vertices and 5 edges.

35
3. a) Let S = {2, 3, 4, 7, 11, 13} Construct the multigraph M whose vertex set is S and where ij
is an edge for distinct elements i and j in S whenever i + j ∈ S.

b) Answer the same problem with this rule: i and j are joined by two edges if both i + j ∈ S
and |i − j| ∈ S

4. Draw a graph or pseudograph with 4 vertices (all odd) and 5 edges.

5. Draw a graph with 4 vertices (all even) and 5 edges (no loops).

36
2.5 Families of Graphs
There are many families of graphs.

a b a b a 1

b
d c d c 2

Figure 8: C4 (left), K4 (center), and K3,2 (right)

2.5.1 Path Graphs

A graph is called a path graph if the vertices of G of order n can be labeled v1 , v2 , ...vn so that
its edges are v1 v2 , v2 v3 , v3 v4 , ..., vn−1 vn . We denote path graphs as P1 , P2 , P3 , ....

2.5.2 Cycle Graphs

A graph is called a cycle graph if the vertices of G of order n ≥ 3 can be labeled v1 , v2 , ...vn so
that its edges are v1 v2 , v2 v3 , v3 v4 , ..., vn−1 vn , AND vn v1 . We denote cycles graphs as C1 , C2 , C3 , ....

37
2.5.3 Complete Graphs
A complete graphs of order n are graphs in which every pair of vertices is connected by an edge. In
other words, every two distinct vertices are adjacent. We denote complete graphs as K1 , K2 , K3 , ...
and these graphs have the maximum possible size for a graph with n vertices.

Question: What is the size of Kn ?

Depending on the circumstance, sometimes it is helpful to change how one represents a graph. For
example, the Complete Graph on 4 Vertices, K4 , can be represented by both graphs in the Figure
9 and, in some cases, it may be nicer to have a representation of K4 without intersecting edges.

a b a

d
d c
b c

Figure 9: Two representations of K4 .

Example 2.18. Can you draw some different representations of path graphs, cycle graphs, and
complete graphs?

38
2.5.4 Bipartite Graphs
Some graphs are classified by rules which have their vertex sets partitioned in particular ways. We
call graphs in which the vertices can be partitioned into two sets A and B (called partition sets) in
which every vertex in A connects with every vertex in B, but no vertex in A connects to another
vertex in A, and no vertex in B connects with another vertex in B, bipartite graphs.

Examples:

If G is a bipartite graph with partition sets A and B and if every vertex in A connects with with
every vertex in B and every vertex in B connects with every vertex in A, say G is a complete
bipartite graph. If A has n vertices and B has m vertices, we denote G as Kn,m .

Example 2.19. Is C4 a bipartite graph?

(a) Yes and I am very confident!
(b) Yes, but I am not very confident.
(c) No and I am very confident!
(d) No, but I am not very confident.
(e) Does this graph come with snacks?

Example 2.20. Is C5 a bipartite graph?

(a) Yes and I am very confident!
(b) Yes, but I am not very confident.
(c) No and I am very confident!
(d) No, but I am not very confident.
(e) Is it time to go home yet?

39
Theorem 2.1. A nontrivial graph G is a bipartite graph if and only if G contains no odd cycles.

Proof in Textbook. We will prove the statement: If a nontrivial graph G is a bipartite graph then
G contains no odd cycles. by proving the contrapositive statement:

2.5.5 K-partite Graphs

We can generalize bi-partite graphs for k partition sets instead of 2.
Examples:

40
2.5.6 Star Graphs
If G is a complete bipartite graph in which one of the partition sets has order 1, G is called a star
graph.
Examples:

There are many other bipartite graph families including Mongolian Tent Graphs, Book Graphs,
Cross-Prism Graphs, Crown Graphs, and more...

2.6 Creating New Graphs

There are several ways to create a new graph from a given graphs. Here are a few:

2.6.1 Union Graph

Given graphs G and H, the union graph of G and H, denoted, G ∪ H, is the new graph whose
vertex set is V (G) ∪ V (H) and whose edge set is E(G) ∪ E(H).
Example: G = 2C4 ∪ P3 ∪ K4 :

2.6.2 Complement Graph

Given a graph G, the of G, denoted Ḡ, is a graph whose vertex set is V (G) for which
each distinct vertex u, v ∈ V (G), edge uv is in Ḡ if and only if uv is NOT an edge of G.

Example:

41
Example 2.21. Draw C¯3 .

Example 2.22. What is K¯n ?

2.6.3 Join Graph

Given graphs G and H, the join G + H consists of G ∪ H and all edges joining a vertex of G and
a vertex of H.

Example 2.23. Draw C4 + P1 .

2.6.4 Cartesian Product of Graphs

Given graphs G and H, the Cartesian Graph Product G × H (sometimes denoted GH) is the
graph with vertex set V (G × H) = V (G) × V (H). That is every vertex is an ordered pair (u, v)
where u ∈ V (G) and v ∈ V (H). We also have the following edge property: distinct vertices (u, v)
and (x, y) are adjacent in G × H if either
1) u = x and vy ∈ E(H) or
2) v = y and uxy ∈ E(G).

Example 2.24. Draw P2 × P3 .

42
2.6.5 Hypercubes Graphs
Example 2.25. Draw C4 × K2

The graph above is sometimes denoted Q3 and is called a 3-cube. We can generalize this in the
following way.
The graphs Qn are called n-cubes or hypercubes and they are defined recursively in the following
way:

Q1 = K2 and Qn = Qn−1 × K2 .

Examples:

These hypercubes can also be defined as the graphs whose vertex set is the set of ordered n-tuples
of 0s and 1s (called n-bit strings) and where two vertices are adjacent if their ordered n-tuples differ
in exactly one coordinate.

2.7 Digraphs
Remember the “Poor Sportsmanship Graph”? This was an example of a directed graph or
digraph.

Definition 2.14. A digraph or directed graph D is a finite nonempty set of vertices connected
together, where all the edges are directed from one vertex to another. This means our Edge set can
be represented by ordered pairs of distinct vertices which are called arcs or directed edges.

Examples:

43
Terminology: Given an edge (u, v) in E(D), we say u is adjacent to v and v is adjacent from u.
We say u and v are both incident with edge (u, v).

Note: We could have both edges (u, v) and (v, u) in a digraph. But if at most one of (u, v) or
(v, u) is in the digraph for each pair of distinct vertices of the digraph, we say the digraph is an
oriented graph.

Example 2.26. Suppose Team A beats teams B and C. Team B beats Team D. Team D beats A
and C. And C beats B. Draw the directed graph representing the Poor Sportsmanship model AND
list the Edges of the digraph as ordered pairs. Is this graph oriented?

44
2.8 ICE 3: Classes of Graphs
1. Draw P3 ∪ P2

2. Draw P3 + P2

3. Is P3 × P2 the same as P2 × P3 from our notes?

4. What is P2 × P2 ?

5. What is K¯n + P1

6. What is K̄3 + P¯2

45
.

46
3 Degrees
There are many numbers which can describe a graph. We have already discussed order and size
and degree. Often this isn’t enough to describe the full structure of the graph (as we will see in
our next section), but knowing the degree of each vertex is a big help for determining information
about a given graph.

Recall, the degree of a given vertex is the number of edges incident to that vertex.

Definition 3.1. A vertex of degree 0 is called an vertex and a vertex of degree 1 is an

vertex or .
Definition 3.2. We denote the minimum degree of a graph G, by
We denote the maximum degree of a graph G, by

Example 3.1. Given a graph of order n what are the upper and lower bounds for δ(G) and ∆(G)?
(a) 0 ≤ δ(G) ≤ ∆(G) ≤ n
(b) 0 ≤ δ(G) ≤ ∆(G) ≤ n − 1
(c) 1 ≤ δ(G) ≤ ∆(G) ≤ n
(d) 1 ≤ δ(G) ≤ ∆(G) ≤ n
(e) 0 ≤ δ(G) ≤ ∆(G) ≤ n + 1
Example 3.2. Given a graph of order n what are the upper and lower bounds for any vertex?

Example 3.3. Draw 4 different graphs (not all from the same family). Count the number of edges
and vertices for each graph and count the sum of the degrees of the vertices.

Question: Does there seem to be a relationship between any of these numbers?

47
Example 3.4. If you could draw a graph with exactly one odd vertex, what would the sum of all
the degrees be?

Theorem 3.1 (The First Theorem of Graph Theory). If G is a graph of size m, then
X
deg v =
v∈V (G)

Proof:

Example 3.5. Suppose we want to construct a graph with 20 edges and have every vertex of degree
4. How many vertices must the graph have?

48
Example 3.6. Suppose a certain graph G has order 14 and size 27. The degree of each vertex is
3,4, or 5. There are six vertices of degree 4. How many vertices of G have degree 3 and how many
have degree 5?

What are some other ways we could have attempted to solve this problem?

Corollary 3.1. Every graph has an even number of odd vertices.

Proof:

49
Theorem 3.2. Let G be a graph of order n. If for every two nonadjacent vertices u and v of G,
deg u+ deg v ≥ n − 1, then G is connected and the diam(G)≤ 2.

Proof: See text.

n−1
Corollary 3.2. If G is a graph of order n with δ(G) ≥ , then G is connected.
2
Proof:

Example 3.7. True or False: A graph of order 7 such that δ(G) ≥ 3 is connected.

(a) True and I am very confident!

(b) True, but I am not very confident.

(c) False and I am very confident!

(d) False, but I am not very confident.

50
3.1 ICE 4: Degrees
1. Can a graph have nine vertices of which 4 have degree 2, three have degree 3 and two have
degree 4?

(a) True and I am very confident!

(b) True, but I am not very confident.
(c) False and I am very confident!
(d) False, but I am not very confident.

2. Must the number of people at a party who do not know an odd number of other people be
even?

3. Must the number of people ever born who had (have) an odd number of brothers and sisters
be even?

4. Must the number of families in Alaska with an odd number of children be even?

5. Is it possible to have a group of seven people such that each person knows exactly three other
people in the group?

51
6. How many vertices will the following graphs have if they contain:

(a) 12 edges and all vertices of degree 3.

(b) 21 edges, three vertices of degree 4, and the other vertices of degree 3.

(c) 24 edges and all vertices of the same degree.

7. Consider a group of people at a party. Suppose some guests shook hands with some of the
other guests. If we asked everyone at the party how many people they shook hands with and
added them all up, what would be the total number of hands shook?

52
3.2 Regular Graph
Definition 3.3. If the vertices of a graph G all have the same degree, we say G is . If
all vertices of G are of degree k, we say G is .

Examples:

Example 3.8. Could you have a k-regular graph of order n if k and n are odd?

(a) True and I am very confident!

(b) True, but I am not very confident.

(c) False and I am very confident!

(d) False, but I am not very confident.

Theorem 3.3. Let k and n be integers with 0 ≤ k ≤ n − 1 Then there exists a k-regular graph of
order n if and only if at least one of r and n is even.

Proof: See text.

3.3 Degree Sequences

Often we don’t have regular graphs, thus it makes sense that we may want to describe our graphs
by the degrees the vertices in the graph have.

Definition 3.4. If the degrees of the vertices of a graph G are listed in a sequence s, then s is
called a degree sequence of G.

Definition 3.5. A finite sequence of nonnegative integers are said to be if it is the

degree sequence of some graph.

Note: There are a lot of algorithms and applications which use this which we will get to later in
the course.

Example 3.9. What is the degree sequence of the graphs below:

53
Example 3.10. Is the degree sequence: s1 = 3, 3, 2, 2, 1, 1 graphical?

(a) Yes and I am very confident!

(b) Yes, but I am not very confident.

(c) No and I am very confident!

(d) No, but I am not very confident.

Example 3.11. Is the degree sequence: s2 = 6, 5, 5, 4, 3, 3, 3, 2, 2 graphical?

(a) Yes and I am very confident!

(b) Yes, but I am not very confident.

(c) No and I am very confident!

(d) No, but I am not very confident.

Example 3.12. Is the degree sequence: s3 = 7, 6, 4, 4, 3, 3, 3 graphical?

(a) Yes and I am very confident!

(b) Yes, but I am not very confident.

(c) No and I am very confident!

(d) No, but I am not very confident.

Example 3.13. Is the degree sequence: s3 = 3, 3, 3, 1 graphical?

(a) Yes and I am very confident!

(b) Yes, but I am not very confident.

(c) No and I am very confident!

(d) No, but I am not very confident.

54
Theorem 3.4 (Havel-Hakimi Theorem). A non-increasing sequence s = d1 , d2 , ..., dn where n ≥ 2
of non-negative integers where d1 ≥ 1 is graphical if and only if the sequence is graphical:

s1 : d2 − 1, d3 − 1, ..., dd1 +1 , dd1 +2 , ..., dn

Proof: See Text.

Example 3.14. Is the sequence graphical: s : 7, 7, 4, 3, 3, 3, 2, 1

Example 3.15. Is the sequence graphical: s : 5, 3, 3, 3, 3, 2, 2, 2, 1

55
3.4 Matrices
We can describe a graph using V (G) and E(G), but we can also describe a graph using matrices
which is especially nice when using computers.

Definition
( 3.6. The adjacency matrix of G is the n × n matrix A = [aij ] such that
1 if vi vj ∈ E(G)
aij =
0 otherwise

Definition
( 3.7. The incidence matrix of G is the n × m matrix B = [bij ] such that
1 if vi is incident with ej
bij =
0 otherwise

Example 3.16. Find the adjacency and incidence matrices of the graph below.

v1 v2 v5

v4 v3

What are some observations/properties of these two matrices?

56
It turns out power of these matrices also have information!

Theorem 3.5. Let G be a graph with vertex set V (G) = {v1 , v2 , ..., vn } and adjacency matrix
(k)
A = [aij ]. Then the entry aij in row i and column j of Ak is the number of distinct vi − vj walks
of length k in G.
(2) (3)
Note: This means aii =deg vi and aii is twice the number of triangles in G that contain vi .
Proof: Uses Induction, but we won’t go over it because it involves linear algebra which not everyone
has taken.

v1 v2 v5

v4 v3

 
3 1 1 2 1
1 2 2 1 1
A2 = 
 
1 2 2 1 1
2 1 1 4 0
1 1 1 0 1
 
4 5 5 6 2
5 2 2 6 1
A3 = 
 
5 2 2 6 1
6 6 6 4 4
2 1 1 4 0

57
58
4 Isomorphic Graphs
So far we have been dancing around the following question:
“How can we tell if two graphs are really the same graph, but are just drawn differently and with
different names for the vertices?”

We have noticed that graphs are essentially the same graph even if they aren’t drawn the same
way. We call graphs that have the same number of vertices and identical connections (but may
look different) isomorphic.

Example:

Who cares? Here are some examples where the isomorphism question comes into play:
“Researchers working with organic compounds build up large dictionaries of compounds that they
have previously analyzed. When a new compound is found, they want to know if it is already in
the dictionary. Large dictionaries can have many compounds with the same molecular formula but
differing in their structure as graphs (and possibly in other ways). Then one must test the new
compound to see if its graph-theoretic structure is the same as the structure of one of the known
compounds with the same formula (and the same in other ways)that is, whether the new compound
is graph-theoretically isomorphic to one of a set of known compounds.

A similar problem arises in designing efficient integrated circuitry for a computer. If the design
problem has already been solved for an isomorphic circuit (or if a piece of the new network is
isomorphic to a previously designed circuit), then valuable savings in time and money are possible.”7

Informal definition of isomorphic graphs: Two graphs are isomorphic if one can be “stretched”
or “shifted” to look like the other without adding, breaking, or removing any parts.

Note: You can think of the vertices as steel balls and the edges as rubber bands. We assume balls
will remain in whatever position we place them and that the rubber bands never break. Using this
idea, two graphs are isomorphic if we can move the balls around to make one graph look like the
other.

Definition 4.1. Formal definition of isomorphic graphs: Graphs G and H are isomorphic
if there exists a bijective function f : G → H that preserves adjacencies of the vertices of G.

In other words, f is injective (or ) and surjective (or ) and if vertices x and
y in G are adjacent, so are f (x) and f (y) in H.

7
Tucker’s Applied Combinatorics

59
Definition 4.2. We say that the function f : X → Y is injective if the following property is true:
Whenever f (x1 ) = f (x2 ) it must be true that = .

Notice, in Figure 10, we see that the transformation represented on the left is 1-1, but the trans-
formation represented on the right is not because both v1 and v2 map to w4 , but v1 6= v2 .

w1 w1
v1 v1
w2 w2
v2 w3 v2 w3
v3 v3
w4 w4
v4 v4
w5 w5
v5 v5
w6 w6
v6 v6
w7 w7

Figure 10:

Process to show Injectivity: Start with two arbitrary elements in the range of f and suppose
they are equal, that is suppose f (u) = f (v) and work backwards to see if u = v. In other words,
can we find two possible input values that map to the same output? If yes, f is NOT injective.

Definition 4.3. We say that the function f : X → Y is surjective if every element in Y is

mapped to. That is, if y ∈ Y , then there exists a x ∈ X so that = .

Idea: Dr. Harsy is going to do her absolute worst to present you with the most obscure element
of the codomain. Are you certain that this function will be able to map to this element no matter
what Dr. Harsy presents you with?

Process to show surjectivity: Pick an arbitrary element in co-domain, find an element that
maps to it.

Example 4.1. Consider the function f : R → R where f (x) = x2 . Note that this is neither
injective nor onto. Why?

60
Example 4.2. Create a bijective map between these two graphs in order to show these graphs are
isomorphic. 8

Example 4.3. Are these graphs isomorphic?

Example 4.4. There are 11 non-isomorphics graphs of order 4. Let’s find them!

8
image from Tucker’s Applied Combinatorics

61
Example 4.5. If two graphs have the same size and order, they are isomorphic.

(a) True and I am very confident!

(b) True, but I am not very confident.

(c) False and I am very confident!

(d) False, but I am not very confident.

(e) Snack time?

Example 4.6. Are these two graphs isomorphic?

Example 4.7. Are these two graphs isomorphic?

4.1 Useful Theorems to help with the Isomorphism Question

Determining whether two graphs are isomorphic is one of the fundamental and notoriously chal-
lenging questions in the field of graph theory. Luckily we have some tools (theorems) which can
help us!

4.1.1 Degree Sequences

Theorem 4.1. If two graphs are isomorphic, they have the same degree sequence.

Proof:

62
Example 4.8. Is the converse of this statement true, “If two graphs have the same degree sequence
are the graphs isomorphic?” In other words, if two graphs have the same size, order, and degree
sequence, must they be isomorphic?

(a) Yes and I am very confident!

(b) Yes, but I am not very confident.

(c) No and I am very confident!

(d) No, but I am not very confident.

(e) More snacks?

Example 4.9. For the graphs below, give reasons why the graphs could be isomorphic and give
reasons why the graphs could not be isomorphic.

Example 4.10. Consider the two graphs below. What is the degree sequence of each graph and
determine if they are isomorphic.

63
4.1.2 Subgraphs
Definition 4.4. A subgraph G’ of a graph G is a graph formed by a subset of vertices and edges
of G. We write G0 ⊆ G.

Definition 4.5. If a subgraph G’ of a graph G has the same vertex set as G, we say G0 is a
of G.

Definition 4.6. We say F is an subgraph of G if whenever u and v are vertices of G

and uv is an edge of G, then uv is an edge of F.

Examples:

Theorem 4.2. If two graphs are isomorphic, then subgraphs formed by corresponding vertices and
edges must be isomorphic.

Proof: Omitted.

Subgraphs can be used to test for isomorphism in the following way. If a graph G has a set of six
vertices forming a chordless circuit of length 6 (chordless means there are no other edges between
these six vertices except the six edges forming the circuit), then any graph isomorphic to G must
also have a set of six vertices forming such a chordless 6-circuit.

Example 4.11. Use a subgraph argument to determine whether the graphs below are isomorphic:

64
4.2 Other Theorems
Theorem 4.3. Two graphs are isomorphic if and only if their complements are isomorphic.

Proof: Omitted.

Example:

Theorem 4.4. If G and H are isomorphic then G is bipartite if and only if H is bipartite

Theorem 4.5. If G and H are isomorphic then G is connected if and only if H is connected

Definition 4.7. Two Digraphs D and D’ are isomorphic if there exists a one to one correspon-
dence f : V (D) → V (D0 ) such that (u, v) ∈ E(D) if and only if (f (u), f (v)) ∈ E(D0 ).

65
.

66
4.3 Ice 5: Isomorphisms
1. For the graphs below, give reasons why the graphs could be isomorphic and give reasons why
the graphs could not be isomorphic.

2. Use the complement of these two graphs to determine if the graphs below are isomorphic.

3. Determine whether the following graphs are isomorphic. Explain your reasoning or state the
isomorphism.

(a)

67
(b)

(c)

(d)

(e)

68
5 Trees
The most commonly used graph is called a Tree. They are very useful for organizing information
and search procedures (directories in a UNIX operating systems, word sorting, computer networks
and parallel computing). We will also revisit these graphs after we discuss Euler and Hamilton
Circuits.
There are several ways to define a tree graph.
Definition 5.1. Tree Definition 1: A Tree Graph is a connected graph with no circuits.
Definition 5.2. Tree Definition 2: A Tree Graph is a graph with a designated vertex called a
root such that there is a unique path from the root to any other vertex in the tree.

Figure 11: Example of a tree.

Figure 12: Is this graph a tree?

We say a Tree Graph is if any vertex can be the root of the graph. If we have a directed
tree graph we call the graph a tree.

69
 Figure 13: Two pictures of the same graph (Picture from Applied Combinatorics).

We usually draw a rooted tree with the root at the top of the figure. We say the root is at level
0. The vertices directly below the root are at level 1, and so on. For any non-root vertex, x, in
a rooted tree the parent of x is the vertex y with edge (y, x) into x. The children of x are the
vertices z with directed edge from x to z. Two vertices with the same parents are called siblings.
Vertices of trees with no children are called leaves. If every internal vertex of a rooted tree has m
children, we call the tree an m-ary tree.

Figure 14: Rooted Trees (Picture from Applied Combinatorics.).

70
Definition 5.3. A graph which has no circuits is called .

So you can also define a tree as a connected acyclic graph. We call acyclic graphs .

Figure 15: Example of a forest.

Theorem 5.1. Let T be a connected graph. Then the following statements are equivalent.
a) T has no circuits.
b) Let a be any vertex in T. Then for any other vertex x in T, there is a unique path Px between a
and x.
c) There is a unique path between any pair of distinct vertices x, y in T.
d) T is minimally connected, in the sense that the removal of any edge of T will disconnect T.

Proof:

71
Theorem 5.2. A tree with n vertices has n − 1 edges.

Proof:

Theorem 5.3. Let T be an m-ary tree with n vertices, of which i vertices are internal, then
n = mi + 1.

Proof: Homework

Corollary 5.1. Let T be an m-ary tree with n vertices consisting of i internal vertices and l leaves.
If we know one of the parameters n, i, or l, then we can find the other parameters by the following
formulas:

a) Given i, then l = (m − 1)i + 1 and n = mi + 1

l−1 ml − 1
b) Given l, then i = and n =
m−1 m−1
(m − 1)n + 1 n−1
c) Given n, then l = and i =
m m

Example 5.1. Suppose 63 people sign up for a tennis tournament, how many matches will be
played in the tournament?

72
Definition 5.4. The of a rooted tree is the length of the longest path from the root. In
other words, it is the largest level number of any vertex.

Definition 5.5. A rooted tree of height h is called if all leaves are at levels h and h − 1.

Balanced trees are “good” trees. Why is it important for a tennis tournaments tree to be balanced?

Theorem 5.4. Let T be an m-ary tree of height h with l leaves. Then,

(a) l ≤ mh , and if all leaves are at height h, l = mh .
(b) h ≥ dlogm le, and if the tree is balanced, h = dlogm le.

Proof of a)

Determining the number of leaves and height of more complex search trees is a major concern in
the field of computer science called analysis of algorithms. Below is formula for the number of
different undirected trees on n labeled vertices. This formula was first proved by Cayley in 1889.

Theorem 5.5. There are nn−2 different undirected trees on n labels.

Proof: Omitted.

73
5.1 Search Trees and Spanning Trees
Trees are used to find solutions to problems that involve a sequence of choices, whether hunting
through a graph for a particular vertex (which may represent something like the cheapest solution).
Many classic graph theory problems which involve isomorphisms, Hamilton circuits, minimal col-
orings, and optimal graphical modeling use tree-based searching for computerized solutions. By
letting the sequential choices be internal vertices in a rooted tree and the solutions and representing
“dead ends” by leaves, trees help us better organize and solve such problems. The main concern
we have when using trees to solve these problems is that we want to design it so the search is
exhaustive -that is we want to make sure we have checked ALL possible options.

The field of Optimization Research uses very large trees and special “pruning” algorithms to solve
their problems. We won’t go into any major research questions, but the ideas from this section
may show up later when we discuss network algorithms and the traveling salesman problem.

Recall, A spanning subgraph of G is a subgraph that contains all of G. Thus the

following definition makes a lot of sense:

Definition 5.6. A of G is a tree containing all vertices of G.

Note, unlike spanning subgraphs, if a graph is not connected, we will not be able to construct a
spanning tree.

There are 2 traditional ways to construct a spanning tree:

Creating a depth-first spanning tree:

1. Pick some vertex as the root and begin building a path from the root composed of edges of
the graph.

2. Continue this path until it cannot go any further without repeating a vertex already in the
tree.

3. We will have a leaf at the vertex where this path is forced to stop.

4. Now backtrack to the parent of the leaf and try to build a path from the parent, say vertex
y, in another direction.

5. Continue until all possible paths from this parent (y) and children have been built.

6. Now backtrack to find the parent of vertex y and repeat this process.

7. Stop when we have come back to the root and have checked all other possible paths from the
root.

74
Creating a breadth-first spanning tree:
1. Pick some vertex as the root, say x and put in all edges leaving x (along with the vertices at
the ends of these edges) in the tree.
2. Add to the tree the edges leaving the vertices adjacent from x, unless such an edge goes to
a vertex already in the tree.
3. Continue to build the tree in this level-by-level manner.
Note: If we can create a spanning tree using either of these methods, this means the graph is
connected. So this process could be used to check whether a graph is connected.

Example 5.2. Find a depth-first spanning trees for the graph below:

Example 5.3. Find a breadth-first spanning trees of the graph above.

75
Example 5.4. Use Spanning Trees to determine whether the graph with the following adjacency
matrix is connected:

Proposition 5.1. Show that in an n-vertex graph, a set of n-1 edges that form no circuits is a
spanning tree.
Proof:

Proposition 5.2. A breadth-first spanning tree consists of shortest paths from the root to every
other vertex in the graph.
Proof: Omitted.

76
Example 5.5. Suppose we are given three pitchers of water, of sizes 10 quarts, 7 quarts, and 4
quarts. Initially the 10-quart pitcher is full and the other two empty.We can pour water from one
pitcher into another, pouring until the receiving pitcher is full or the pouring pitcher is empty. Is
there a way to pour among pitchers to obtain exactly 2 quarts in the 7- or 4-quart pitcher? If so,
find a minimal sequence of pourings to get 2 quarts in the 7-quart or 4-quart pitcher.

77
5.1.1 Minimal Spanning Tree Algorithms for Weighted Trees
There are several algorithms for finding a minimum spanning tree. We will talk about 2 of them.

Kruskal’s Algorithm:
Given a connected weighted graph G, we can construct a spanning tree T of G in the following
way:

1. Start with the vertex set of G.

2. Have the first edge e1 be any edge of G of minimum weight.

3. For e2 select any remaining edge of G with minimal weight.

4. For e3 select any remaining edge of minimum weight that does not produce a cycle.

5. Continue until a spanning tree is produced.

Example 5.6. Use Kruskal’s Algorithm to create a spanning tree for the following graph:

78
Prim’s Algorithm:
Given a connected weighted graph G, we can construct a spanning tree T of G in the following
way:

1. Start with the vertex set of G.

2. Pick an arbitrary vertex, v, of G.

3. For e1 select an edge incident with v with minimal weight.

4. For e2 , e3 , etc. select incident edges of minimum weight among the edges having exactly one
of its vertices incident with an edge already selected.

5. Continue until a spanning tree is produced.

Example 5.7. Use Prim’s Algorithm to create a spanning tree for the following graph:

79
5.2 Ice 6- Trees
1. Suppose a telephone chain is set up among 100 parents to warn of a school closing. It is
activated by a designated parent who calls a chosen set of three parents. Each of these three
parents calls given sets of three other parents, and so on. How many parents will have to
make calls? Repeat the problem for a telephone tree of 200 parents.

2. A tree T has order 13 with degrees of its vertices only be 1,2, and 5. If T has exactly 3
vertices of degree 2, how many end-vertices does it have?

3. Find a depth-first spanning tree for K6 .

4. Find a breadth-first spanning tree for K6 .

80
5. Use Kruskal’s Algorithm to create a spanning tree for the following graph:

6. Use Prim’s Algorithm to create a spanning tree for the following graph:

81
82
6 Traversability
6.1 Hamiltonian Graphs
The figure below shows a diagram of an art museum that is divided into 15 exhibition rooms. At
the end of the day, a security officer enters the reception room by the front door and checks each
exhibition room to make certain that everything is in order. Can the officer make a round trip to
visit each room only once and return to the reception room?

How could we turn this situation into a graph? What does the question become?

Definition 6.1. A cycle in a graph G that contains every vertex of G is called a Hamiltonian
cycle of G. A Hamiltonian graph is a graph that contains a Hamiltonian cycle. A path in G
that contains every vertex of G is called a Hamiltonian path.

Example 6.1. List some families of graphs which are Hamiltonian.

83
Example 6.2. The graph below is the art museum situation from the first page.

Is G Hamiltonian? Does G have a Hamiltonian path?

Question: If G contains a Hamiltonian cycle, must G contain a Hamiltonian path?

Example 6.3. Show that the following graph is not Hamiltonian.

84
6.2 Activity: Constructing Disjoint Hamiltonian Cycles of K2n+1
Recall a partition of a nonempty set A is a collection of nonempty subsets of A such that every
element of A belongs to exactly one of these subsets. In other words, a partition of A is a collection
of pairwise disjoint nonempty subsets of A whose union is A.

Definition 6.2. A graph G is said to be decomposable into the subgraphs H1 , H2 , . . . Hk if

{E(H1 ), E(H2 ), . . . , E(Hk )} is a partition of E(G). Such a partition produces a decomposition
of G.

Example 6.4. We can decompose G below into a cycle and a path.

See handout for activity.

Theorem 6.1. The complete graph K2n+1 has a decomposition into n disjoint Hamiltonian cycles.

Proof Idea. The algorithm that proves this was our class activity.

Theorem 6.2. The complete graph K2n has a decomposition into n disjoint Hamiltonian paths
and a decomposition into n − 1 disjoint Hamiltonian cycles and a matching.

(We will study matchings on a graph at a later time.)

Proof Idea. Run the algorithm from above and delete vertex 2n + 1.

Example 6.5. If 17 students dine together at a circular table during a conference, and if each
night each student sits next to a pair of different students, how many days can the conference last?

85
Definition 6.3. A graph G that is not connected is called disconnected. A connected subgraph of
G that is not a proper subgraph of any other connected subgraph of G is a component of G. The
number of components of G is denoted k(G).

Thus a graph G is connected if and only if .

Theorem 6.3. If G is a Hamiltonian graph, then for every nonempty proper set S of vertices of
G,
k(G − S) ≤ |S|.

Proof:

The above theorem provides a condition for G to be Hamiltonian. Thus we can use
the contrapositive to prove G is not Hamiltonian.

Contrapositive:

Example 6.6. Show the graph below is non-Hamiltonian.

86
The following result provides a sufficient condition for a graph to be Hamiltonian.

Theorem 6.4. Let G be a graph of order n ≥ 3. If

deg u + deg v ≥ n

for each pair u, v of nonadjacent vertices of G, then G is Hamiltonian.

Proof Sketch.

Example 6.7. Show the following graph is Hamiltonian using the Theorem above.

Corollary 6.1. Let G be a graph of order n ≥ 3. If deg v ≥ n/2 for each vertex v of G, then G is
Hamiltonian.

Proof.

Conclusion: Although Eulerian circuits and Hamiltonian cycles seem to have similar definitions, it
is now clear they are dramatically different. Determining whether G is Euclidean is straightforward;
however, determining whether G is Hamiltonian can be extremely difficult because there is no
useful characterization of Hamiltonian graphs to this day.

87
6.3 ICE 7: Eulerian and Hamiltonian Graphs
1. Determine if the following graph is Eulerian, Hamiltonian, both, or neither. Support your
answer.

2. Find an Eulerian trail and a Hamiltonian path in the following graph.

3. Prove that the following graphs are not Hamiltonian.

88
4. Show the following complete tripartite graph, K3,3,3 is Hamiltonian.

5. When is K`,m,n for `, m, n ∈ Z Hamiltonian? Provide some families of K`,m,n which are and
are not Hamiltonian.

6. Try to answer question Question 5 for any multipartite graph.

89
7 Digraphs
Recall a digraph, also known as a directed graph, D consists of a vertex set V and an edge set
E. The ordered pairs of E are arcs or directed edges. If for each distinct u, v ∈ V , at most one
of (u, v) and (v, u) is a directed edge of D, then D is an oriented graph. Here, D is referred to
as an orientation of G.

Definition 7.1. If (u, v) ∈ E(D), then u is adjacent to v and v is adjacent from u. The
number of vertices to which a vertex v is adjacent is the outdegree of v and is denoted od(v).
The number of vertices from which v is adjacent is the indegree of v, denoted id(v).

Example 7.1. Draw two different orientations of the following graph. Then find the indegrees and
outdegrees for each orientation.

Theorem 7.1 (The First Theorem of Digraph Theory). If D is a digraph of size m with V (D) =
{v1 , v2 , . . . , vn }, then
Xn n
X
od(vi ) = id(vi ) = m.
i=1 i=1

Proof. Each (u, v) ∈ E(D) contributes to and .

Example 7.2. Consider the following digraph to find the following:

1. (directed) 2 − 4 walk of length 6:

2. (directed) trail of length 6:

3. (directed) path of length 4:

4. (directed) cycle of length 3:

5. (directed) circuit of length 5:

The underlying graph G of a digraph D is obtained by replacing each arc (u, v) or the pair
(u, v), (v, u) by the edge uv. Draw the underlying graph from the digraph above.

90
7.1 Strong Digraphs
Definition 7.2. A digraph D is connected (or weakly connected) if the underlying digraph of
D is connected. A digraph is strong (or strongly connected) if D contains both a u − v path
and a v − u path for every pair u, v of distinct vertices of D.

Example 7.3. Classify each of the digraphs below as connected, strong, or neither.

Theorem 7.2. If a digraph D contains a u − v walk of length `, then D contains a u − v path of

length at most `.

Proof Idea. Refer to Example 7.2

Theorem 7.3. A digraph D is strong if and only if D contains a closed spanning walk.

Proof.

91
Example 7.4. Two people A and B play a game on the complete graph K4 . These two people
alternate assigning a direction to an edge, beginning with A. It is the goal of A to produce a strong
orientation of K4 while it is the goal of B to stop A from doing this. If both players play perfectly,
who will win the game? On which turn will the game end?

Come up with a variation of this game and determine the outcome.

Definition 7.3. An Eulerian circuit in a (strong) digraph D is a circuit containing every arc of
D. If a D contains an Eulerian circuit, then D is an .

Example 7.5. Determine if the following graphs are Eulerian. Can you come up with a charac-
terization of Eulerian digraphs?

92
Theorem 7.4. A nontrivial connected digraph D is Eulerian if and only if od(v) = id(v) for every
vertex v of D.

Proof Idea.

Theorem 7.5. A nontrivial connected graph G has a strong orientation if and only if G contains
no bridges.

Proof.

93
7.2 Tournaments
Definition 7.4. A tournament is an orientation of a complete graph. Thus a tournament can
be defined as a digraph such that for every pair of distinct vertices u, v, exactly one of (u, v), (v, u) ∈
E(D). A tournament T models a round robin tournament. The vertices of T are the ,
and (u, v) is an arc in T if .

Example 7.6. The top 8 male tennis players compete in the ATP world finals at the end of each
year. The players are split into two groups of 4 where they complete a round robin stage. The
top 2 players in each group advance to a semifinal round. Last year one of the groups consisted of
players: Cilic, Djokovic, Isner, and Zverev.

1. Draw the corresponding tournament from the results shown below.

Winner Loser Score

Zverev Cilic 7-6(5), 7-6(1)
Djokovic Isner 6-4, 6-3
Djokovic Zverev 6-4, 6-1
Cilic Isner 6-7(2), 6-3, 6-4
Zverev Isner 7-6(5), 6-3
Djokovic Cilic 7-6(7), 6-2

2. Use the digraph to determine which two players advanced. In general, which team will win
the tournament?

**Zverev ends up beating Djokovic in the final round.

3. There are 4 non-isomorphic tournaments of order 4. Draw the other 3.

4. In each tournament of order 4 above, indicate the winning vertex (if one exists).

94
Definition 7.5. A tournament T is transitive if whenever (u, v) and (v, w) are arcs of T , then
(u, w) is also an arc of T .

Example 7.7. Which order 4 tournaments from above are transitive?

Example 7.8. Draw a transitive tournament of order 3 and a transitive of order 5. In general,
how many transitive tournaments of order n are there?

Remark 7.1. If T is a transitive tournament of order n and i is an integer with 0 ≤ i ≤ n − 1,

Theorem 7.6. A tournament T is transitive if and only if T contains no cycles.

Proof.

95
Definition 7.6. Let u, v be vertices of a digraph D. The distance d(u, ~ v) from u to v is the
~
length of a shortest u − v path in D. A u − v path of length d(u, v) is a u − v geodesic.
~ v) to be defined for every pair of vertices u, v ∈ V (D), then D must
Remark 7.2. In order for d(u,
be .
~ v) ≤ 2 for
Theorem 7.7. If u is a vertex of maximum outdegree in a tournament T , then d(u,
every vertex v of T .

Proof.

Theorem 7.8. Every tournament contains a Hamiltonian path.

Proof.

Theorem 7.9. A nontrivial tournament T is Hamiltonian if and only if T is strong.

Proof. See book.

96
7.3 ICE 8: Digraphs and Tournaments
1. How many orientations does a simple graph G of order n have?

2. Show that each of the graphs G1 and G2 is orientable by assigning a direction to each edge
so that the resulting digraph is strong.

3. Use the proof of the converse from Theorem 7.5 to find a strong orientation of the graph
below. Start by directing the cycle C 00 = (v1 , v2 , . . . , v10 , v1 ).

4. Is it possible to have a tournament with 6 teams where 3 of the teams win 3 games and the
other 3 teams win 2 games?

~ v) and d(v,
5. Let u and v be distinct vertices in a tournament such that d(u, ~ u) are defined.
~ ~
Show that d(u, v) 6= d(v, u).

97
6. Consider the tournament below to answer the following questions.

(a) Which team wins this tournament?

(b) Is this digraph strong? Explain your reasoning.

(c) Is this tournament transitive? Explain your reasoning.

(d) Is this digraph Eulerian? Explain your reasoning.

(e) Is this digraph Hamiltonian? Explain your reasoning.

98
8 Network Algorithms
Previously we studied the problem of finding a minimum spanning tree using Kruskal’s and Prim’s
Algorithms. In this chapter, we continue the discussion of algorithms for the solutions of the
important network optimization problems: shortest paths and maximum flows. By a network we
mean a graph with a non-negative integer k(e) assigned to each edge e. This integer will typically
represent the length or cost or capacity of an edge in various units.

8.1 Shortest Path

The following algorithm gives shortest paths from a given vertex a to all other vertices. Let k(e)
denoted the length of edge e. Let the variable m be a distance counter.

For increasing values of m, the algorithm labels vertices whose minimum distance from vertex a
is m. The label on each vertex v is of the form (−, −), where the first blank will be the previous
vertex on the shortest path from a to v, and the second blank will be the length of the shortest
path from a to v.

Shortest Path Algorithm (due to Dijkstra):

1. Set m = 1 and label vertex a with (−, 0).

2. Check each edge e = (p, q) from some labeled vertex p to some unlabeled vertex q. Suppose
p’s labels are (r, d(p)). If d(p) + k(e) = m, label q with (p, m).

3. If all vertices are not yet labeled, increment m by 1 and go to Step 2. Otherwise go to Step
4. If we are only interested in a shortest path to z, then we go to Step 4 when z is labeled.

4. For any vertex y, a shortest path from a to y has length d(y), the second label of y. Such a
path may be found by backtracking from y (using the first labels).

Example 8.1. Find the shortest path from a to z using the shortest path algorithm.

99
Example 8.2. Suppose you wish to find a shortest path from point N (Niagara Falls) to point R
(Reno) in the road network shown below. You apply the shortest path algorithm to get the labels
shown below. What is the shortest path?

Example 8.3. Use the shortest path algorithm to find the shortest path between vertex c and vertex
m in the previous example.

100
8.2 Network Flows
In this section, all networks are direced. We interpret the integer k(e) associated with edge e as a
capacity. Our goal is to maximize a flow between a source vertex and a sink vertex such that the
flow in each edge does not exceed that edge’s capacity.

Definition 8.1. An a − z flow in a directed network N is an integer-valued function f (e) defined

on each edge e, f (e) is the flow in e, together with a source vertex a and a sink vertex z satisfying
the following conditions:

1. 0 ≤ f (e) ≤ k(e)

2. For every non-source, non-sink vertex x,

X X
f (e) = f (e).
e∈In(x) e∈Out(x)

In(x), Out(x) denote the sets of edges directed into and out from vertex x, respectively.

3. f (e) = 0 if e ∈ In(a) or e ∈ Out(z).

Definition 8.2. The value of the a − z flow f (e), denoted |f |, equals the sum of the flow in edges
coming out of a.

Example 8.4. A sample single-source, single-sink network with flow.

Definition 8.3. Let (P, P ) denote the set of all edges (x, y) with vertex x ∈ P and y ∈ P . The set
(P, P ) is a cut. Further we call (P, P ) an a - z cut if a ∈ P and z ∈ P .
The capacity k(P, P ) of the cut (P, P ) is
X
k(P, P ) = k(e).
e∈(P,P )

Find the capacity of the cut displayed in Example 8.4.

101
Theorem 8.1. For any a − z flow f (e) in a network N , the flow out of a equals the flow into z.

Proof Idea.

Example 8.5. What is the maximum value of any flow in the network below?

Remark 8.1. All flow from a to z must cross each a − z cut. Thus, the combined capacity of the
edges in any a − z cut is an on how much flow can get from a to z.

Theorem 8.2. For any a − z flow f and any a − z cut (P, P ) in a network N , |f | ≤ k(P, P ).

Proof Idea.

102
Remark 8.2. The conservation of flow at a vertex (condition 2. from Definition 8.1) can be
extended to any set of vertices P not containing a or z:
X X
f (e) = f (e).
e∈(P ,P ) e∈(P,P )

That is, the flow into P equals the flow out of P .

Corollary 8.1. For any a − z flow f and any a − z cut (P, P ) in a network N , |f | = k(P, P ) if
and only if

1. For each e ∈ (P , P ), f (e) = 0.

2. For each e ∈ (P, P ), f (e) = k(e).

Further, when |f | = k(P, P ), f is a maximum flow and (P, P ) is an a − z cut of minimum capacity.

Proof.

Flow Maximizing Algorithm Attempt # 1

Definition 8.4. An a − z unit flow fL along an a − z path L is defined as fL (e) = 1 if e is in L

and 0 if e is not in L. We call an edge unsaturated whenever the present flow does not equal the
capacity. The slack s(e) of an edge e in flow f is s(e) = k(e) − f (e).

Remark 8.3. All normal flows can be decomposed into a sum of a − z unit flows. Then to build an
a − z flow, we can build up the flow as much as possible by successively adding a − z flows together
while being sure not to exceed any edge’s capacity.

Strategy: If s is the minimum slack among the edges in the a − z unit flow fL , we can add:

If f1 , f2 , . . . , fm are a − z unit flows, then:

103
Example 8.6. The capacities are shown for each arc in the network. Find a maximum flow.

Example 8.7. Switch positions d and e, and let’s try the same strategy this same network.

104
Augmenting Flow Algorithm

1. Give the vertex a the label (−, ∞).

2. Call the vertex being scanned p with second label ∆(p). Initially, p = a.

(a) Check each incoming edge e = (q, p). If f (e) > 0 and q is unlabeled, then label q with
[p, ∆(q)], where ∆(q) = min[∆(p), f (e)].
(b) Check each outgoing edge e = (p, q). If s(e) = k(e) − f (e) > 0 and q is unlabeled, then
label q with [p+ , ∆(q)], where ∆(q) = min[∆(p), s(e)].

3. If z has been labeled, go to Step 4. Otherwise choose another labeled vertex to be scanned
(which was not previously scanned) and go to Step 2. If there are no more labeled vertices to
scan, let P be the set of labeled vertices, and now (P, P ) is a saturated a − z cut. Moreover,
|f | = k(P, P ), and thus f is maximum.

4. Find an a−z chain K of slack edges by backtracking from z as in the shortest path algorithm.
Then an a − z flow chain fK along K of ∆(z) units is the desired augmenting flow. Increase
the flow in the edges of K by ∆(z) units (decrease flow if edge is backward directed in K).

Example 8.8. Apply the augmenting flow algorithm to the previous example.

Remark 8.4. If the algorithm is executed by a computer, it start with zero flow. When we imple-
ment the algorithm by hand, start with a flow by inspection to speed up the process.

105
Example 8.9. Apply the augmenting flow algorithm to the figure below given the initial flow. Find
a maximum a − z flow and a minimum capacity a − z cut.

106
Theorem 8.3. For any a − z flow f , a finite number of applications of the augmenting flow
algorithm yields a maximum flow. Moreover, if P is the set of vertices labeled during the final
(unsuccessful) application of the algorithm, then (P, P ) is a minimum a − z cut.

Corollary 8.2 (Max Flow - Min Cut Theorem). In any directed flow network, the value of a max-
imum a − z flow is equal to the capacity of a minimum a − z cut.

8.3 ICE 9: Network Algorithms

1. The network below shows the time it takes to walk along blocks in a city. The number on the
edge uv is the time (number of minutes) it takes to walk from intersection u to intersection
v.

Use the shortest path algorithm to find the

quickest path between intersections a and y.
How long will it take to walk?

2. Apply the augmenting flow algorithm to the flow in the figure below.

107
3. Consider the network of solid edges below. The vertices b, c, and d can supply up to 60,
40, and 40 units of flow, respectively. Vertices h, i, and j have flow demands of 50, 40, and
40 units, respetively. The sources could be factories and sinks warehouses. Or perhaps the
sources are oil refineries and the sinks oil-truck distribution centers.

(a) Can we meet all the demands? Find such a flow upon inspection, if possible.

(b) Now suppose the vertices b, c, d have unlimited supplies. How much flow can be sent to
the set {h, i, j}? Explain your new model.

108
9 Combinatorics
9.1 Basic Counting Principles
Combinatorics is a branch of mathematics dealing with counting different outcomes of a some
task called an activity with an observable outcome (called an experiment).

Principle 9.1 (Addition Principle). If there are r1 different objects in the first set, r2 different
objects in the second set, ..., and rm different objects in the mth set, and if the different sets are
disjoint, then the number of ways to select an object from one of the m sets is r1 + r2 + ... + rm

Principle 9.2 (Multiplication Principle). If a procedure can be broken into m successive (ordered)
stages, with n1 different outcomes in the first stage, n2 different outcomes in the second stage, ...,
and nm different outcomes in the mth stage. If the number of outcomes at each stage is indepen-
dent of the choices in previous stages and if the composite outcomes are all distinct, then the
total procedure has n1 × n2 × n3 × ... × nm different composite outcomes.

Principle 9.3 (Subtraction Principle). If S ⊂ U , then S c := S = U S = {x ∈ U : x ∈

/ S} is called
the complement of S in U. The number of objects in S, |S| = |U | − |S|

Principle 9.4 (Division Principle). Let S be a finite set partitioned into k parts such that each part
|S|
contains the same number of objects, say n, then the number of parts in the partition is k =
n

Example 9.1. What are the total possible outcomes when you flip a coin 5 times?

Example 9.2. Eva is deciding on what canned wet food she will have for dinner. The box has 3
different flavors of fish and 2 different flavors which are chicken based. How many choices does she
have?

Example 9.3. Baby Soraya has 10 baby bodysuits, 9 pairs of baby pants, 3 jackets, and 6 baby
headbands. How many different outfit combinations does Soraya have?

109
Example 9.4. Suppose you roll two dice, one red and one green.
a) How many different outcomes are there?

b) How many outcomes are there in which there are no doubles?

c) What is the probability that there are no doubles rolled?

Example 9.5. A soccer team is numbering their jerseys by ironing on 1’s, 2,’s, 6,’s, 7’s, and 9’s.
How many 2-digit numbers can be made if

1. repetition is allowed?

2. repetition is NOT allowed?

3. Only odd 2-digit numbers can be used and repetition is NOT allowed?

4. In which the number 9 must be used and repetition is NOT allowed?

Example 9.6. A computer password needs to consist of a string of 6 symbols taken from the
digits 0, 1, 2, ..., 9 and lowercase letters a, b, c, ..., z. How many computer passwords have a repeated
symbol?

110
Example 9.7. Suppose 11 people consisting of 5 men, 2 women, and 4 kids are going to a Lewis
Volleyball game and have 11 seats in a row.

a) How many ways there to seat them if there are no restrictions?

b) If the grown-ups sit together and the kids sit together?

Example 9.8. Given 10 different English books, six different French books, and four different
German books, (a) How many ways are there to select one book?

(b) How many ways are there to select three books, one of each language?

(c) How many ways are there to make a row of three books in which exactly one language is missing
(the order of the three books makes a difference)?

111
.

112
9.1.1 Ice -Counting Principles
2 sided!

1. How many 5 digit codes are there if

(a) repetition is not allowed?

(b) repetition is allowed?

(c) the code number must start with a 2 and end with a 9 and repetition is not allowed?

(d) The code number must be odd but cannot have any 5’s in it and repetition is not
allowed?

2. How many numbers of any length less than 500 can be formed from 2, 4, 7, and 9 if repetition
of digits is allowed?

113
3. Suppose we have 3 different colored chips, Red, Blue, and White.

(a) How many ways are there to grab 1 chip?

(b) How many ways are there to grab one chip and then a second chip without replacement?

(c) How many ways are there to grab a chip, then a second, then a third without replace-
ment?

(d) How many ways are there to grab 2 of the 3 chips at once?

(e) How many ways are there to grab all 3 chips at once?

(f) How many ways are there to put the chips in a circle?

114
9.1.2 Ice -Does Order Matter
For the following situations, determine where order matters.
1. Lewis has 3 soccer scholarships of $1,000. How many ways are there to award these after
they’ve narrowed the candidates down to 6 athletes?

2. Lewis has 3 soccer scholarships of $1,000, $2,000, and $3,000. How many ways are there to
award these after they’ve narrowed the candidates down to 6 athletes?

3. How many ways are there for 10 soccer teams to end the season in first, second, or third
place?

4. After being dealt 5 cards, how many ways can someone arrange the cards in their hands?

5. How many different groups of 4 can be formed from 6 cats?

6. How many different 5-card poker hands are there?

7. How many different 9 player batting line ups are possible on a softball team with 13 players
on the roster?

8. How many different outcomes are there when you flip 3 identical nickels at the same time?

9. How many different outcomes are there when you flip a penny 3 times in a row?

10. How many different 2 card blackjack hands are there (first card dealt down and second dealt
face up)?

11. How many ways are there to grab 3 marbles one at a time out of a bag with 6 red marbles,
5 green marbles, and 2 blue marbles?

12. How many ways are there to grab 3 marbles one at a time out of a bag with 6 red marbles, 5
green marbles, and 2 blue marbles so that the first is red, the second is green, and the third
is blue?

13. How many different 3 piece ski outfits could you come up with if you have 2 jackets, 3 hats,
and 4 pairs of gloves?

14. How many different words can be formed using all of the letters in BIKE? (nonsense words
are ok)?

115
.

116
9.2 Permutations and Combinations
Definition 9.1. A permutation of n distinct objects is an arrangement, or ordering, of the n
objects. An r-permutation of n distinct objects is an arrangement using r of the n objects, denoted
P (n, r) =
If r > n, P (n, r) =

P (n, 1) =

Example 9.9. What is P (n, n)?

Example 9.10. If there are 9 cats in a cutest cat competition, how many ways are there for them
to finish in 1st, 2nd, and third? (What if Archer is guaranteed to with first?)

Example 9.11. What is the probability that a 4-digit campus telephone number has one or more
repeated digits?

Example 9.12. Suppose Archer has 3 kitty friends over to play poker. How many ways are there
for the 4 to sit at Archer’s round poker table?

117
Definition 9.2 (Circular Permutations). To place n objects in a circle, there are
number of ways.

Definition 9.3. An r-combination of n distinct objects is an unordered selection, or subset, of r

out of the n objects, denoted C(n, r) and we say, “n choose r.”

We can derive the formula for C(n, r):

Example 9.13. From a standard deck of playing cards. How many 10-card hands are there?

Example 9.14. Suppose 11 people consisting of 5 men, 2 women, and 4 kids only have 6 tickets
for a Blackhawks game. How many ways are there to choose who gets to go to the game if

a) there are no restrictions? c) exactly 3 men get to go?

b) 2 men, 1 women, and 3 kids get to go? d) at least 2 kids get to go?

118
Theorem 9.1 (Permutations of Objects with some being identical). If there are n objects r1 of
type 1, r2 of type 2, ... , and rm of type m, where r1 + r2 + ... + rm = n, then the number of
arrangements of these n objects, denoted P (n; r1 , r2 , ..., rm ), is C(n, r1 ) · C(n − r1 , r2 ) · C(n − r1 −
n!
r2 , r3 ) · ...C(n − r1 − ...rm−1 , rm ) =
r1 ! · r2 ! · ...rm !

Example 9.15. How many different ways are there to select 6 cats from 3 different breads of cats
(say Siamese, the Archer line, and the Eva Line)?

Theorem 9.2. The number of selections with repetition of r objects chosen from n types of objects
is C(r + n − 1, r).

Example 9.16. How many ways are there to fill a box with a dozen doughnuts chosen from five
different varieties with the requirement that at least one doughnut of each variety is picked?

Example 9.17. How many different full house hands are there in a standard deck of cards?

119
.

120
9.2.1 Ice -Permutations and Combinations
1. Suppose you have a kickball team which has 12 kids, 7 boys and 5 girls.

(a) Suppose you have chosen 3 kids to rest while the others are playing. How many ways
are there to assign your remaining 9 players to either play the outfield (3 positions) or
the infield (6) where their actual positions are not important only infield or outfield?

(b) Suppose you have chosen 3 kids to play outfield (left, center, and right). How many
ways can you assign these 3 selected for these outfield positions?

(c) Suppose you have chosen 9 kids to play, 5 of which are girls. How many ways can you
assign the 9 players to the infield if exactly 3 of the boys must play infield?

(d) How many different 9 person batting line ups are there?

(e) How many different 9 person line ups are there if it must alternate first a girl, then a
boy, then a girl, etc.?

turn over!

121
2. From a group of 5 Maine Coon cats, 4 Siamese cats, and 3 Ragdoll cats, Eva needs to make
a committee of 6 cats. In how many ways can this be done with:

(a) there are no restrictions?

(b) 3 Maine Coons, 2 Siamese, and 1 Ragdoll?

(c) exactly 2 Ragdolls?

(d) at least 4 are Maine Coons?

3. How many different 2 pair hands are there in a standard deck of cards?

4. How many ways are there to distribute six different books among 13 children if no child gets
more than one book?
(a) C(13, 6)
(b) 136
(c) C(13, 6)6!
(d) None of the above.
(e) Yay! Dr. Harsy is back!

122
9.3 Pigeon Hole Principle
The Pigeon Hole Principle (sometimes called Dirichlet drawer principle or the shoebox principle)
basically says if a lot of pigeons fly into not as many pigeonholes, then at least one hole will have
at least two pigeons.

Theorem 9.3 (Pigeon Hole Principle: Simple Form). If n + 1 objects are distributed into n boxes,
then at least one box contains two or more objects.

Proof:

Example 9.18. Among 13 people there are 2 who have their birthdays in the same month.

Example 9.19. Suppose we have n married couples. How many of the 2n people must be selected
to guarantee that a married couple is selected?

Abstract Algebra Tie-In:

The Pigeon Hole Principle is used to prove the Chinese Remainder Theorem:

Let m, n be relatively prime9 positive integers and let a, b ∈ Z where 0 ≤ a ≤ m − 1 and

0 ≤ b ≤ n − 1. Then there is a positive x ∈ Z such that x can be written in the form x = pm + a
and x = qn + b for some p, q ∈ Z.

It’s proof uses a similar method to the following example:

9
Recall we say m, n are relatively prime if their greatest common divisor is 1

123
Example 9.20. Given m integers a1 , a2 , ..., am there exits k, l ∈ Z with 0 ≤ k < l ≤ m such
that ak+1 + ak+2 + ... + al is divisible by m. (That is, there exists consecutive a’s in the sequence
a1 , a2 , ..., am whose sum is divisible by m.)

Example: Let m = 7, and integers be 2,4,6,3,5,5, and 6.

124
9.4 ICE- More Counting and Pigeon Hole Principle
1. How many ways to select a subset of eight doughnuts from three types of doughnuts if at
most two doughnuts of the first type can be chosen?

2. How many ways are there to have a collection of eight fruits from a large pile of identical
oranges, apples, bananas, peaches, and pears if the collection should include exactly two
different kinds of fruits?

3. Show that if n + 1 integers are chosen from the set {1, 2, ..., 2n} then there are always two
which differ by 1.

125
126
10 Generating Functions
10.1 Generating Function Models
Generating functions can be used to solve combinatorial problems with special constraints in selec-
tion and arrangement problems with repetition.

Suppose ar is the number of ways to select r objects in a certain procedure. Then g(x) is a
for ar if

g(x) = a0 + a1 x + a2 x2 + ... + ar xr + ... + an xn

If g(x) has infinitely many terms, it is called a .

Recall from Calculus II,

       
n n n 2 n r n n
(x + 1) = 1 + x+ x + ... + x + ... + x
1 2 r n

Thus g(x) = (x + 1)n is the generating function for ar = C(n, r), the number of ways to select r
things from a set of n objects.

Example:
(1 + x)(1 + x)(1 + x) = 1 · 1 · 1 + 1 · 1 · x + 1x1 + 1xx + +x11 + x1x + xx1 + xxx

To find the coefficient of xr is the number of different products with r x’s and (3 − r) 1’s.
Example: Find x2 ’s coefficient:

Note we could replace 1 with x0 .

This is great when we are only counting one thing. But for this section, we will mostly be discussing
multiplying polynomial factors... fun stuff right?

Note that (1 + x)(1 + x)(1 + x) = 111 + 11x + 1x1 + 1xx + +x11 + x1x + xx1 + xxx
could be written as xe1 xe2 xe3 where 0 ≤ ei ≤ 1.

And (1 + x + x2 + x3 )3 could be written as xe1 xe2 xe3 where 0 ≤ ei ≤ 3.

127
Example: The coefficient of x5 in (1 + x + x2 )4 is the number of formal products xe1 xe2 xe3 xe4
where 0 ≤ ei ≤ 2 equaling x5 and this is the same as the number of integer solutions of

This is equivalent to asking how many ways to select 5 objects from a collection of 4 types with at
most 2 objects in each type (or distributing 5 identical objects into 4 distinct boxes with at most
3 objects in each box).

Example 10.1. Find the generating function for the number of ways to select r balls from 3 green,
3 white, 3 blue, and 3 gold balls.

Example 10.2. Find a generating function for the number of ways to select r cats from 5 Siamese,
5 Maine Coons, 3 Tabbies, and 3 Calicos.

128
Example 10.3. Find a generating function for the number of ways for Soraya to select r outfits
from 5 onesies, 4 pants, 4 jackets, 3 pairs of socks, and 3 headbands. Note this has the implied
constraint that there must be at least one of each type.

Example 10.4. Find a generating function for ar , the number of ways to distribute r identical
objects into five distinct boxes with an even number of objects not exceeding 10 in the first two boxes
and between three and five in the other boxes.

Note if there isn’t an upper bound, we have a power series.

129
.

130
10.1.1 ICE -Generating Function Models
1. Build a generating function for ar the number of r selections from

(a) 5 red, 5 black, and 4 white balls.

(b) Five jelly beans, five licorice sticks, eight lollipops with at least one of each type of
candy.

(c) Six types of light bulbs with an odd number of the first and second types. Hint: There
is no upper bound for the number of light bulbs.

2. Use a generating function for modeling the number of distributions of 16 chocolate bunny
rabbits into four Easter baskets with at least three rabbits in each basket. Which coefficient
do we want?

131
3. Find a generating function for ak , the number of k-combinations of n types of objects with
an even number of the first type, an odd number of the second type, and any amount of the
other types.

4. Find a generating function for the number of integers between 0 and 999,999 whose sum of
digits is r. Hint: 6 digits total with values between 0 and 9.

132
10.2 Calculating Coefficients of Generating Functions
Recall we can find coefficients of products of polynomials using

That is long and tedious so instead we are going to use algebra to rewrite products in a way so
that we can use Table 10.2 to determine coefficients of generating functions.

Step 1: Use algebra to rewrite your function to one of (or a product of) the following functions:
(1 + x)n , (1 − xm )n , or (1 − x)−n .

Step 2: Use Table 10.2 and polynomial multiplication to determine the desired coefficient.

Figure 16: Table from Tucker’s Applied Combinatorics

133
Example 10.5. Find the coefficient of x16 in (x2 + x3 + x4 + ...)5 . What is the coefficient of xr ?
What does (x2 + x3 + x4 + ...)5 represent combinatorially?

Example 10.6. Use generating functions to find the number of ways to collect $15 from 20 distinct
people if each of the first 19 people can give a dollar (or nothing) and the twentieth person can give
either $1 or $5 (or nothing).

134
Example 10.7. How many ways are there to select 25 toys from seven types of toys with between
two and six of each type?

135
Example 10.8. How many ways are there to distribute 25 identical balls into seven distinct boxes
if the first box can have no more than 10 balls but any number can go into each of the other six
boxes?

136
10.2.1 Coefficients of Generating Functions
1. Find the Coefficient of x11 in

(a) x2 (1 − x)−10

x2 − 3x
(b)
(1 − x)4

(1 − x2 )5
(c)
(1 − x)5

2. Find the coefficient of x10 in (1 + x + x2 + x3 + ...)n .

137
3. Use generating functions to find the number of ways to select 10 balls from a large pile of red,
white, and blue balls if the selection has at least two balls of each color. State the generating
function.

4. Use generating functions to find the number of ways to select 10 balls from a large pile of
red, white, and blue balls if the selection has at most two red balls. State the generating
function.

138
5. Find the coefficient of xr in (x5 + x6 + x7 + ...)8 .

6. Find the coefficient of x32 in (x3 + x4 + x5 + x6 + x7 )7 .

139