Contents

1 Introduction 4
1.1 Pointless module? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 The pedantic mathematician . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Disclaimer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.4 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 Logic 5
2.1 Truth tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Universe of Discourse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3 Numbers, numbers, all kinds of numbers 8

3.1 The naturals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.2 The integers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.3 The rationals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.4 The irrationals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.5 The reals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.6 The complex numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1
4 Domino-like proof, and the FUNda MENTAL Theorem of Arithmetic 9

4.1 Primes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2 The Well-Ordering Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.1 Proving stuff with the WOP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.3 Proof by Induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

4.3.1 A slight variant of the Principle of Induction . . . . . . . . . . . . . . . . . . . . . . . . . 11

4.4 The FUNda MENTAL Theorem of Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.5 HCF and LCM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4.6 Euclidean Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

4.6.1 The extended Euclidean Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

5 Integers and Modular Arithmetic 14

5.1 Subgroups of Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.2 The wonders of Modular Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

5.3 Congruence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.4 A last thing you should know . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.5 Some Fun Stuff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.5.1 Multiplicative Inverse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

5.5.2 Gaussian Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

6 Rationality + Irrationality = Reality 20

6.1 Decimal Expansions,to infinity.......and beyond- Buzz Lightyear . . . . . . . . . . . . . . . . . . 21

7 Going Beyond Reality: The Complex Realm 22

7.1 The Complex Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
7.2 The Conjugate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

7.3 The Argument and Absolute Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

7.3.1 Evaluating the Argument and the Absolute Value . . . . . . . . . . . . . . . . . . . . . . 25
7.3.2 Some Properties of |z | . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

7.4 Polar Co-ordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

7.5 The Exponential Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.6 Roots of numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
7.6.1 Using De Moivres Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

7.6.2 Using the Exponential Form of Complex Numbers . . . . . . . . . . . . . . . . . . . . . . 30

7.6.3 nth Roots of Unity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

8 Functions 32
8.0.4 Injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
8.0.5 Surjection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

8.0.6 Bijection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
8.1 Some Interesting Stuff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

2
 8.1.1 Counting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

8.1.2 The Pigeonhole Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

8.1.3 Composite Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
8.2 Inverses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
8.3 Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

9 Relations 43

9.1 Equivalence Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

9.2 Order Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

10 Polynomials 45
10.1 Polynomial Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
10.1.1 Polynomial Addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

10.1.2 Polynomial Multipication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

10.1.3 Polynomial Division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.2 Prime Polynomials, HCF and LCM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

10.2.1 Polynomial LCM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

10.3 The Remainder Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
10.4 Roots of Polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
10.4.1 Algebraic Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

11 Counting 59

11.1 Countable and uncountable sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

11.2 Different Infinities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

12 Foundations 66
12.1 Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
12.2 Subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

12.3 Sets and Venn Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

12.4 Mathematical Induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
12.5 Remainder Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

12.6 Pascals Triangle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

12.7 Injections and Surjections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
12.8 Counting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
12.9 Samir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

13 Analysis 69

13.1 True / False Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

13.2 Give an Example Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
13.3 Toy Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

14 Conclusion 76

3
1 Introduction

1.1 Pointless module?

Foundations is often considered by many first year MORSE students as the most difficult module. The aim of
this guide is to try make it more fun (and by fun I mean fun) so that it becomes easier for you. The notes will
try create a link with modules/materials that you will come across later in your first year so that you dont feel
that what you are doing is pointless. Often, how MORSE students feel about foundations determines how they
feel about mathematics in general.

Many will be discouraged by this module while some will be inspired. The best thing to do concerning Foun-
dations is to keep in touch with the material by attempting and discussing the exercise sheets. And yes,
unfortunately youll have to study a bit over christmas for the January exams.

1.2 The pedantic mathematician

You will find during the course of your degree that mathematicians are the most pedantic academic creatures
that you will ever come across. The weirdest part of it all is how they set an injection between Mathematics
and the set of arts.(Dont worry, by the end of these notes, you will get the joke).

During your first year, you will often find yourselves proving obvious things like:

1. For any number , prove that 0   0 (MA106)

2. Any odd degree polynomial crosses the x-axis at least once (MA131)

You will also be proving more interesting stuff like lHopitals Rule (MA131), De Morgans Law (ST115), and
for those taking (Number Theory MA246), the Chinese Remainder Theorem. The main reason for making you
prove stuff is so that you develop a mentality of questioning the validity of statements. It is important to always
look for a reason for the way things are. In my opinion, studying rigorous maths helps you do that.

If all goes well, you will, by the end of the second term, be able to understand most Maths jokes (often a
worrying sign of dementia). I wish all of you the best of luck in the marvelous world of Mathematics.

1.3 Disclaimer

This guide might not be 100% accurate although it follows the syllabus closely. Its purpose is to help deal with
ideas that may not be clear in the formal notes. It should not be the sole point of reference for the module,
but should be used alongside the lectures and the formal lecture notes. Moreover it is not possible to pass the
exams just by reading this guide. The weekly exercises need to be attempted to gain further understanding of
the module.

1.4 Acknowledgements

I wish to thank all those who proofread this guide. Without them, this guide would be a gammatical mess (that
includes mathematical grammar).

Desislava Ivanova
Ryan Balgobin
Mikoai Kasprzak
Kevin Chan
Gaurav Maheshwari
Dmitrijs Murins

I would also like to thank Keegan Kang for letting us use his compilation of practice questions and Iain Carson,
who edited this compilation.

To Those Who Believe In The Elegance Of The Universe

4
2 Logic

In this section, we will focus on logical thinking. Logic in mathematics is very important, not only because it
allows you to work through the subject, but also because most mathematics students inherit a very logical way
of communicating their ideas. To get a broader view of logic, you can have a look at Metamagical Themas:
Questing For The Essence Of Mind And Pattern by Douglas Hofstadter. It can also be obtained from the
library.

2.1 Truth tables

Truthtables are really annoying, but they help a lot when you have to proof logical relations. The first step is
knowing the truthtables for a few basic relations. These are:

_
a _ b
T T T
T T F
F T T
F F F
The symbols a and b represent 2 statements. Then, T represents true and F represents false. So, a
truth table analyses all possible permutations of the truth or falsity of statements when the statements
are related by a logical statement. Now the T and F below the _ sign shows the validity of the logical
argument. What the table shows is this: the statement a _ b is true when either a or b or both are true
and a _ b is false when both a and b are false. The symbol _ can be interpreted as or.
^
a ^ b
T T T
T F F
F F T
F F F
The truthtable says that the statement a ^ b is true only when both a and b are true. _ can be interpreted
as and.

a b
T T T
T F F
F T T
F T F
This is perhaps the weirdest truthtable. The implication sign shows the direction of causation in a
logical argument. For example, let us consider the logical relation between death and a bullet in the head.
Our universe of discourse (i.e. the world in which the relationship is being discussed) is a time lag of 24
hours in real life. Let a represent getting a bullet in the head and b represent dying. Getting a bullet in
the head necessarily results in death. So, saying that getting a bullet in the head does not result in death
is false, hence the second line in the truth table. Now for the last two lines, consider not getting a bullet
in the head. A person may still die as death can be the result of many things, such as diseases. Moreover,
the person may not die as the causes of death may not operate in the 24 hours. So the last two lines hold
the true value as they are possibilities in our universe of concourse.

a b
T T T
T F F
F F T
F T F
The also shows the direction of causation. The difference between and is that the relates two
statements that cause the other to happen. That is, a happens only if b happens and vice versa. For
example, the statements a and b can be 5x 3  23 and x  4 respectively.

5

a
T F
F T
a can be interpreted as not a. So whatever the truth value of a, a will be the opposite.

It is very important to master the use of these logical relations as you will use them a lot in proofs. For example,
often you will have to prove statements of the type a b. It can sometimes be easier to prove p bq p aq,
which is equivalent to proving a b.

Example Show that p bq p aq is the same as a b

Solution: First, fill out the truth values for the individual statements.
a b b a
T T T F T F
T F F T F F
F T T F T T
F T F T T T
Next, consider the truth values of the whole individual statements. The values of the statement
are the values below .
ab b a
T T T
F T F
T T T
T T T
We just showed that whatever the truth values of a and b, a b and b a are equivalent.
When such a scenario occurs, we say that the equivalence relation is a tautology, meaning that
the whole statement is logically true. The word tautology refers to something that is intuitively
and obviously true. A tautology in the english language would be something that does not require
justification. All mathematical statements that are true are tautologies in that same sense. The
difference is that, as human beings who unfortunately have a limited processing power, we need
to formerly and rigorously establish these tautologies. This is why, mathematicians are often
considered as pedantic.

Exercises
1. Show that the following relations hold (hint: start assigning truth values to the innermost
bracket and work your way out)

(i) pa ^ pa bqq b
(ii) ppa bq ^ pb cqq pa cq
2. One of the following is equivalent to a b. Which one? Use a truth table to justify your
answer. (Jan 2011 exam)
(i) a ^ b (ii) p aq _ b (iii) p bq _ a (iv) p aq ^ b

3. One of the following is equivalent to pa bq. Which one? Use a truth table to justify your
answer. (Jan 2010 exam)
(i) a ^ p bq (ii) b a (iii) p aq _ b (iv) p bq a
4. One of the following is equivalent to pp aq bq. Which one? Use a truth table to justify
your answer. (Jan 2009 exam)
(i) a b (ii) a _ b (iii) p bq _ p aq (iv) p aq ^ p bq
5. One of the following is equivalent to a p bq. Which one? Use a truth table to ustify your
answer. (Jan 2008 exam)
(i) pa _ p bqq (ii) a _ b (iii) p bq _ p aq (iv) p aq _ b

2.2 Universe of Discourse

In mathematics, it is very important to be precise about the objects we are talking of. This is why, it is
important to define clearly the realm in which our objects are being discussed. If you do not do so, there might
be a lot of confusion. One example of this can be found in the formal lecture guide. You will come across many
such confusing objects in your first year. One such would be about the divergent series 1 2 3 4 .....
(MA131). According to Srinivasa Ramanujan, this series converges to

1 . There needs to be some abstract
12

6
definition of the universe of concourse for this to be true. Another interesting manipulation of universe of
concourse is a paddock (completely unrelated to horses). It has been defined as a set that satisfies all 4 addition
axioms and all 4 multiplication axioms of number systems but in which 1  0 (MA106). You actually come
across a paddock in one of your assignments and so, we will not discuss them further.
There are many more such examples and it is thus important that you are accurate in whatever statements you
make. This will help you a lot in developing your mathematical and general thinking skill.

Theorem 2.1 (De Morgans Law). For any well defined sets A and B, we have

(i) pA X B qc  Ac Y B c or in logic language, p pa ^ bqq p aq _ p bq

(ii) pA Y B qc  Ac X B c or in logic language p pa _ bqq p aq ^ p bq
Proof of De Morgans Law. There are three ways to prove the above statement

Using Venn Diagrams

Using truthtables

Using a rigorous usage of set theory

The first two ways are easy and we will not prove it this way. We shall only prove the first statement and the
second is left as exercise to the reader (annoying, I know, but it is a good exercise!).
To show that 2 sets P and Q are equal, we need to show 2 things:

P Q that is, px P P q px P Qq
Q P that is, px P Qq px P P q

So we do the same for our 2 sets, pA X B qc and pAc Y B c q

1) Assume x P pA X B qc . Then x R A X B. So, either x R A or x R B x P Ac or x P B c . This is like saying

x P pAc q Y pB c q
2) Assume x P pAc q Y pB c q. Then x P Ac or x P B c Rightarrow x R A or x R B x R A X B x P pA X B qc
We just showed that pA X B qc Ac Y B c and Ac Y B c pA X B qc , thus proving De Morgans law.
You will make use of De Morgans Law in ST115. There, you will encounter the extended version of De Morgans
Law.
Theorem 2.2 (Extended De Morgans Law). Consider a finite number of sets P1 , P2 , ......, Pn , then

n
n
(i) pP1 X P2 X .... X Pn qc  P1c Y P2c Y .... Y Pnc . A better way to write this is p Pi qc  Pic

i 1 
i 1

n
n
(ii) pP1 Y P2 Y .... Y Pn qc  P1c X P2c X .... X Pnc . A better way to write this is p Pi q c  Pic

i 1 
i 1

Exercises
1. Prove the simple version of De Morgans Law using truth tables.
2. Prove the extended De Morgans Law. (Hint: The proof structure should be like the one just
given)

7
3 Numbers, numbers, all kinds of numbers

3.1 The naturals

The natural numbers are numbers as we all know them. They are the ones that were originally invented for the
most basic mathematics, i.e. counting. Those who invented them probably used them to have a more concrete
measure of quantity. So instead of saying I have few apples and you have many (highly subjective right?),
they would start saying, I have 100 apples and you have 10. As you will later see in this module and in ST115,
the naturals are a really powerful counting tool. They are actually used to know the cardinality of other number
systems.

The set of natural numbers is denoted by N

N  t1, 2, 3, 4, 5, 6, ...u (You will find those dots a lot now and when they are placed like this, it means that the
sequence is never ending)

It is to be noted that different persons adopt different conventions about whether to include zero or not in
the naturals. It is sometimes out of caprice that mathematicians adopt their own conventions concerning this.
However, you will find that certain branches of mathematics require the adoption of a specific convention. (By
require, I mean, you have less to write if you adopt a certain convention in your explanations)

The naturals are closed under addition and muliplication only, meaning that the sum or multiplication of 2
natural numbers always gives a natural number.

3.2 The integers

The integers are a bit like natural numbers, except that they have a sign attached to them. The set of integers
is denoted by Z. By definition, Z always contains zero.

Z  t...
3,
2,
1, 0, 1, 2, 3, ...u

The integers are closed under addition, multiplication, subtraction but not division.

3.3 The rationals

The concept of rationality is very interesting. The set of rational numbers, denoted by Q are numbers that can
be expressed as fractions. In formal mathematical language,
Q  tq P Q : q  for some m, n P Zu
m
n

The rationals are what we call a field as Q is closed under addition, subtraction, multiplication and division.

3.4 The irrationals

There are those weird numbers that cannot be expressed as fractions. They form the set of irrational numbers.
? the complement of Q. It is denoted by Qc or ZzQ.
The set of irrationals is
Notable members are 2, e and

3.5 The reals

The real numbers is the set of all the above numbers. A more hype definition would be the set of square roots.
It is denoted by R. One interesting feature of the reals is that they are uncountably infinite (dont worry, youll
learn about that soon enough).

8
3.6 The complex numbers

What happens when?you want to take the square root of
1? Well you are left with the imaginary number i.
By definition then,
1  i.

Complex numbers consist of 2 parts, the imaginary part, and the real part. A complex number is then expressed
as ai b where a, b P Z. For such a complex number z, =pz q  a, <pz q  b. The set is denoted by C

What we then have is N Z Q R C

Note that Q, R, C are all fields.

4 Domino-like proof, and the FUNda MENTAL Theorem of Arith-

metic
Big words: PROOF BY INDUCTION and FUNDAMENTAL THEOREM OF ARITHMETIC. You will come
across many fundamental theorems in your wonderful mathematical journey. They really are fundamental in
the sense that they are obvious to us in an intuitive manner but are more or less annoying when it comes to
justification. In the run-up to these 2 words, we need to learn a bit about prime numbers

4.1 Primes

Definition 4.1. A prime number is a natural number which can be divided by exactly two positive numbers,
i.e, one and itself. In other words, if p is a prime number, its natural factors are 1 and p

Prime numbers have many interesting properties such as their infinite number, which we are going to study.
There are many other statements about primes that need to be proven such as the claim by the German math-
ematician Goldbach in 1742 that every even integer greater than 2 can be written as a sum of prime numbers.
We do know that every integer greater than 2 is a product of 2 prime numbers but the above claim, known as
Goldbachs conjecture has not yet be proved (not even by the great Euler, to whom Goldbach made the original
claim).

Another interesting property of prime numbers is that there is no known formula or algorithm that can be
used to list all the primes. However here is an ingenious way of showing that no list of primes can be exhaus-
tive.
Suppose you have a list of n prime numbers, p1 , p2 , ....., pn where n P N. Define a new number,
pn 1  pp1  p2  p3  ....  pn q 1
Clearly, pn 1 is not divisible by any of the the primes found in the previous list, as it gives a remainder of 1
on division by any of them. So it must be divisible by a prime that is not in the list.
We just proved there are infinitely many primes.

4.2 The Well-Ordering Principle

Definition 4.2. Every non-empty subset of N has a least element

The Well Ordering Principle (WOP) is quite intuitive in the sense that the set N, by definition is bounded
below, that is, there always exists a number in the set of naturals which is smaller than or equal to any other
natural. Any subset of the naturals thus will have a least element. The interesting thing about the Well-Ordeing
Principle is that it can be used in proofs.

4.2.1 Proving stuff with the WOP

The Well-Ordering Principle is very useful when you have to prove something about natural numbers. Care
must be taken, though, so as not to use WOP where induction should be used and vice versa. Generally,
Induction is used when there is a law that applies to an ordered pattern of naturals. The WOP is used to prove
existence of a number and sometimes, its uniqueness. Often, contradiction is used along side the WOP.
Whatever the proof is, if you are using the WOP, then you should follow these steps:

9
1) Define a set of natural numbers. Labelling it with a letter, say S can help to save ink/lead. Usually, the set
should be that which contains the natural number whose existence/uniqueness you are trying to prove.
2) Then, you say, by the WOP, as the set is not empty (here, you will have to justify this assumption), there
exists a least element. The letter for that least element is usually r0 . the subscript 0 is used a lot for least
element or fundamental numbers.
3) Then, you assume the opposite of what has to be proven. You work your way through the maths and end
up with a contradiction, which should take the form of r0 is not the smallest element of S.
4) In the end, you will have proven that r0 is the only element of the set.

The steps seem easy enough to follow. However, defining the set and working the maths often prove to be very
challenging. It is important that you memorise these steps so that you know what you are looking for in your
proof and hence, do not get lost while doing it.

Example
Proposition 4.1 (Division with Remainder). If m and n are natural numbers with n 0,
then there exists natural numbers q and r with 0 r n m such that m  qn r

Proof. What we want to prove here is the existence of the number r, or rather, we want to prove
that r has the stated properties.
1) We define our set S  tr P N : r  m
qn for some q P Nu
What this means is that, we fix numbers n and m and for a feasible set of natural numbers,
q, we build a natural number r. Our set S contains all those rs. For example, we may choose
n  2 and m  9. p can take values 1, 2, 3, 4 and the corresponding r values would be 7, 5, 3, 1.
We see that for r  1, r n. We have that this holds for all natural numbers m, and n
2) It is obvious that S is not empty and so, by WOP, S has a least element r0 given by r0  m
q0 n
for some q0 P N.

3) Now we assume that r0 n. Then, r0 can be written as n r1 where r1 P N. Here, r1 takes

the form m
pq0 dqn and so r1 P S.
Clearly, r1 r0 and so, r0 is not the least element of S. This is a contradiction.
4) We have shown that the least element of the set S is always less than n. We thus proved the
statement.

Exercise: Prove the following lemma using the WOP

Lemma 4.1. Every number greater than 1 can be divided by some prime number

4.3 Proof by Induction

You will use proofs by induction to prove a statement about an infinite number of consecutive naturals. Imagine
a row of standing dominoes. If you push the first one, it will fall over to hit the next one. Of course the next
one must be in range for that to happen. So if these two conditions are satisfied, it is possible to make all of
the dominoes fall just by touching the first one. In a proof by induction,

you need to show that the statement applies for the first natural number in your set (just like nudging
the first domino).
Then you show that if the statement is true for a number k, then it is true for the next number k 1
(testing the distance between 2 dominos).

If you manage to show these 2 facts, you showed that the statement applies for all the naturals in your set.
(The set can be even naturals, naturals starting from a number n0 or just N

Unfortunately, presentation is important when using this technique. We will label the statement about a number
n as P pnq. The steps for a well presented proof is as follows:

10
 Base Case
We need to show that the statement applies to the first natural number in the set, n0 . It is usually very
easy and will generally involve replacing values into an equation. Here you need to write:
P p0q is true
once you are done with your logical argument
Inductive step
Here we show that the fact that the statement is true for an arbitrary number, n implies that it is also
true for the next number n 1. We need to assume that it is true for n (inductive assumption) and using
that assumption, we will show that it is true for n 1. You write:
P pnq is true P pn 1q is true.

n
Example Show that k  12 npn 1q

k 1

n
Solution: Let P pnq be the statement k  12 npn 1q

k 1

Base Case
Replacing n  1 in the RHS yields 1. Similarly, the LHS yields 1. So, P p1q is true
Inductive Case
Assume P pnq is true.
n1
n
k p kq pn 1q

k 1 
k 1

 12 npn 1q pn 1q from the inductive assumption

(1)
 pn 1qp1 n2 q
 12 pn 1qpn 2q

The last line is exactly what we expect for P pn 1q. We therefore say that P pn 1q is true.

Our Proof is now complete.

4.3.1 A slight variant of the Principle of Induction

Corollary 4.1 (Principle of Induction II). Suppose T N and

0PT
for every natural number, n t0, 1, 2, 3, ..., n
1u T nPT
Then T N
This is very much like the previous version of the Principle of Induction, but the only difference is that, here
we assume that the statement is true for a set of numbers up to n
1. What the POI II is actually saying, is
that, if you manage to show that the truth of a statement for a natural number is implied by the truth of the
statement for all preceding natural numbers, then the statement is true for all natural numbers.

4.4 The FUNda MENTAL Theorem of Arithmetic

Theorem 4.1 (Fundamental Principle of Arithmetic). It says the following about natural numbers:

1) Every natural number greater than 1 can be written as a product of prime numbers
2) The prime factorisation of any number is unique up to the ordering of the factors

You can see how obvious this theorem is. However, as students, you are required to learn the proof. We basically
have to prove 2 things, existence and uniqueness. The first is easy and is left as an exercise. You should use
the POI II to do it. If you find using the POI II difficult, then follow the example below.

11
Proof. Proof of uniqueness of prime factorisation

Base case: The first prime number is 2 and it is obvious that it has a unique prime factorisation. The
same applies for all numbers that come after it. We can show, with enough effort that this s true up to a
number n
1
Inductive case: We need to show that is is true for the next number n, using the assumption we made
about the preceeding numbers.
(i) We assume first that there are 2 prime factorisation for the number n, i.e.
n  p1 p2 p3 ....ps  q1 q2 q3 ....qt .
We order the prime factors such that p1 p2 ... ps and q1 q2 ... qt
(ii) If the prime factorisation is unique, then, we need to have pi  qi @i. We will first assume that
p1  q1 . We then consider only p1 q1 and we will extend the argument to p1 q1 by symmetry.
(iii) This is the crucial part, as it is the key to the proof. Often, you will find that complicated proofs
have a key part that requires memorising. Those who came up with this idea did so after years of
work, and you will not be expected to come up with such stuff in your exams.
p1 pp2 p3 ....ps
q2 q3 ....qt q  p1 p2 p3 ....ps
p1 q2 q3 ....qt
p1 pp2 p3 ....ps
q2 q3 ....qt q  q1 q2 q3 ....qt
p1 q2 q3 ....qt
p1 pp2 p3 ....ps
q2 q3 ....qt q  pq1
p1 qq2 q3 ....qt p1q
As q1
p1 P N, we can factorise it.
pq1
p1 q  r1 r2 ...ru p2q
Combining (1) and (2), we get,
p1 pp2 p3 ....ps
q2 q3 ....qt q  r1 r2 ...ru q2 q3 ....qt p3q

(iv) Now we analyse what we got. We know from (3) that p1 divides the RHS, otherwise, we would find
ourselves with a fraction. So either p1 is one of the qi or p1 is one of the ri .
The first case is impossible as we already assumed that p1 q1 q2 ... qt . So p1 must be an ri .
What we have then is p1  rm where 1 m u.
r1 ...rm ...ru  pq1
p1 q rm  q1 . This is impossible as q1 is prime and the only factor it has is
itself .
(v) By symmetry, we can show that p1 q1 .
(vi) The only possibility left is p1  q1 . (BY the axiom of trichotomy, this must be true). The same
procedure can be extended to show that q2  p2 , q3  p3 , ...., ps  qt .

4.5 HCF and LCM

You must be telling yourself that this feels like going back to primary school. Well, unfortunately, it is not that
simple. Many of you will be moaning about how complicated this has become. You should remember however,
that the principle aim of this mocule is to introduce you to the rigours of maths.
Definition 4.3 (Highest Common Factor and Least Common Multiple). Let m and n be natural
numbers. Then

The largest natural number to divide both m and n is called the highest common factor and is denoted
by hcf(m,n)
The smallest natural number that can be divided by both m and n is called the least common multiple
and is denoted by lcm(m,n).

Note that the HCF is sometimes called the greatest common divisor, gcdpm, nq.
We shall employ the following notation in prime factorisation:
n  2i2  3i3  5i5  ......  prpr where each powers of the primes are natural numbers (including zero).
i

Also, @n 0 we shall use these two conventions:

1) hcfpn, 0q  n
2) lcmpn, 0q  0

12
Below are some interesting facts about HCFs and LCMs. It is left to you to prove them. Odds are they will be
proved in the formal lecture guide.

A Suppose
n  2i2  3i3  5i5  ......  prpr
i

m  2j2  3j3  5j5  ......  prpr , then,

j

hcfpm, nq  2minti2 ,j2 u  3minti3 ,j3 u  5minti5 ,j5 u  ......  pmin

r
tipr ,jpr u
B hcfpm, nqlcmpm, nq  m  n
C If g is any common factor of m and n, then g divides hcfpm, nq
D If l is any common multiple of m and n, then lcmpm, nq divides l

E If m  qn r, then hcfpm, nq hcfpr, nq

4.6 Euclidean Algorithm

So far, weve had a look at division with remainder and at a few interesting properties of HCFs and LCMs.
These actually help us in the Euclidean algorithm, which is used to find the HCF of large numbers. You should
practise the Euclidean algorithm as much as possible as it is the most scoring exercise in the exams. Sadly, it
is the one where, without enough pratice, you will unnecessarily lose time.
The basis of the Euclidean algorithm is to break down the two large numbers until they cannot be broken down
any further. To understand how to use it, it is best that you are introduced through examples.

Example Find the HCF of 870 and 750

870  1  750 120
From what we already know, hcfp870, 750q hcfp750, 120q. We then break down 750 and 120, and repeat until
we no longer obtain remainders.
750  6  120 30
120  4  30 0
hcfp870, 750q hcfp750, 120q hcfp120, 30q hcfp30, 0q  30

There is no better way to learn the Euclidean algorithm than practice.

Example Find the HCF and LCM of 1066 and 4360

4360  4  1066 96
1066  11  96 10
96  9  10 6
10  1  6 4
614 2
422 0
Hence, hcfp1066, 4360q  2
hcfp1066, 4360q lcmp1066, 4360q  1066  4360
lcmp1066, 4360q  2323880

4.6.1 The extended Euclidean Algorithm

The Euclidean algorithm requires some notion of negative numbers, which all of you should have.
Corollary 4.2. Let m, n P Z, then Da, b P Z such that hcfpm, nq  am bn

We shall prove this in the next section. However, we already have the necessary tools to develop a method to
find the numbers a and b. The idea is to work backwards using the Euclidean algorithm.

Example Find integers a and b s.t hcfpm, nq  am bn when n  1066, m  4360

We already know that hcfp1066, 4360q  2. So as not to confuse numbers, we will denote the HCF by h. Then we
start with the last line of the Euclidean algorithm and express h in terms of each fragmentation of 4360 and 1066.

13
 Extended Euclidean Algorithm Step Euclidean Alrgorithm Origin Comment
h2
h16
14 614 2 Express 2 in terms of 6 and 4
10  1  6 4
h  2  6
1  10
Express 4 in terms of 10 and 6
and replace in previous equation
96  9  10 6
h  2  96
19  10
Express 6 in terms of 10 and 96
and replace in previous equation
1066  11  96 10
h  211  96
19  1066
Express 10 in terms of 96 and 1066
and replace in previous equation
4360  4  1066 96
h  211  4360
863  1066
Express 96 in terms of 1066 and 4360
and replace in previous equation
We worked backwards until the first step so that we got an expression in the form 4360a 1066b for hcfp1066, 4360q.
Here, a  211, b  863

A good question to ask is whether the integers a and b in the equation hcfpm, nq  am bn are unique. It turns
out that they are not. The proof for this statement is usually not examinable. However, you may be asked to
find a second such pair of integers.
Proposition 4.2. Let hcfpm, nq  a1 m b1 n and hcfpm, nq  a2 m b2 n, then,
a2  a1
n and
b2  b1 m

5 Integers and Modular Arithmetic

We have first to introduce some basic notion of subgroups of Z. Then, we will look at the more interesting
notion of Modular Arithmetic.

5.1 Subgroups of Z

Definition 5.1 (Subgroup of Z). A set S is a subgroup of Z if

S H
0PS

SZ
if x, y P S, then, px y q P S and px
y q P S
Proposition 5.1. Let S be a subgroup of Z and x P S, then @n P Z, nx P S

Proof. Exercise
Proposition 5.2. Let S be a subgroup of Z, then Dg P N s.t S  gZ  tgn : n P Zu
The above proposition is a stronger version of proposition 1.1. Proposition 1.1 says that all the integer multiples
of an element of a subgroup belong to that subgroup, whereas proposition 1.2 says that a subgroup can be created
by putting all the integer multiples of a natural number into a set.

Proof of Proposition 1.2. The case where the only member of S is zero is trivial. Otherwise, we know that
S has both positive and negative numbers, as @a P S, p0
aq P S. We denote S as the set of positive members
of S and S
as the set of negative numbers in S.
By the WOP, S has a smallest element, g. What we need to prove now, is that S  gZ, which is done by
proving S gZ and gZ S.
Our sets S and S
does not contain 0 but note that 0 P S and 0 P gZ.

gZ S
We already proved that, g P S gn P S @n P Z. This means that all the elements of gZ are in S. Thus
gZ S.

14
 S gZ This part is a bit more tricky. We need to show that any element found in S is necessarily a
member of gZ, in logic terms, we need x P S x P gZ.
Assume that q P S and q R gZ. Then, we know that q  gn r where n, r P N, 0 r g. From the
properties of subgroups, we know that q
gn, hence r P S . Then, g is not the smallest member of S .

We actually proved x P S x P gZ. We still need to show x R S
x R gZ.
Again, assume that q P S and q R gZ. Then, we know that q 
gn r where n, r P N, 0 r g. We
come to the same conclusion about g not being the smallest element of S .

We are moving closer to justifying what we did in the Extended Euclidean Algorithm.

Proposition 5.3. Let G1 and G2 be subgroups of Z, then

(i) G1 X G2 is a subgroup of Z
(ii) G1 G2 given by tx1 x2 where x1 P G1 , x2 P G2 u is a subgroup of Z
Proof. We need to show that each of these two sets are closed under addition

(i) Let x, y P G1 X G2 . Then we know that x, y P G1 and x, y P G2 . This means that @m, n P Z, mx ny P
G1 , mx ny P G2 . This is like saying mx ny P G1 X G2 .
(ii) Let x1 x2 , y1 y2 P G1 G2 with x1 , y1 P G1 and x2 , y2 P G2 . Then also, x1 y1 P G1 and x2 y2 P G2 .
We can add members of G1 and G2 so that they are in G1 G2 . Hence, px1 y1 q px2 y2 q 
px1 x2 q py1 y2 q P G1 G2

Proposition 5.4. Let G1 and G2 be subgroups of Z, then,

(i) G1 X G2 contains every subgroup contained in both G1 and G2 . In other words, it is the largest subgroup
contained in both G1 and G2 .
(ii) G1 G2 is contained in every subgroup containing both G1 and G2 . In other words, it is the smallest
subgroup that can contain both G1 and G2 .

Proof. (i) When 2 sets intersect, all the members that appear in both of them will be in the intersection.
Simillarly, all the numbers which are in both G1 and G2 will appear in G1 X G2 . So any subgroup that is
contained in both G1 and G2 will appear in G1 X G2 . For example, take 12Z and 9Z. 9Z X 12Z will be the
set 36Z. A subgroup that is contained in both 12Z and 9Z could be 108Z. 108Z is also contained in 36Z.

(ii) Any subgroup that contains G1 and G2 must contain all the members of the 2 subgroups, and at the
same time, will contain the sums x1 x2 where x1 P G1 , x2 P G2 . So G1 G2 contains G1 and G2 .
That G1 G2 is the smallest such subgroup is less clear. Consider any subgroup that contains G1
and G2 . What we know is that such a subgroup contains x1 x2 where x1 P G1 , x2 P G2 . Then this
subgroup contains all the members of G1 G2 and thus contains G1 G2 itself.

Lemma 5.1. For 2 integers m and n, m  n if and only if nZ mZ

Proof. We will prove the logical argument in the 2 directions for naturals. To prove it for integers, repeat the
same proof but using |n| and |m|.

m  n nZ mZ
m  n n is a multiple of m. Then n P mZ. This means that all multiples of n, and thus nZ is contained
in mZ.

m  n nZ mZ
nZ mZ n P mZ. We know that m is the smallest member of mZ and so, as n m, n must be a
multiple of m.

15
Now that we are done with these boring propositions and lemmas, we can move to the fun stuff. We can now
justify the Extended Euclidean Algorithm. First we will introduce the following convention:

Let x, yP Z. Then hcfpx, yq hcfpm, nq where m, n P N and m  |x|, n  |y|

Let x, y P Z. Then lcmpx, y q lcmpm, nq where m, n P N and m  |x|, n  |y |
Theorem 5.1. Let m, n P Z. Then,

(1) mZ nZ  hZ where h hcfpm, nq

(2) mZ X nZ  lZ where l lcmpm, nq

Proof. The propositions we looked at above help us to prove this.

(1) What we know about mZ nZ  hZ is that it contains mZ and nZ. This means that h  m and h  n. h
is a common divisor of m and n.
We need to show that it is the greatest common divisor. Let d be any common divisor of m and n, then
dZ contains hZ. This means that d  h.
(2) We know that mZ X nZ  lZ is contained in both mZ and nZ. So, m  l and n  l. l is a common multiple
of m and n.
Now let c be any common multiple of m and n. We know that, as n  c and m  c, cZ is contained in both
mZ and nZ. cZ will then also be contained in lZ l  c. So l is the lcm.

A good exercise would be to state the proposition that supports each argument in each of the above proofs.
This will prove useful in MA106, where a lot of the theorems are supported by many preceding propositions
and lemmas.
Exercise Using the previous theorem, try formulating an argument that proves the existence of
integers a and b such that hcfpm, nq  am bn

5.2 The wonders of Modular Arithmetic

We have so far seen inifinite number system. An interesting class of finite number systems are congruence
classes. To get a general idea of what congruence classes are, consider the 24 hour clock. If the time is 22:00,
then traditional arithmetic would tell us that 5 hours from now, the time should be 27:00. But, we all know
that in a 12 hour clock, the time should be 03:00. This is an example of addition modulo 12. What we did is
we added the two numbers 22 and 5 to get 27. Then we divided the result by 12 and the result of the operation
was the remainder we got.
Definition 5.2 (Addition and Multiplication modulo n). Let n P N and r, s P Z. Also, a r s and
b  r  s Then

Addtion modulo n, n
If a  pm1  nq p for some m1 P N, then r n s  p.
In other words, we added r and s and took the remainder on division by n as answer to addition modulo
n
Multiplication modulo n, n
If b  pm2  nq q for some m2 P N, then rn s  q.
We multiplied r ans s, and divided the result by n. The remainder is then the result of multiplication
modulo n

Example Here are a few examples of simple modular arithmetic

(i) 10 7 51

16
 9 0 1 2 3 4 5 6 7 8
0 0 1 2 3 4 5 6 7 8
1 1 2 3 4 5 6 7 8 0
2 2 3 4 5 6 7 8 0 1
3 3 4 5 6 7 8 0 1 2
4 4 5 6 7 8 0 1 2 3
5 5 6 7 8 0 1 2 3 4
6 6 7 8 0 1 2 3 4 5
7 7 8 0 1 2 3 4 5 6
8 8 0 1 2 3 4 5 6 7

Table 1: Addition modulo 9

(ii) 50 7 75  25
(iii) 16 1 05  0
(iv) 121 0011  32
(v) 82 43  24

Note that the results of modular arithmetic always lie in the set t0, 1, 2, ...., n
1u. This set is called Z{nZ,
pronounced zed mod n.

Also, we can put any number that is not in this set into it by doing the modulo n of the number. For any
integer, q, q(mod n) is the remainder upondivision of q by n

Now we will look at a few propositions that simplify the process of modular arithmetic. Knowing how to carry
out the computations in this chapter is important because these will be the most scoring exercises in the exams.
You may be asked the proofs also, and these are more or less easy once you manipulate the abstract notion of
division with remainder well.
Definition 5.3. Let  represent the operation of addition or multiplication.
 is Commutative means that x  y  y  x
 is Associative means that px  yq  z  y  px  zq
Proposition 5.5. Addition and multiplication modulo n are both associative.

Proof. We want a n pb n cq  pa n bq n c and an pbn cq  pan bqn c

Let n, p P N, then from the definition of division with remainder, we have, p q kn for some k, q P N.
Note that, a n p  a n q. This follows from:

a p  a pq knq
a p  pbn rq pq knq where a  bn r
a p  pb k qn pr q q,
Also, a q  bn pr q q. Taking modulus is simply dividing the RHS by n, which yields the same result.
Now all this implies that
@a, b, c P Z, a n pb n cq  a n pb cq and pa n bq n c  pa bq n c. It is obvious that both expressions
are just pa b cq(mod n)
Let n, p P N, then from the definition of division with remainder, we have, p  q kn for some k, q P N.
Note that, an p  an q. This follows from:
a  p  pbn rqpq knq where a  bn r
a  p  npbn rk qbq qr
We also have a  q  bn qr. Cleary, applying mod n to both RHS yields the same result.
Then, we have an pbn cq  an pb cq and pan bqn c  pabqn c. This has the same conclusion as
for addition modulo n

There are some interesting observations you can make about the above tables. Can you make any?

17
 9 0 1 2 3 4 5 6 7 8
0 0 0 0 0 0 0 0 0 0
1 0 1 2 3 4 5 6 7 8
2 0 2 4 6 8 1 3 5 7
3 0 3 6 0 3 6 0 3 6
4 0 4 8 3 7 2 6 1 5
5 0 5 1 6 2 7 6 8 4
6 0 6 3 0 6 3 0 6 3
7 0 7 5 3 1 8 6 4 2
8 0 8 7 6 5 4 3 2 1

Table 2: Multiplication modulo 9

5.3 Congruence

Definition 5.4. Let a, b P Z. a and b are said to be congruent mod n if they leave the same remainder on
division by n. Then we express this relationship as
a  b (mod n)

Note 2 numbers, a and b are congruent if and only if a
b is a multiple of n. This means that a
b (mod n) 0.
A good exercise would be to prove this

Actually, congruence shows that 2 numbers exist in a same set, which we are going to discuss soon. The next
lemma is very helpful in applying modulo on large numbers.
Lemma 5.2. Suppose that a1  b1 (mod n) and a2  b2 (mod n), then
(i) a1  a2  b1  b2 (mod n)
(ii) a1 a2  b1 b2 (mod n)

Proof. (i) pa1  a2 q  ka n ra and b1  b2  kb n rb . We want ra  rb .

a1
b1  k1 n and a2
b2  k2 n
We will use a very popular and clever trick, which is adding and subtracting a number in an expression
(which does not change the value) so as to be able to factorise the expression.
a1 a2
b1 b2  a1 a2
a1 b2 a1 b2
b1 b2
a1 a2
b1 b2  a1 pa2
b2 q b2 pa1
b1 q
a1 a2
b1 b2  a1 k2 n b2 k2 n
So, a1  a2  b1  b2 (mod n)
(ii) a1
b1  k1 n and a2
b2  k2 n
pa1 a2 q
pb1 b2 q  pa1
b1 q pa2
b2 q
Clearly, the RHS is a multiple of n. So, a1 a2  b1 b2 (mod n)

The lemma allows us to say that, for example, 4  13 (mod 9) 4k  13k (mod 9).
The above lemma allows us to simplify large numbers into smaller ones and carry out calculations with large
numbers. You will need to have a lot of practice with those as you are sure to get such exercises in the exams
and in the tests. You will have very little time to do these computations.

Example Here are a few examples of the potency of the above lemma. Simplify the expressions so that the
RHS lies in the set Z{n.

(1) 56 (mod 7)
56  p
2q6 (mod 7)
56  163 (mod 7)
56  23 (mod 7)
56  8(mod 7)
Hence, 56 (mod 7) 1

18
(2) 93  87 (mod 11)
93  87  p
2q3  p
3q7 (mod 11)
93  87  p
8q  p
3q7 (mod 11)
93  87 
3  p33  33  3q(mod 11)
93  87  p
9q  27  27(mod 11)
93  87  2  6  6(mod 11)
93  87  12  6(mod 11)
93  87  1  6(mod 11)
Hence, 93  87 (mod 11) 6

If you think hard enough, you will find that for a given number n, there are infinitely many numbers that are
congruent to m modulo n. Having that many numbers means that we can, for a given modulo n, separate
numbers into sets, or rather classes. Then we denote rms as the set of numbers that are congruent to m (mod
n). These separate the numbers into classes for a given modulo.

For example, for modulo 2, we can separate the numbers into even and odd. r0s would be the set of even
numbers and r1s that of odd numbers.

Given the appropriate modulo, you can separate numbers into subgroups of Z using these conventions.

5.4 A last thing you should know

Theorem 5.2 (Fermats Little Theorem). Let p be a prime number. Then @a that cannot be divided by p,

ap
1  1 (mod p)
You will find this theorem useful in your tests and your exam. You might be asked to state it, but not to prove it.

5.5 Some Fun Stuff

5.5.1 Multiplicative Inverse

Definition 5.5. The multiplicative inverse modulo n of a congruence class [a] in Z{n is the congruence class
ra
1 s which yields

a  a
1  1 (mod n)
It is to be noted that for a to have a multiplicative inverse modulo n, it needs to be coprime with n, meaning
that hcfpa, nq  1

Computing the above is easy for small modular classes, but gets tougher for larger numbers. For small numbers,
one can draw the multiplicative table to find the inverse. But for larger numbers, using the Extended Euclidean
Algorithm is better.

Given a and a modulus n, we want to find x such that

a  x  1 (mod n)
What we know is that ax
1 is a multiple of n. So,
ax
1  qn for some q P N. Rearranging gives
ax
qn  1
The RHS is actually hcfpa, nq. Using the Extended Euclidean Algorithm, we can find x, the multiplicative
inverse, and q.

19
5.5.2 Gaussian Algorithm

Ever wondered how you could determine the day of the week just by having the date? Well, the gaussian
algorithm does exactly that, and it uses modular arithmetic!

Definition 5.6 (The Gaussian Algorithm). w  pd t2.6m
0.2u y

y
4
u t
c
4
u
2cq (mod 7)

where:

w: name of the day (0  Sunday,...,6 Saturday)

d: day (1 to 31)

m: month (March=1,...,Feb=12)
y: last 2 digits of year (20 for 2012)
c: first 2 digits of year (12 for 2012)

The symbol txu means the greatest integer that is smaller than x.
The algorithm is still tedious to calculate, but luckily, you will be introduced to mathematica. The code for the
above algorithm is:

f2[day , month , year ] := Module[tk, y, c, mu,

k = day;
m = Mod[(month) - 2, 12];
y = If[Mod[month, 12] < Mod[(month - 2), 12], Mod[year - 1, 100],
Mod[year, 100]];
c = (year - Mod[year, 100])/100;
If[Mod[Floor[2.6*m - 0.2] + k + y + Floor[y/4] + Floor[c/4] - 2*c,
7] == 1, Mon,
If[Mod[Floor[2.6*m - 0.2] + k + y + Floor[y/4] + Floor[c/4] - 2*c,
7] == 2, Tues,
If[Mod[Floor[2.6*m - 0.2] + k + y + Floor[y/4] + Floor[c/4] - 2*c,
7] == 3, Wed,
If[Mod[
Floor[2.6*m - 0.2] + k + y + Floor[y/4] + Floor[c/4] - 2*c,
7] == 4, Thurs,
If[Mod[
Floor[2.6*m - 0.2] + k + y + Floor[y/4] + Floor[c/4] - 2*c,
7] == 5, Fri,
If[
Mod[Floor[2.6*m - 0.2] + k + y + Floor[y/4] + Floor[c/4] -
2*c, 7] == 6, Sat,
Sun]
]
]
]
]
]]

For example, if you want to know the day on which September 5, 2121 falls, you input the following

In[6]:= f2[5, 9, 2121]

Out[6]= Fri
Cool stuff huh?

6 Rationality + Irrationality = Reality

Some philosophy: life is full of rational people (those who take wise decisions) and irrational people (those who
do not know how to make wise decisions, like people who choose to do MORSE).

20
 m
Proposition 6.1. For a given rational number, q, there exists a unique expression where m and n are
n
coprime. This expression is called the minimal form of q.

m
Proof. Existence: For any rational , it is possible to divide both the denominator and the numerator
n
by h hcfpm, nq. It is obvious that
m n
and are coprime. If not, it is easy to rigorously work this out.
h h
Uniqueness: Suppose that we have q   m1 n2  m2 n1
m1 m2
n1 n2
Then from the above we have m1  n1 m2 and m2  n2 m1 . As mi and ni are coprime for i  1, 2, we have
m1  m2 and m2  m1 .

Proposition 6.2. If l P N is not a perfect square, then there is no number q P Q such that q2  l
m2
Proof. Suppose that l  n2 l  m2
n2
We know that each prime factor in n2 and m2 comes in pair and that at least one prime factor in l appears
only once. From the Fundamental Theorem of Arithmetic, we know that all the prime factors that appear on
the LHS must appear on the RHS (uniqueness of prime factorisation). Here it is not the case.
So Eq P Q such that q 2  l

We have just shown that there exist numbers that are not rational, that is, there exist irrational numbers. Then
we need a bigger number system that allows us to take square roots. Hence the reals. An important property
of the real numbers is that it constitute a complete space, which you wil discuss in MA131.

6.1 Decimal Expansions,to infinity.......and beyond- Buzz Lightyear

Here, we will briefly talk about the bijection between rationalty and decimal expansions. We all know what a
decimal number represents. But, as we need to be pedantic, it would be wise to be able to define a decimal
expansion.
Definition 6.1. The decimal number mk mk
1 ....m1 m0  n1 n2 ....nt .... where each mi and nj is a single digit
between 0 and 9 is the real number p10k  mk q ...... p101  m1 q p100  m0 q p 1  n1 q ..... p t  nt q
1 1
10 10

Now that we have defined a decimal number, we can prove the following theorem. The proof you will see is
very straightforward and actually provides us with a way to turn recurring decimals back to fractions. It would
be wise to follow this particualr lecture if you have difficulty grasping the idea.
Theorem 6.1. The number mk mk
1 ....m1 m0  n1 n2 ....nt .... is rational if and only if the sequence
mk , mk
1 , ...., m0 , n1 , n2 , ...., nt , ...
is eventually periodic.

This means that for any rational number, we want the last bit of the decimal expansion to repeat itself in-
finitely many time. Note that a repeating tail consisting of only zeroes would fit the above theorem. Before

going through the proof, let us consider a reccurring decimal number, say 0.4765.

Let us denote it by x.
10000x  4765  555555......
1000x  476  55555555.....
Then we can get rid of the recurring tail:
10000x
1000x  4289
9000x  4289
x
4289
9000

4
Let us now consider a rational number in minimal form, . Long division seems the easiest way to produce
11
a recurring decimal, and it is perhaps true in this case. However, for the purpose of introducing the proof,
consider the following method:
We create an infinite list by multiplying the numerator by powers of 10:
4, 40, 400, 4000, 40000, 400000, ...

21
Now we reduce the list modulo 11 (the denomenator) to get an infinite list with finite elements.
4, 7, 4, 7, 4, 7, ....
The first repeat is at 4  101 and 4  102 .
p102
101 q  11  9 So, denoting 114 by y,
p103
101 qy  4  9  36
If we do the opposite of what we did in the previous example, we can deduce that y  0.36.
Why this works will be explained in the proof. The example was just to remove the abstract part of it.

Proof of Theorem 6.1. We need to prove the statement in both directions

Recurring decimal rational number

Assume that the recurrence in the number x starts immediately after the zero and the sequence is d
numbers long
x  mk mk
1 ....m1 m0  n1 n2 ....nd n1 n2 ....nd ..
We can then multiply x by 10d and subtract x from the result so as to eliminate the recurring tail.
10d x
x  pmk mk
1 ....m1 m0 n1 n2 ....nd n1 n2 ....nd n1 n2 ....nd ..q
pmk mk
1 ....m1 m0 n1 n2 ....nd n1 n2 ....nd ..q

10d x
x  a P Z
x d
a
10
1
So we showed that x P Q
Now consider a number whose recurrence starts after e steps and the length of the recurrence is d.
x  mk mk
1 ....m1 m0  n1 n2 ....ne n1 n2 ....nd n1 n2 ....nd ..
10d e x
10d x  pmk mk
1 ....m1 m0 n1 n2 ....ne n1 n2 ....nd n1 n2 ....nd n1 n2 ....nd ..q
pmk mk
1 ....m1 m0 n1 n2 ....ne 
n1 n2 ....nd n1 n2 ....nd q
10d e x
10d x  b P Z
x d e
b
10
10d
Hence, x P Q
We just shifted the decimal places by multiplying by a power of 10 so that we can eliminate the recurring
tail.
Rational number recurring decimal
Assume that we have a rational number x 
m
in minimal form. An important consequence of what we
n
showed above was that if we can find naturals d and e such that p10d e
10d qx P Z, then we have shown
that x can be broken down to a recurring decimal number, with period d starting after e digits. (e and
d must be the smallest such numbers for this to apply). Now our claim is that such numbes exist for m
and n.
Consider the sequence m, 10m, 102 m, 103 m, ..., 10k , .... This sequence is clearly infinite, but if we apply
mod n to each of those numbers, we get a sequence consisting of only t0, 1, ...., n
1u. Clearly there needs
to be repeats.
We are interested in the first repeat that occurs, say at k1 and k2 . This means that
10k1 m  10k2 m (mod n)
10k1
10k2 P Z
m
n
The double implication mark we got in the previous part guarantees a recurring decimal in such a case.

7 Going Beyond Reality: The Complex Realm

?
The most important thing to remember is
1  i. Now this seems dubious, to have a number that has no
apparent value. Why this works in real life is that often, we find ourselves faced with polynomial equations
that do not have real roots when solving real life problems. Using the imaginary numbers then bridges the gap
between the solution and problem. It works in real life problems, so there is no reason not to consider it as a
possibility. What we will do in this section is simply look at the properties of the complex numbers.
?
Definition 7.1. A complex number z is in the form a bi where a, b P R and
1  i. Morevoer, the real
part of z, <pz q  a and the imaginary part of z, =pz q  b
Definition 7.2 (Equality and Arithmetic). Let z1  a1 b1 i and z2  a2 b2 i be 2 complex numbers.
Then,

22
 z1 z2  pa1 a2 q pb1 b2 qi
z1  z2  pa1 b1 iqpa2 b2 iq  a1 a2 a1 b2 i b1 a2 i b1 b2 i2  pa1 a2
b1 b2 q pa1 b2 a2 b1 qi
z1  z2 if and only if a1  a2 and b1  b2

Note that if a1 , a2  0, it makes no sense to say a1 i b1 a2 i b2 or vice versa. We will develop another way
of ranking complex numbers later.

7.1 The Complex Plane

Figure 1: Plotting Complex Numbers

Just like we are able to draw graphs for pairs of real numbers, we are able to do the same for complex numbers.
Except that the plane depicts one complex number only. You will learn in MA106 that complex numbers can
be viewed as 2-dimensional vectors. This also means that we can plot complex numbers on a 2-dimensional
axis. The imaginary axis is the vertical (or y-axis) and the real axis is the horizontal (or x-axis). The complex
plane allows us to geometrically interpret operations on complex numbers. For example, addition is merely a
case of vector addition. The complex plane also allows us to represent complex numbers using trigonometry
and, to a certain extent, using the exponential e.

7.2 The Conjugate

Definition 7.3. Let z a bi, then the conjugate of z is the complex number z  a
bi.

Some Properties of Conjugates:

1. z  z z P R
2. z
1 z2  z1 z2
3. 1  z2  z1  z2
z

23
 Proof. Let z1  a bi and z2  c di
Then z1  z2  pac
bdq pad cbqi z 1  z2  pac
bdq
pad cbqi
Also, z1  z2  pa
biqpc
diq  pac
bdq p
ad
cbqi

4. z a bi z  z  a2 b2

Proof. z  z  pa biqpa
biq  a2 b2

The conjugate of a complex number is very important when having to carry out division in C. Consider a non
zero complex number z  a bi. We already said earlier that Z is a field. So, we should be able to carry out
division in it, or rather, we should be able to evaluate the reciprocal of z. The simplification process involves
removing the imaginary number from the denominator by multiplying both the numerator and denomenator
by z

1
z
a 1
bi
1  pa
biq
 pa biq  pa
biq (2)

a
bi
 a2 b2
Definition 7.4. For a non zero complex number a bi,
1
z
 aa2
bib2 (3)

Example Let z1 4 3i and z2  5
2i. Evaluate

z1
z2
Solution:
1
z2
 5
1 2i
 552 2i
22
(4)

 5 2i
29
The division process has been reduced to a multiplication problem.


z1
z2
 5
29
2i
 p4 3iq
(5)
 1
29
p14 23iq

The geometric interpretation of the conjugate is that it is a reflection in the real axis on the complex plane.

7.3 The Argument and Absolute Value

When we think of coordinates in a cartesian plane, we think of positions relative to the origin. It is possible to
define the coordinates using the distance and direction relative to the origin. This also applies in the complex
plane. Defining complex numbers in such a way leads to interesting properties.

Definition 7.5. Let OP be the vector joining the origin and z a bi.

The argument of z, argpz q is the angle that OP

~ makes with the positive x-axis

The absolute value of z, |z | is the length of the vector.

24
 Figure 2: Argument and Absolute Value of a Complex Number

7.3.1 Evaluating the Argument and the Absolute Value

Let z a bi. argpz q and |z | are calculated using trigonometry.

argpz q  tan
1
b
(6)
a

We know that tan has a period of and so, care must be taken when calculating the argument. There are
2 conventions you may adopt. Either 0 argpz q 2 or
argpz q . It all depends on taste than
effectiveness here, as you will deduce in the next section.

a
|z|  | a2 b2 | (7)

7.3.2 Some Properties of |z |

|z z|  |z |2
|z |  0 z 0
|z1 z2 | |z1 | |z2 |

Proof.
a Letting z1  a1 b1 i and z2  a2 b2 i, we get on the LHS,
pa1 a2 q2 pb1 b2 q2 We will use Minkowskis Inequality, which is outside the syllabus of this course
to prove this. Here is the inequality. Note that in our case, p  2

p1 
p1 
p1

n
n
n
|ak bk |p
|ak |p
|ak |
p

  
Setting n  2, we get
k 1 k 1 k 1

a b b
pa1 a2 q2 p b1 b2 q2 a21 a22 b21 b22 , which is exactly what we expect to get.

|z1
z2 | ||z1 |
|z2 ||

25
 Proof. We actually want ||z1 |
|z2 || |z1
z2 |
|z1
z2 | |z1 |
|z2 | |z1
z2 |, which we can do
using what we just proved above.
|z2 z1
z2 | |z2 | |z1
z2 |
|z1 | |z2 | |z1
z2 |
|z1 |
|z2 | |z1
z2 |
We now subtract z1 instead of z2
|z2 z1
z1 | |z1 | |z2
z1 |
|z2 | |z1 | |z1
z2 |
|z2 |
|z1 | |z1
z2 |
|z1 |
|z2 |
|z1
z2 |
The first two are obvious. Their proof would be a nice exercise. It is worth noting that the subtracting in-
side modulus sign to break it down to an inequality will be used a lot in Analysis. Do not worry about the
Minkowskys Inequality too much, that was just to blow your mind. It will not be assessed.

The absolute value of a complex number is in fact a measure of its size.

7.4 Polar Co-ordinates

Knowing the co-ordinates of a complex number helps you find the argument and aboslute value of it. Is it
possible to reverse the process? That is, given the argument and modulus, can we give the coordinates of a
complex number? The answer is yes. It makes sense because coordinates represent the position of an object
relative to a fixed one. This requires direction (argument) and distance (absolute value) from the said fixed
point. The transformation has to go through an intermediate stage called polar coordinates.

Figure 3: Polar Coordinates of a Complex Number

Using trigonometry, it is clear that b  r sin and a  r cos . So we can write

z  r cos ir sin (8)

This way of writing complex numbers gives rise to many interesting properties, which we shall look at. Note
that regular coordinates x-y can be written this way too.
Lemma 7.1. These 2 identities will be useful in proving the following properties of polar coordinates.

26
 sinp1 2 q  sin 1 cos 2 sin 2 cos 1
sinp1
2 q  sin 1 cos 2
sin 2 cos 1
cosp1 2 q  cos 1 cos 2
sin 1 sin 2
cosp1
2 q  cos 1 cos 2 sin 1 sin 2

Suppose we have 2 complex numbers, z1  r1 pcos 1 i sin 1 q and z2  r2 pcos 2 i sin 2 q. We do know how to
convert the complex numbers into regualr form and carry out the task of multiplication and division. Something
more interesting is to understand the geometrical workings of such an operation. We shall consider this next.
Proposition 7.1. Consider z1 and z2 with |z1 |  r1 , |z2 |  r2 and argpz1 q  1 , argpz2 q  2 , then

(i) |z1  z2 |  r1 r2 and argpz1  z2 q  1 2

  

 z1 
(ii)  
z
 rr1 and arg
z1
z2
 1
2
2 2

Figure 4: Geometrical Interpretation of Multiplication

27
 Figure 5: Geometrical Interpretation of Division

The implications of the proposition can clearly be seen above. Adopting the convention
argpzq better
helps to work out multiplication and division this way.

Proof. The proof requires the use of polar coordinates.

(i) z1  r1 pcos 1 i sin 1 q and z2  r2 pcos 2 i sin 2 q. Then we have,

 
 r1 pcos 1 i sin 1 q r2 pcos 2 i sin 2 q
z1 z2

 r1 r2 pcos 1 cos 2
sin 1 sin 2 q ipsin 1 cos 2 sin 2 cos 1 q (9)

 r1 r2 cosp1 2 q i sinp1 2 q
Clearly, |z1  z2 |  r1 r2 and argpz1  z2 q  1 2
(ii) First note that
1
cos i sin
 cos
i sin
r1 pcos 1 i sin 1 q
z1
z2
 r2 pcos 2 i sin 2 q


 rr1 pcos 1 i sin 1 qpcos 2
i sin 2 q

2

(10)

 r pcos 1 cos 2 sin 1 sin 2 q ipsin 1 cos 2
sin 2 cos 1 q
r1
2



 r cosp1
2 q i sinp1
2 q
r1
2
  

 z1 
Clearly,  
z
 rr1 and arg
z1
z2
 1
2 .
2 2

Theorem 7.1 (De Moivres Theorem). Let n P Z, P R, then

pcos i sin qn  cos n i sin n (11)

28
Proof. We shall prove it for non-negative n first. As with most proof that make a statement about all natural
numbers, we shall use induction to prove the above. Let P pnq be the statement pcos i sin qn  cos n i sin n.

Base Case
P p0q says that
pcos i sin q0  cos 0 i sin 0
which is true, as both LHS and RHS equal one
Inductive Step We assume that P pnq is true. We need to show then, that P pn 1q is true. For that,
we use the previous proposition.

pcos i sin qn 1
 pcos i sin qpcos i sin qn
 pcos i sin qpcos n i sin nq (12)
 cospn 1q i sinpn 1q

The proof for negative n, uses what we just proved for positive n and the fact that sinp
q 
sin and
cosp
q  cos .

pcos i sin q
n  pcos 1

i sin qn
 cos n 1 i sin n (13)

 cos n
i sin n
The last line can be written as cosp
nq i sinp
nq

We will see that De Moivres Theorem has many applications, the most important of which is finding the nth
root of numbers. First, we shall introduce the exponential form of complex numbers, which further simplifies
multilpication and division.

7.5 The Exponential Form

Let us consider some taylor expansions, which you will cover at the end of your Analysis course.
Definition 7.6. The functions ex , sin x and cos x can be written as polynomials:

x2 x3 x4
8 xn

ex 1 x ..... 
2! 3! 4! n o  n!
x3 x5 7 8 x2n
1
sin x  x

x7! .....  p
1qn
no
3! 5! 2n 1

x2 x4 6 8 x2n

cos x  1

x6! .....  p
1qn
2! 4! 
n o
2n

The above will just help us justify using exponential forms to represent complex numbers.

Let us consider a complex number z  rpcos i sin q




2 4 6
3 5 7
rpcos i sin q  r 1

6! ..... i
i i
i .....
2! 4! 3! 5! 7!
 2 3

4 5 6 7
 r 1 i
2!
i 3! 4!
i
5!


6!

i 7! ..... (14)

 rpei q

29
Definition 7.7. Given a complex number z with |z |  r and argpz q  , then,

 rpei q z (15)

Lemma 7.2. Given complex numbers z1  r1 pei q and z2  r2 pei q,

1 2

z1  z2  r1 r2 eip q 1 2
(16)

z1
z2
 rr1 eip
q 1 2
(17)
2

The lemma just follows from proposition 7.1 and definition 7.7.

Consider the complex number with absolute value 1 and argument . Plotting it on the complex plane shows
that it is the coordinates (-1,0). This gives us Eulers Identity

Eulers Identity: ei 1

Mathematicians see it as the most beautiful expression because it has the 5 most important numbers in math-
ematics while being symmetric (no negative signs). You will see that and e keep popping up everywhere in
maths. The first is known as the geometrical constant, showing the relationship between circles and all other
shapes while the second one is the analytical constant, being the limit of many sequences and series.

7.6 Roots of numbers

There are 2 ways of finding the nth root of a number. You can either use De Moivres Theorem or the Exponential
Form of Complex Numbers. This is what we will do with a complex number z having modulus r and argument
.

7.6.1 Using De Moivres Theorem



z
1
n r 1
n i sinp

n
2k
cosp

n
2k
(18)

Here, k  0, 1, 2, ....., n
1. This is because sin and cos have a period of 2, and so many roots will repeat once
the argument exceeds 2.
 

n 

r n i sinp cosp r i sinp 2k q cosp 2k q

1 2k 2k
Proof.
n n

From the periodicity of sin and cos, we know, sinp 2k q  sin and cos  cosp 2k q

7.6.2 Using the Exponential Form of Complex Numbers

zn
1
r 1
n ei
2k
n (19)

Proof. From what we deduced from the previous section was that argpz n q  and |z n |  r n . Putting
1 2k 1 1

n
the nth root in exponential form yields the desired expression.

1
What we can note from the above is that z n has n roots and that each root will be separated on an argand
2
diagram (complex plane) by radians.
n

30
 Figure 6: Plotting the four fourth roots of unity

7.6.3 nth Roots of Unity

Another name for the number 1 is unity as it represents the unit of our counting system. It is a special
number, and consequently, its roots are special too, thus the special name roots of unity. An interesting
feature of the roots of unity is that they all have an absolute value of 1. Whatever the value of n is, the roots
of unity will be on the edge of the circle with radius 1 and centre the origin.

Let us see how we can calculate the roots of unity. We will use the exponential form of complex numbers for
that.
We know that 1  1  e0i

Then, 1 n  1 n ei n for k  0, 1, 2, ...., n
1.

1 1 0 2k

As expected, 1 is always an nt h root of unity. What we are more interested in, though is when k  1, denoted
by .

 ei 2
n

This is the second root you get on an argand diagram as you move anticlockwise, starting at 1, along the circle
4 6 8 pn
1q
(having center the origin and radius 1). As continue this movement, the roots are ei n , ei n , ei n , ....., ei n .
A clear pattern emerges whereby all the roots can be expressed as powers of .

Proposition 7.2. The nth roots of unity are 1, , 2 , 3 , 4 , ......, n
1 where,
 ei n .
2

31
The geometric series
Many of you might have heard of the geometric series in High School. A geometric series, is the
sum of the first consecutive powers of a number.

n
Sn 1 x x2 .... xn  xk

k o

We will derive the general expression for this sum.

 1 x x2
Sn 1 .... xn 1

Sn 1  1 x 1 x x2 .... xn
Sn xn 1  1 xSn
(20)
Sn
xSn  1
xn 1
1
xn 1
Sn 
1
x
Hence, we have,

 1
1
x x

n n 1
xk

k o

When we consider the sum of the nth roots of unity, we see that they are a geometric series. We just have to
replace x above by .

n
1  11

 11

1  0
n
1 2 ....

This makes sense from a geometrical point of view. If you look at the argand diagram for the fourth roots of
unity, you see that the top and bottom ones cancel and the right and left ones cancel. It becomes less clear
for bigger nth roots as the cancellation seems less obvious. For those of view, however, who have done physics,
you can view the lines joining the origin and the roots as representing forces that are trying to pull the origin
away. These forces each have the same magnitude and when you break them into their horizontal and vertical
components, you see that these cancel out.

8 Functions

Many of you already have an idea of what a function is. The notion that you should have of a function is of
a link between 2 sets. A function basically links a member of a set A to only one member of a set B. By
definition, a function does not link a member of A to more than one member of B. Note that A and B may be
the same set. The notation we use to show such a relationship between sets A and B is: f : A B. The set A
here is called the domain of f and B is called the co-domain of f .

Given a domain A, it is possible to build a set which is called the image of f (Impf q) where
Impf q  tf paq : a P Au
Now what we are really interested in is the nature of the maps. For instance, there may be a map that links
every element of A to only one element of B.

8.0.4 Injection

An injection is a type of mapping whereby each element in the domain is mapped onto exactly one element
of the co-domain. It is sometimes called a one-to-one function. This must not be confused with a
one-to-one correspondence. The notion you should have is summarised in the figure below.

Definition 8.1. Let f : A B. Then f is Injective if

32
 Figure 7: A one-to-one function

a1  a2 f pa1 q  f pa2 q
or equivalently (remember logic)
f pa1 q  f pa2 q a1  a2

This definition is a more complicated way of saying that an element of the domain is mapped to only one
element of the co-domain. It is, however, used in proofs.

Example Here are a few examples of Injective functions:

(i) f : N Q
f pnq 
1
n
(ii) g : Z Z
f pxq  ex
(iii) h : R R
hpxq  x3

8.0.5 Surjection

A surjection is a mapping where each element in the co-domain has an origin in the domain. In other words,
there cannot exist elements in the co-domain that is not a mapping by the function of an element in the domain.
The difference with injections is that an injection allows the co-domain to contain b : @a P Af paq  b where A
is the domain.

Definition 8.2. Let f : A B. Then f is Surjective if

@b P B Da P A such that f paq  b

B can thus be built from A by applying f to the latter.

33
 Figure 8: A many-to-one function

Example Here are a few examples of surjections:

(i) f : Q N
m
f
n
 |m| where m, n P Z
(ii) f : R R
f pxq  logpxq
(iii) f : R R
f pxq  2x

8.0.6 Bijection

Consider the function f : R Z where f pxq  2x.

x1  x2 2x1  2x2 . And also, @y P R, Dx P R such that y  2x.
Clearly, f is both injective and surjective. Such a function is called a bijection, or a one-to-one correspon-
dence between 2 sets. This is different from a one-to-one function. A bijection is a link between 2 sets
whereby each element in A is linked to only one element in B and vice versa.
Definition 8.3. A function is bijective if it is both Injective and Surjective.

Example Here are a few examples of bijection

(i) f : R R
f pxq  x3
(ii) f : Z Z
f pxq  x
3

It is possible to change pure injections and pure surjections to bijections by simply changing the domain and
co-domain. We will do that for all the pure injections and surjections we have considered so far.

Turning injections to bijections

34
 Figure 9: A one-to-one correspondence

(i) f : Nzt0u Q
f pnq 
1
n
becomes a bijection if we change the co-domain to the set tx  where n P N and n  0u
1
n
(ii) g : Z Z
f pxq  ex
becomes a bijection if we change the co-domain to the set R0

Turning surjections to bijections

(i) f : Q N
m
f
n
 |m| where m, n P Z
becomes a bijection if we change the domain to Q0
(ii) f : R R
f pxq  logpxq
becomes a bijection if we change the domain to R0

8.1 Some Interesting Stuff

8.1.1 Counting

When you are counting apples, you are actually labelling each apple with a number from the set of natural
numbers. So each apple is linked to a unique natural number and vice versa. This is a bijection.
Definition 8.4. Counting is simply establishing a bijection between the set of objects being counted and the
set of natural numbers, N
Definition 8.5. (i) A set A is finite if there exists a bijection between it and some subset of N, t1, 2, 3, ..., nu.
Then, we say that the cardinality of A (the number of memebers of A), denoted by pAq or |A| is n.
(ii) A set is infinite if it is not finite.

35
8.1.2 The Pigeonhole Principle

By now, all of you are familiar with your pigeonholes, where your lecturers or tutors will be putting your
corrected work. When your ST113 asiignment will be corrected, your lecturers will find himself with n distinct
scripts to put in n different pigeonholes. Assuming that all of you submitted the work on time and that no one
did 2 versions of the same assignment, one script will ned to go into only one pigeonhole. This automatically is
equivalent to each pigeonhole receiving exactly one script. This looks vaguely like a bijection. And it is using
this idea that we define the following:
Definition 8.6 (The Pigeonhole Principle). If A and B are finite sets with |A|  |B | and if f : A B is
a mapping, then

f is injective f is surjective
Proof. Label the elements of B so that we can write B  tb1 , b2 , b3 , ...., bn u.

f is injective f is surjective
Let mi be the number of mappings f paq  bi . In other words, it is the number of origins bi has for
1 i n. Note that mi 0. We know that f is injective which means that mi 1 as one element of A
is mapped to only one element of B, and so, each element of B cannot have more than one origin in A.
We also have that m1 m2 ....mn n. This is because, A has n elements and so, B can only have a
maximum of n origins from A.
mi 1 and m1 m2 ....mn n mi  1 where 1 i n
f is surjective f is injective
We know that f is surjective which means that mi 1 as . This is because each element of B must have
at least one origin in A.
Furthermore, for the same reason as above, m1 m2 ....mn n.
mi 1 and m1 m2 ....mn n mi  1 where 1 i n

8.1.3 Composite Functions

If we have 3 sets, A, B and C with functions:

f : A B
g : B C

then it is possible to define a new function

h : A C with
hgf

Definition 8.7. Let A, B, C and f and g be mappings with f : A B and g : B C, then the composition of
f and g, g  f : A C is given by


gf  g f paq where a P A

This only means that we apply f to a member of A so that it is found in B, then apply g to that new number
so that we are left with an element in C.
Proposition 8.1. If A, B, C, D are sets with functions f : A B, g : B C and h : C D, then,
h  pg  f q  ph  g q  f

36
 Figure 10: Composite Functions

Proof. The proof isreally straight.    

@a P A, h  pg  f q paq  h pg  f qpaq  h g f paq  ph  gq f paq  ph  gq  f paq
The composition of 2 functions is itself a function as the result is still a link between 2 sets. Something really
useful would be the ability to deduce their nature given that of the functions that compose them. The next
lemma does that.
Lemma 8.1. Supppose that f : A B, g : B C are mappings. Then
(i) f and g both injective g  f injective
(ii) f and g both surjective g  f surjective
(iii) f and g both bijective g  f bijective

Proof. (i) We know f and g are injective, meaning that,  

a1  a2 f pa1 q  f pa2 q and f pa1 q  f pa2 q g f pa1 q  g f pa2 q .
This is equivalent to a1  a2 g  f pa1 q  g  f pa2 q
(ii) We know f and g are surjective.
@c P C D b P B such that gpbq  c and @b P B D a P A such that f paq  b.
This means that take any element in C. It is possible to find an origin for that element in B. Then for
this particular element of B, it is possible to find an origin in A. The combined effect is that we are able
to find an origin for any element of C in A. So we can say that
@c P C D a P A such that g  f paq  c
(iii) Combine the above 2 proofs.

37
8.2 Inverses

Consider the function f : A B given by f pxq  2x with A  B  Z. We can also define a new function
g : B A given by g pxq  x.
1
2
Then, g  f pxq  g p2xq  x. The composition g  f acts as an identity map on the set A.

Definition 8.8. Consider a set A. The identity map, idA : A A is given by idA pxq  x @ x P A.
In the above example, g  f  idA . Also, g inverts what f does to x. Moreover, it is written to the left of f . So we
call it a left inverse of f . Similarly, f inverts g and it is placed to the right of g. It is thus a right inverse of g.

It worth noting that for any element x P B, f  gpxq  x. This leaves us with the conclusion that f is a left
inverse of g and g is a right inverse of f .
Definition 8.9. Consider the function f : A B and g : B A. Then we say that g is
a left inverse of f if g  f  idA
a right inverse of f if f  g  idB
an inverse of f if it is both a right and left inverse of f

Proposition 8.2. Consider f : A B. If there exists functions g : B A and h : B A that are Left
inverse and Right inverse to f respectively, then g  h and both are inverses to f .

Proof. Things that we know, (you should be able to justify those):


@ b P B, gpbq  g idB pbq

@ b P B, hpbq  idA hpbq
f  h  idB
g  f  idA

Using those (and associativity), we get :

g  g  idB  g  pf  hq  pg  f q  h  idA  h  h

Lemma 8.2 (Uniqeness of inverses). Consider f : A B. If g : B A and h : B A are inverses of f,

then g  h.

Proof. If g is an inverse of f , then it is also a left inverse of f . Similarly, h is a right inverse of f . By proposition
8.2, g  h.

Proposition 8.3. Let f : A B be a map, then:

If f is a surjection, then there exists an injection g : B A which is a right inverse to f .
Proof. We know f : A B is a surjection. We know that for each element in B, there is an element of A that
acts as origin. There may be many such origins for one particular element of B as shown in the diagram. This
also means that we can define a function g : B A by choosing for each b P B one unique element a P A such
that f paq  b. We then define g pbq to be that particular a
We are then left with an injective function (the diagram shows that). This is because we chose to map elements
of B to a single one of its origin(s) in A using g. So if we have b1  b2 , then g pb1 q  g pb2 q.

It is clear that g is then a right inverse to f .

38
 Figure 11: The Surjection f

Figure 12: We define a new function that is injective

Proposition 8.4. Let f : A B be a map, then:

If f is an injection, then there exists a surjection g : B A which is a left inverse to f .
Proof. It is basically the same thing as the previous proof. We define a new function g : B A by choosing
for each element a P A, the destination b P B where f paq  b. Define g pbq to be that origin a P A. Now, as you
can see from the diagram, it possible that elements from B do not get mapped to any element in A, in which
case, g is not a mapping. So for the unmapped bu P B, define g pbu q  aa where aa is any element in A (the
dotted line in the diagram).
Basically, we then have;

#
if f paq  binB
g pbq 
a
any element of A if b is not mapped to any point in A

39
 Figure 13: The Injection f

Figure 14: The Left Inverse of f

Clearly, g is a surjection as every element in A has an origin in B. What is perhaps less clear is how g is a left
inverse to f .
The composition g  f paq  g pbq  a. This is true for all elements of A. This is exactly the definition of a left
inverse. We can very rightfully neglect the unmapped elements of B, bu as g just needs to bring elements that
f maps to B back to A. The dummy bu are just there to make g a function.

Proposition 8.5. Let f : A B be a map, then if f has a left inverse then it is an injection.

Proof. Let us make use of some logic to do that. The opposite of an injection would be a function which obeys
the following rule:
There exists elements a1 , a2 P A such that a1  a2 f pa1 q  f pa2 q. This can be different from a surjection.
We will show: f not injective f has no left inverse.

40
 Figure 15: f is NOT an injection

If we try to define a function that maps elements of B back to A, we end up with a picture that looks like this.

Figure 16: g is NOT a function

Basically, what happens is that there will exist b P B such that g pbq  a1 and g pbq  a2 . This does not make
sense and g cannot be a mapping. What we could do to rectify this would be to choose which of a1 or a2 to
define as g pbq. Well choose a1 . (For the dummy elements, just map them to any element in A).
g is now a function, but when we do g  f pa2 q  g pbq  a1 . g is not a left inverse of f
Proposition 8.6. Let f : A B be a map, then if f has a right inverse then it is a surjection.

Proof. You should try this as an exercise. Here is a sketch proof:

f not surjective f has no right inverse.

Use the following definition for not surjective: For all elements a P A, there exists an element b P B
such that f paq  b

41
 Go on to show that in whatever way you try to define a function g : B A, there will always be a b P B
such that f  g pbq  b

Summary
Here is a small summary of what might be useful for the test on functions.

A function cannot map one element of a set to more than one element of another set.
The opposite of an injection is not always a surjection. For example, the function sin : R R is neither
an injection nor a surjection.
f has a left inverse f is injective
f has a right inverse f is surjective

8.3 Graphs

Definition 8.10. Let f : R R be a function. Then the graph of f is the set tpx, yq P R2 : y  f pxqu
Definition 8.11. Given sets X1 , X2 , X3 , Xn the cartesian X1  X2  X3  ...  Xn is the set tpx1 , x2 , x3 , ..., xn q :
xi P Xi u

Findind the equation of a 3D Plane

We can find the equation of a plane in 3D given three points A, B, C found on the plane. Let A  pxA , yA , zA q, B 
pxB , yB , zB q, C  pxC , yC , zC q.

Calculate two direction vectors in the plane,

AB  pxB
xA , yB
yA , zB
zA q

BC  pxC
xB , yC
yB , zC
zB q

Then find the normal N to the plane by calculating the determinant below. Note that i is the component of
the x-axis, j the y-axis and k the z-axis.
i j k

N  xB
xA yB
yA zB
zA
xC
xB yC
yB zC
zB
 i yB
yA zB
zA
j xB
xA zB
zA k B
x
xA yB
yA

yC
yB zC
zB xC
xB zC
zB C
xB yC
yB
N

x 
N  i pyB
yA qpzC
zB q
pyC
yB qpzB
zA q
j pxB
xA qpzC
zB q
pyC
yB qpzB
zA q

k pxB
xA qpyC
yB q
pxC
xB qpyB
yA q
 
pyB
yA qpzC
zB q
pyC
yB qpzB
zA q
 
pxB
xA qpzC
zB q pyC
yB qpzB
zA q
So, N
pxB
xA qpyC
yB q
pxC
xB qpyB
yA q
Now choose an arbitrary point on the plane P  px, y, z q and create a direction vector found on the plane with

any of the points A, B, C we already have. Let us choose P A.

Set P A  N
 0 and you will get an expression in the form ax by cz  d which is the equation of the plane.

Example Find the equation of the plane containing A  p0, 0, 3q, B  p2, 0, 0q, C  p0, 1, 0q

(i) Finding the normal
AB  p2, 0,
3q, BC  p
2, 1, 0q

i j k

3  i 0
3
j 2
3 2 0
 3i
2

2
0
1 0
1 0
2 0 k

2 1
6j 2k

Hence N  p3, 6, 2q
(ii) The dot product

Let P  px, y, z q be a point on the plane. P C  px, y
1, z q.


PC  N
0
3x 6py
1q 2z  0
3x 6y 2z
6  0 Is the equation of the plane.

42
 Figure 17: z
x
y 30

Finding the equation of a 3D line

We are given 2 points A  pxA , yA , zA q, B  pxB , yB , zB q. We find the direction vector

AB  pxB
xA , yB
yA , zB
zA q  pa, b, cq. Then we write the equation in parametric form,

x  xA at
y  yA bt
z  zA bt

you can replace pxA , yA , zA q by pxB , yB , zB q

Example Find the equation of the line through A  p1, 2, 5q, B  p2, 4, 3q

AB  p1, 2,
2q

x1 t
y  2 2t
z  5
2t

We can write that in cartesian form by expressing t in terms of x, y, z.

tx
1
y
2
t
2
5
z
t
2
y
2
The cartesian equation is x
1 
2
 5
2 z
Proposition 8.7. Let f : A B where A and B are subsets of the reals. Then the graph of f
1 is just the
reflection in the line y  x of the graph y  f pxq.

Proof. Graph of f  tpa, bq P A  B : f paq  bu and graph of f
1  tpb, aq P B  A : f
1 pbq  au.

This just tells us to invert the pair px, y q to py, xq to get the graph of f
1 .

9 Relations

Mathematics has loads of symbols. They can be split into groups, depending of what they represent. There are
symbols that represent an operation, a mathematical object or a relationship between 2 mathematical objects.

43
x y z

The above just means that the sum of 2 numbers is equal to another number. x, y, z represent mathemtical
objects; represents an operation; and  represents a relationship between the LHS and RHS.
We are interested in the relations symbol. You can use whatever symbols you want too represent relations, as
long as you define them. To remove the need for defining symbols each time you use them, conventions have
been adopted regarding them. A relation between 2 numbers could be end in the same digit and could be
represented by z, then we can say
15 z75
3 z1003

9.1 Equivalence Relations

This is one type of relation that needs to have the following three properties. We will denote the relation by 
(1) Symmetry: a  b b  a
(2) Transitivity: a  b and b  c ac
(3) Reflexitivity: a  a@a. In other words, a must be equivalent to itself.

The most common equivalence relation is the  sign. It has all 3 properties. The following are all examples of
equivalence relations:

Congruence modulo n
 where x  y means x and y have the same number of prime factors
 where x  y means x and y are both less than a number a
We can then split mathematical objects into equivalence classes where each member of a class is equivalent to
the other member of the same class. This is exactly what we did when we put numbers into congruence classes.

9.2 Order Relations

Distinct mathematical objects clearly have at least one property that is different from that of other objects. To
demonstrate this difference, order relations are quite useful. The most common difference between real numbers
concerns their size. So we use the sign to show this.
x y simply means that x is less than or equal to y. It is worth noting that this relation has the reflexitivity
and the transitivity property.
Another interesting fact is x y and y x x  y. This is called antisymmetry.

A relation  has the property of antisymmetry when x  y and y  x x  y.

Most order relations have the property of antisymmetry. The equality relation is both symmetric and anti
symmetric. It is also possible to find relations that are neither symmetric nor antisymmetric. One example is
the food chain (not a mathematical object, but still of some interest).

What you should know from this section is the definition of an equivalence relation.

44
10 Polynomials

This topic deals with a special kind of function. It is basically full of complicated syntax which you will have
to learn. This facilitates proving stuff about polynomials.
Definition 10.1. A polynomial P pxq is an expression of the form an xn an
1 xn
1 ... a1 x a0 where
a0 , a1 , a2 , ..., an are constants which we will refer to as the coefficients and x is the variable.

Important Terminology and symbols

n
We can write a polynomial as a summation ai xi
i 0 
The nature of the coefficient can be either natural, integer, rational, real or complex. In this case, the
polynomial will belong to different sets Nrxs, Zrxs, Qrxs, Rrxs, Crxs respectively.
So if we have a polynomial P1 pxq  x5 6x 1, we say that P1 pxq P Qrxs
1 3 3
x
8 7
We can assign a numerical value to x so that we give the polynomial a value. Note that the value we
? to x must be from the number system of the polynomial. So for the polynomial P1 above, x cannot
assign
be 3

We can group polynomials according to their degree also. The degree of a polynomial P (degpP q) the
highest power of x that appears in the expression. Note that the coefficient of the highest power of x is
not zero.
By convention, the zero polynomial has no degree. A polynomial with degree zero is just a constant as it
takes the form a0 x0

The leading term in a polynomial is the term with the highest power of x. The coefficient of this term is
called the leading coefficient. The leading term determines the degree of the polynomial. For obvious
reasons, the zero polynomial has no leading term. Polynomials with leading term 1 are called monic. This
is merely a fancy word which highlights, once again, the importance of unity (the number 1).

10.1 Polynomial Arithmetic

We are going to treat polynomials as algebraic objects and not as functions (although, to a certain extent, all
functions can be treated as algebraic objects). This merely means that we will carry out operations we normally
reserve for numbers on them. We will now look at some complicated notation for addition and multiplication.

10.1.1 Polynomial Addition

If you want to add 2 polynomials, you just add the terms which have the same power of x.
So, if we have 2 polynomials P1 pxq  an xn an
1 xn
1 ... b1 x b0 and P2 pxq  bm xm bm
1 xm
1 ... b1 x b0
with m n, then
P1 P2  bm xm .... pan bn qxn ... pa1 b1 qx pa0 b0 q
This is quite a messy way of defining addition, so we will use a fancier looking and shorter notation P1 pxq 

n
m
ai xi and P2 pxq  bi xi

i 0 
i 0




n
m
m
ai x i
bi x i
 pai bi qxi

i 0 
i 0 
i 0

If ai stops existing before the index m is reached, we just add zero to b.

10.1.2 Polynomial Multipication

Each term in the polynomial must be multiplied by all the terms in the other polynomial. The resulting list is
compressed by adding each term with the same power of x together.
We again have P1 pxq  an xn an
1 xn
1 ... b1 x b0 and P2 pxq  bm xm bm
1 xm
1 ... b1 x b0 (we

45
do not have to specify which of m or n is bigger).

P1  P2  an bm xm n pan bm
1 an
1 bm qxm n
1 .... pa1 b0 a0 b1 qx a0 b0

This is again very messy and we can find a better way of writing this down.



 k

n
m n
m
ai x i
bi x i
 ai bk
i xk

i 0 
i 0 
k 0 
i 0

The above expression means that we choose a k between 0 and m n. Then we list all numbers i such that
0 i k. We next compute all possible values of ai bk
i and add them together. The resulting number is the
coefficient of xk . Repeat for all values of k. This is exactly what we do in polynomial multiplication.

3
2
Example Let P1 pxq  ai xi and P2 pxq  bi xi



i 0 

i 0 

3
2
5
k
ai xi bi xi  ai bk
i xk

i 0 
i 0 k 0  
i 0

We choose k  0. Then we will have pi, k
iq  p0, 0q.

So the coefficient of x0 is a0 b0 .

We choose k  1. Then we will have pi, k
iq  p0, 1q, p1, 0q.

Coefficient of x1 is pa0 b1 a1 b0 q
We choose k  2. Then we will have pi, k
iq  p0, 2q, p1, 1q, p2, 0q.
Coefficient of x2 is pa0 b2 a1 b1 a2 b0 q
We choose k  3. Then we will have pi, k
iq  p0, 3q, p1, 2q, p2, 1q, p3, 0q.
Coefficient of x3 is pa0 b3 a1 b2 a2 b1 a3 b0 q
We choose k  4. Then we will have pi, k
iq  p0, 4q, p1, 3q, p2, 2q, p3, 1q, p4, 0q.
Coefficient of x4 is pa0 b4 a1 b3 a2 b2 a3 b1 a4 b0 q
We choose k  5. Then we will have pi, k
iq  p0, 5q, p1, 4q, p2, 3q, p3, 2q, p4, 1q, p5, 0q.
Coefficient of x5 is pa0 b5 a1 b4 a2 b3 a3 b2 a4 b1 a5 b0 q

Proposition 10.1. Let P1 and P2 be polynomials then,

degpP1 P2 q  degpP1 q degpP2 q

Provided P1 P2  0, degpP1 P2 q degpP1 q degpP2 q

10.1.3 Polynomial Division

Proposition 10.2. Let P1 , P2P S rxs and P2  0. There exists Q qnd R such that
P1  QP2 R

with R  0 or degpRq degpP2 q

Note

S can be any field Q, R, C but not Z.

When R  0, there is exact division.
degpP1 q degpP2 q

46

n
m
Proof. We let P1 pxq  ai xi and P2 pxq  bi xi first.

i 0 
i 0
The proof ressembles that of division with remainder for natural numbers. We just want to prove the fact about
R, i.e. that it is either zero or has a degree less than that of P2 . Let Q be any polynomial in S rxs. Then we
can define Tr to be the set of polynomials of the form R  P1
QP2 . (Remember, we already know our P1 and
P2 ).
If 0 P Tr then there is exact division and R  0. Otherwise, we know degpRq 0. We define another set
D  tr  deg R : R P Tr u
As degpRq P N (and clearly D is not empty), we can say that there is a least element r0 P D. We want
r0 degpP2 q. Let us assume on the other hand that r0 degpP2 q.
The polynomial that has degree r0 is R0  P1
QP2 . It is actually the polynomial of least degree possible on
Tr , having leading term cxr0 .
Consider the term
xr0
m P2 . We can add it to both sides of our expression R0  P1
QP2 (Step ).
c
bm
c r0
m
R0
x P2  P1
QP2
bc xr0
m P2
bm m


c r0
m c r 0
m
R0
x P2  P1
Q x P2
bm bm

Clearly, the above polynomial is still in the form R  P1
QP2 , meaning that it is in Tr . The degree of the
polynomial is in D. However we see that the degree of this polynomial is smaller than ro .
Hence, r0 degpP2 q

Remark

In the proof, we showed that the least element of D was less than degpP2 q. We said that there exists an R
having such a degree. There might exist other Rs with higher degree, but we are not interested in them.
Similarly, we are not interested in the other elements of D, because as long as the least element is less
than degpP2 q, we know we have found our R.
We can do step because we assumed that r0 m r0
m 0. Otherwise, we would only be able to
c m
r0
add x P2 and this gives us one of the Rs with higher degree (which is not necessarily less than
bm
the degree of P2 ). We then get an element of D which is bigger than the least element.

We showed the existence of such a polynomial, but we do not yet know how to find the quotient and the
remainder. The easiest way to do this is to use long division.
Let us say we want to divide P1  x3 x2
5x 1 by P2  x
1. We write down the following:

x
1 x3 x2
5x 1

x3
Then we consider the first highest power of P2 . We divide the highest power of P1 by it. We get
x
 x2 . We
write it on the top line as shown.

x2

x
1 x3 x2
5x 1

We then multiply the whole of P2 by the term we just wrote, i.e. x2 and write the result (x3
x2 ) underneath
the terms of corresponding power of x in P1 . We then subtract this result from P1 and write the remainder
down as shown. We also bring down the next (consecutive) power of x from P1 down.

x2

x
1 x3 x2
5x 1

x3 x2

We consider again the highest power of x in P2 , i.e. x. We divide the term with the highest power of x in the
2x2
remainder we got by it.
x
 2x. We multiply what we just got by the whole of P2 and write the result under
our previous remainder.

47
 Figure 18: We will use the right hand notation for long division

x2 2x

x
1 x 3
x2
5x 1

x3 x2
2x2
5x

2x2 2x
We do the subtraction and bringing down process again.
x2 2x

x
1 x 3
x
5x
2
1

x3 x2
2x2
5x

2x2 2x

3x 1
We repeat our algorithm with the highest powers of P1 and P2 .
x2 2x
3

x
1 x 3
x
5x
2
1

x3 x2
2x2
5x

2x2 2x

3x 1
3x
3

2
From our operation, we get
Q  x2 2x
3 and R 
2

Here is another example:

48
P1 pxq  x6
5x4 3x2 , P2 pxq  x3 x
1
Step 1:

x3 x
1 x6
5x4 3x2
Step 2:
x3

x3 x
1 x6
5x4 3x2

x6
x4 x 3

6x 4
x3 3x2
Step 3:

x3
6x
x 3
x
1 x
5x
6 4
3x 2

x6
x4 x3

6x4 x3 3x2
6x 4
6x2
6x
x3 9x2
6x
Step 4:

x3
6x 1
x 3
x
1 x
5x
6 4
3x 2

x6
x4 x3

6x4 x3 3x2
6x4 6x2
6x
x3 9x2
6x

x3
x 1
9x 2

7x 1

Proposition 10.3. For polynomials P1 and P2 where P1  QP2 R, Q and R are unique.

Proof. Let P1  Q1 P2 R1 and P1  Q2 P2 R2 .

Q1 P2 R1  Q2 P2 R2
P2 pQ1
Q2 q  R2
R1
We assume that Q1  Q2 .
We have degpLHSq degpP2 q and degpRHSq maxtdegpR1 q, degpR2 qu.
We know that degpR1 q degpP2 q and degpR2 q degpP2 q. This means that maxtdegpR1 q, degpR2 qu degpP2 q.
Hence, degpLHSq  degpRHSq.
Hence, Q1  Q2 R2  R1

10.2 Prime Polynomials, HCF and LCM

When we were working with natural numbers, we encountered numbers that could not be broken down into a
product of smaller natural numbers. Similarly, there are polynomials that cannot be expressed as products of
polynomials with smaller degrees.
Definition 10.2. Let P P Rrxs and degpP q 0. P is said to be irreducible in R if whenever it is expressed
as a product of polynomials (in R) P1 P2 , then either P1 or P2 is a constant. This is called a trivial factorisation
of P
In other words, it is impossible to find P1 and P2 with both degpP1 q 0 and degpP2 q 0 such that P  P1 P2 .

Definition 10.3. Let P P Rrxs and degpP q 0. P is said to be reducible in R if it can be expressed as a
product P1 P2 where degpP1 q 0 and degpP2 q. This is called a non-trivial factorisation of P .
Lemma 10.1. Every polynomial of degree 1 is irreducible.

Proof. Let P be a polynomial with degpP q  1.

Consider P  P1 P2 . degpRHSq  1. Either degpP1 q  1and degpP2 q  0 or degpP1 q  0and degpP2 q  1

Now we will show that polynomials have some sort of prime factorisation also.
Proposition 10.4. Every polynomial of degree greater than 0 is divisible by some irreducible polynomial.

49
Proof. Either P is irreducible, in which case it divides itself, or it is not.
Consider the set F of non-trivial factors of P in Rrxs. Now consider the set DF of degrees of the polynomials
found in F . DF is obviously not empty. By the WOP, DF has a least element, d0 . This corresponds to a
polynomial P0 P F .
Our claim is that P0 is irreducible. Consider the opposite. Then, P0  Q1 Q2 Q1 , Q2 P F . But then both
have degree smaller than d0 . So P0 is not the polynomial of least degree in F
Hence P0 is irreducible.

Actually, we can extend our statement and say that every polynomial of degree greater than zero can be
expressed as a product of irreducible polynomial.
Definition 10.4. Let P1 , P2 P Rrxs. We say that Q P Rrxs is a highest common factor of P1 and P2 if
It divides both P1 and P2
P1 and P2 have no common factor of higher degree
Lemma 10.2. Let P1 , P2 , Q, R P Rrxs such that P1  QP2 R. Then the set of common factors of P1 and P2 ,
denoted by FpP1 , P2 q is equal to the set of common factors of P2 and R denoted by FpR, P2 q

Proof. FpP1 , P2 q FpR, P2 q

Let F be a common factor of P1 and P2 . Then for some S1 , S2 P Rrxs, P1  F S1 and P1  F S1 .
P1 QP2 R
R  S1 F
S2 F Q
R  pS1
S2 QqF
Clearly, F  R and F is a common factor of P2 and R
FpP1 , P2 q FpR, P2 q Let F be a common factor of R and P2 . Then for some S1 , S2 P Rrxs, P2  F S2
and R  F S1 .
P1  QP2 R
P1  QF S2 F S1
P1  pQS2 S1 qF
Clearly, F  P1 and F is a common factor of P1 and P2

This lemma allows us to use the Euclidean Algorithm for polynomials.

Example We want hcfpP1 pxq, P2 pxqq in each of the following

(1) P1 pxq  x5 6x4 3x2
6 and P2 pxq  x2 3x
2

x5 6x4 3x2
6  px3 3x2
7x

30qpx2 3x
2q p
104x
54q
x 2
3x
2  104
1 x
183 p
104x 54q
2704
467
5408



104x 54 
281216 x 146016 467

0
467 467 2704


So we have, hcfpP1 pxq, P2 pxqq  hcf  2704

467 467
,0
2704

(2) P1 pxq  x4 5x3
2x2 8 and P2 pxq  x2
7x 10

x4 5x3
2x2 8  px2 12x 72qp

x2
7x 10q p384x

712q
x2
7x  384 x
p384x
712q
1 247 1057
10
18432 2304



384x
712  x

884736 1640448 1057
0
1057 1057 2304


So we have, hcfpP1 pxq, P2 pxqq   1057

1057
hcf ,0
2304 2304

50
We have just seen the Euclidean Algorithm for polynomials. Is there an Extended Euclidean Algorithm for
polynomials? The answer is yes. It works exactly as the one for natural numbers. It is very messy.

Example Let us consider again P1 pxq  x4 5x3
2x2 8 and P2 pxq  x2
7x 10

x4 5x3
2x2 8  px2 12x 72qp

x2
7x 10q p384x

712q
x2
7x  384 x
p384x
712q
1 247 1057
10
18432 2304



384x
712  x

884736 1640448 1057
0
1057 1057 2304

hcfpP1 , P2 q  p9x
7x 10q
x
p384x
712q

2 1 247
384 18432



 px
7x 10q
384 x
18432 px 5x
2x 8q
px 12x 72qpx
7x 10q

2 1 247 4 3 2 2 2

 



 1 384 x
18432 px 12x 72q px
7x 10q
384 x
18432 px4 5x3
2x2 8q
1 247 2 2 1 247

We found polynomials A, B such that hcfpP1 , P2 q  AP1 BP2

Proposition 10.5. If P P Rrxs is the hcfpP1 , P2 q, then any common factor of P1 and P2 in Rrxs divides P .
Proof. We just introduced the euclidean algorithm for polynomials. We use it to show this proof.

We know that the algorithm comes to an end. This is because each time we apply the algorithm, we get
polynomials with a smaller degree. As degrees of polynomials are natural numbers, they cannot go below
zero.

The eventual significance of the algorithm is that FpP1 , P2 q  F hcfpP1 , P2 q . This means that the set of
common factors of P1 and P2 is the same as the set of factors of hcfpP1 , P2 q.

Proposition 10.6 (Uniqueness of polynomial HCF). Let P and Q be HCFs of P1 and P2 where P1 , P2 P
Rrxs. Then D P Rz0 s.t P  Q

This shows that the HCF of polynomials are unique up to multiplication by a constant.

Proof. By proposition 10.5, we know that any common factor of P1 and P2 divides hcfpP1 , P2 q. So clearly,
P
Q
 T1 and Q
P
 T2 where T1 , T2 P Rrxs. The only way for this to happen is for T1 and T2 to be constant.

51
Example Let P1 pxq  x2 x
2 and P2 pxq  x3 3x2
x
3
We can factorise both polynomials into irreducible degree 1 polynomials.
P1 pxq  px
1qpx 2q
P2 pxq  px
1qpx 1qpx 3q
Clearly, hcfpP1 , P2 q  x
1  P . Now consider Q  px
1q
1
2
2x 4
1
x
x
2
1
x2
2 2

x 2
x
2x
2

2x 2
0
2x2 8x 6
1
x

x
3
1
x3 3x2
2 2

x 3
x 2

4x2
x

4x2 4x
3x
3

3x 3
0
Quite clearly, Q divides both P1 and P2 and factorising each quotient shows us that there is no
other polynomials of higher degree that divide both P1 and P2 . Hence, Q hcfpP1 , P2 q

The following lemma follows from what we just did. It is up to you to prove it (either using the result we just
got or using the same technique we used for natural numbers).
Corollary 10.1. The factorisation of a polynomial of degree greater than zero into irreducible polynomials is
unique up to the ordering of the factors and multiplication of the factors by non zero constants.

10.2.1 Polynomial LCM

Definition 10.5. Let P1 , P2 P Rrxs, the lcmpP1 , P2 q is a polynomial in Rrxs such that it is the polynomial of
lowest degree that is divisible by both P1 and P2 .
 
Corollary 10.2. P1 P2  hcf P1 , P2  lcm P1 , P2 .


Sketch proof. Let H pxq  hcf P1 , P2

Consider M pxq 
P1 P2
H
M pxq  HQ1 Q2 where Q1  and Q2 
P1 P2
.
H H
We can see that M is divisible by both P1 and P2 It is a common multple of P1 and P2 . Is it the one of least
degree though? The answer is yes.
We know that any multiple of P1 and P2 must also be a multiple of H. If we try to reduce M to a polynomial
of lower degree, we will have to get rid of either a factor of Q1 or Q2 . But then either P1 or P2 will no more
divide M .

10.3 The Remainder Theorem

Theorem 10.1 (The Remainder Theorem). Let P P Rrxs, and let P R such that P  px
qQ R where
Q, R P Rrxs, then,
R  P p q

The theorem says that if you want to divide a polynomial P by a degree 1 polynomial px
q, then the remainder
you get is the numerical value of the polynomial P when x  .

52
Proof. The proof is simple. We know from the definition of division with remainder that if we have

 px
qQ
R
P

Then, degpRq degpx
q  1 or R  0. So R  for some P R

Replacing in either side of the above equation yields

P p q  px
qQp q

P p q 

Corollary 10.3. Let P P Qrxs and P Q. P p q  0 px
q divides P

The above one works for R and C also.

10.4 Roots of Polynomials

We covered the concept of inverse in the section on functions. We already saw that polynomials are functions
and so, for a certain subset of their domains, there exists some notion of inverses. What we are interested in
knowing, however, is the value of x such that

an xn an
1 xn
1 ... a1 x a0

an x n
an
1 x

n 1
... a1 x a0
 0
where P C. We could also work in R and Q
The concept of roots of a function help us in that. A root would be the value of x which makes the polynomial
equal zero.
Definition 10.6. Let P P Krxs and P K. Then, is called a root of P if and only if P pq  0

The geometrical interpretation of roots anre simply the values of x where the grap y  P pxq crosses the x-axis.

Figure 19: y  x3
5x2 7 crosses the x-axis three times

When working in the reals, it may be possble that the graph does not cross the x-axis at all (for even degree
polynomials). Hence we have the following proposition.
Proposition 10.7. A polynomial P P Rrxs of degree n has at most n roots in R
So, number of roots n
53
 Figure 20: y  x3
2x2 7 crosses the x-axis only once

Proof. The proof is an induction on n.

Base Case: For n  0, P pxq  a0 , which is a constant, so P has no roots. So our proposition holds for
n0
Inductive Step: We assume it is true for degrees up to n
1. Now let P be a polynomial of degree n.
If P has no roots in R then the statement is true for degree n. Otherwise for some P R,
P  px
qQ where Q P Rrxs, degpRq  n
1.
We know that Q has at most n
1 roots in the reals. So Q factorises into at most n
1 degree 1 polynomial
factors of the form px
k q . These are also factors of P , and so, P has at most n such factors.
So it is true for all n.

Consider the polynomial Pc pxq  x2 1. Clearly, it has no roots in R. Things change if we move to the complex
realm. We can then factorise P .
Pc pxq  px
iqpx iq where i2 
1.

One way to interpret polynomials that have less roots than their degree is to think of those roots as appearing in
CzR. So missing roots in the reals can be found in the complex realm. This is exactly what the Fundamental
Theorem Of Algebra says.
Theorem 10.2 (Fundamental Theorem of Algebra). Every polynomial in C of degree greater than zero
has a root in C

The proof for this one is quite complex, but intuition sort of justifies it.
Corollary 10.4. Every polynomial in C of degree greater than zero have n roots in C

n
Sketch Proof. We will prove it by induction for a polynomial P pxq  ak xk

k 0

Base Case: For n=1, it is obvious that P has one root.

Inductive Step: We assume that polynomials of degree pn
1q have pn
1q roots in the complex realm.
Then, consider a polynomial P of degree n. We know from the Fundamental Theorem of Algebra that P
has a root P C.
P  px
qQ where Q is a polynomial of degree n
1 which has n
1 factors of the form px
q. So P
has the n
1 factors of Q as its own factors plus px
q as another factor. So P has n roots.

54
There is a direct relationship between the roots of a polynomial and its factors. When we are talking about the
number of roots, we are also talking about the number of polynomial factors of degree 1 it has.

Figure 21: y  x2
2x 1 crosses the x-axis only once

The polynomial x2
2x 1 crosses the x-axis only once. Does this mean it has only one root? No because
if it did have a unique root in the reals, it would have a unique factor, px
1q, appearing in its factorisation.
Clearly, x2
2x 1  x
1. However, it factorises to px
1qpx
1q.

Here, 1 is called a multiple (or repeated) root of the polynomial. The fundamental theorem of algebra does
allow for the existence of repeated roots. The multiplicity of a root is the nmber of times it appears as degree
1 factors of a polynomial. For x2
2x 1, the root 1 has a multiplicity of 2.

Proposition 10.8. Let P P Crxs and P C.

(i) is a multiple root P p q  P 1 p q  0
(ii) P has no multiple roots if P and P 1 have no common factors.

Proof. 1.
Let be a root of P with multiplicity k 2. Then,
P pxq  px
qk Qpxq where Q P Crxs
P 1 pxq  k px
qk
1 Qpxq px
qk Q1 pxq
Clearly, P p q  P 1 p q  0

We have that P p q  P 1 p q  0
This means that P pxq  px
qQpxq where Q P Crxs.
P 1 pxq  Qpxq px
qQ1 pxq. Now, P p q  0 Qp q  0. This means that Qpxq  px
qRpxq
where R P Crxs.
We conclude that P pxq  px
q2 Rpxq. Hence, is a multiple root (mutiplicity greater than 1).
2. We just showed is a multiple root P p q  P 1 p q  0. Not having a multiple root then implies that
P and P 1 have no common factor.

55
10.4.1 Algebraic Numbers

n
Definition 10.7. Let P C and P P C where P  ak xk and ak P Q for all 0 k n.
k 0 
is said to be algebraic if it is a root of P .

That is an algebraic number is one which is a root of a polynomial with rational coefficients. All rational
numbers are algebraic as they are the roots of the polynomial x2
q 2 . Also, the nth root of any rational number
is algebraic.
Numbers that are not algebraic are called Transcendental numbers. e and have been proved to be
transcendental. One of the interesting results of this is that there can be no square whos area is equal to a
given circle.
Corollary 10.5 (Rational Root Test). Let P P Qrxs be a polynomial with integer coefficients and ptx
rq
be a factor of P with r, t P Z and hcfpr, sq  1, then,

(i) r  a0 where a0 is the coefficient (non-zero) of the term with the smallest power of x
(ii) t  an where an is the leading coefficient

(i) We know then, that any multiple of ptx
rq  P , hence, px
qP P p rt q  0

r
Proof.
t

n 
n
1 

an
r
t
an
1
r
t
... a1
r
t
a0 0

n 
n
1 

an
r
t
an
1
r
t

a0 we multilpy both sides by tn
... a1
r
t
an rn an
1 rn
1 t ... a1 rtn
1 
a0 tn we take a factor of r from the LHS
rpan rn
1 an
1 rn
2 t ... a1 tn
1 q 
a0 tn

The RHS and LHS are both integers. Also we have that t and r are coprime. This means that r  a0

(ii)

n 
n
1 

an
r
t
an
1
r
t
... a1
r
t
a0 0

n
1 

n
an
1
r
t
... a1
r
t

an rt
a0 we multilpy both sides by tn

an
1 rn
1 t ... a1 rtn
1 a0 tn 
an tn

tpan
1 rn
1 ... a1 rtn
2 a0 tn
1 
an tn
Once again, it is clear that t  an

Corollary 10.6. Let P pxq  an xn an
1 xn
1 .... al xl where ak P Z@l k n, then if rt is a factor of P ,
then

r  al
t  an

To prove it, just set P pxq  xl pan xn
l an
1 xn
1
l .... al q and use the previous proposition.

The application of the above proposition is that it allows us to initiate the factorisation of a polynomial in Q
that has real roots. We also can find roots using the method.

56
Example We want to factorise P pxq  2x4
12x3 18x2 8x
24.
We want 1 such that P p1 q  0. From our previous proposition, we know that
1 P F2 4  tset of integer factors of 24u.
Using trial and error, we find that P p3q  0 px
3q  P .


2x3
6x2 8
x
3 2x 4

12x 3
18x 2
8x
24

2x4 6x3

6x3 18x2
6x
18x2
3

8x
24

8x 24
0
P pxq  px
3qp2x 3

6x 2
8q  2px
3qpx3
3x2 4q

We can check if 3 is a multiple root. P 1 pxq  8x3
36x2 36x 8 P 1 p3q  8  0. Hence, we need
a different root 2 . This time, 2 P F8 . This is because we want to reduce Q1 pxq  x3
3x2 4
By trial and error we find that P p2q  0 px
2q  P . Is it a multiple root? P 1 p2q  0, so we
can say that px
2qpx
2q  px2
4x 4q  P .

x 1

x 2

4x 4 x 3

3x 2
4

x3 4x2
4x
x2
4x 4

x2 4x
4
0
Hence, P(x) = 2(x-3)(x-2)(x-2)(x+1)

This method does not really help us to find roots that are irrational. We can use the complete the square
method for degree 2 polynomials.
Proposition 10.9. Consider P pxq  ax2 bx c for some a, b, c P C. The root of P is given by
?2


b  b
4ac
2a

Proof. We complete the square.

ax2 bx c0
b 
a x2 x c0
a
b 2 b 2 
a x2
b
a
x
2a

2a
c0
b 2 b2
a x
2a
 4a

c
b 2 b
4ac
2
x 
2a ?4a22
x
b
  b2a
4ac
2a ?
x

b  b2
4ac
2a

The above proposition also allows us to determine the nature of the roots of a quadratic.

b2
4ac 0 roots PR
b2
4ac 0 roots P CzR

57
 b2
4ac  0 multiple roots PR
However, this is not very useful in finding the roots of more complex polynomials. We can approximate the
roots using the newton-raphson method.
Proposition 10.10. Let f : A R be a well behaved function (Analysis 2), then the nmber for which
f pq  0 is given by


f pxn q
 lim
n 8 xn
f 1 pxn q
Where x0 is an arbitrary number that is chosen to be more or less close to the root.

The method can be applied to polynomials. You simply replace f by P . It works over the complex realm also.
Care must be taken when using the newton raphson method as it may not always converge. You will learn this
in St108.

58
11 Counting

I will use the notation pAq as representing the cardinality of set A. This is the size of the set. For finite sets,
it represents the number of elements of the set. The same principle can be extended to infinite sets.

I give you a box (set) of oranges O, and I ask you to count them. The answer you will give me is a natural
number l. How did you arrive to this? You counted, that is, you assigned a natural number 1 to the first
oraneg you saw, the number 2 to the second orange and so on. What you actually did was create a function
that takes elements from the set O to a subset of N. The function assigns one element of O to one and only one
element of S N (otherwise you count oranges more than once). Also, every element in the subset is bound to
an orange otherwise, you might be couning wrongly.
You set a bijection between the set of oranges and the subset of the naturals.
Definition 11.1 (Counting). Counting is simply setting a bijection between the set of objects being counted
and the set A N

Note: The set A usually (but not necessarily) takes the form A  tn P A : n N u for some arbitrary N PN

The notion we have of set is that: A B pAq pB q. This is true only for finite sets. It is not for infinte
sets. What actually determines cardinalities of 2 sets is the existence or absence of an injection or surjection
between the 2 sets.

11.1 Countable and uncountable sets

A set is countable set if given enough time and effort, all its elements can be paired with an element of the
naturals. This pairing should be one-to-one and onto. A set is uncountable if it is impossible to pair its elements
with elements of the naturals. The formal definition is:

Definition 11.2. A set A is said to be countable if there exists a bijection b : A S N.

A is uncountable if there does not exist such a bijection.

Now, there exists sets that are infinite. Some of these sets can still be put into bijection with the naturals
(which is also infinite) and some cant. Those that can are said to be countably infinite while those that
cant said said to be uncountably infinite.
The intuition you should have is that for a countably infinite set, you can still count the members by assigning
natural numbers to them. The counting will never stop. For uncountably infinite sets, you will never be able
to find all the elements because the set is so dense. It is thus impossible to even start counting the elements.
Proposition 11.1. Let B N be an infinite set. Then there exists a bijection h : N B
Here, we have that B is a subset a N (excluding the possibility of equality). So B might be a set that has least
element greater than the least element of the naturals or have gaps in it, i.e, it is not possible to find an infinite
sequence of consecutive naturals by using its members. However it is possible to list all its members.

Proof. The proof is in 2 stages. First we need to create the function h, then we will show that it is bijective.

(i) Consider the set B. By the WOP, it has a least element b0 . We shall define hp0q  b0 .
Next consider B ztb0 u. It also has a least element b1 . Let hp1q  b1 .
We consider B ztb0 , b1 u. It also has a least element b2 . Let hp2q  b2 .
So we have that hpnq  bn where bn is the least element of the set B ztb0 , b1 , ..., bn
1 u.

(ii) h is injective. For, m n hpmq  bm hpnq  bn , which is like saying m  n hpmq  hpnq
h is surjective. Any b P B is equal to a certain hpnq : n P N.

Example Construct a bijection h : A N to show that the set A is countably infinite in each of the following
case:

59
(1) A  Z

Solution:
Let x P Z, and define h : Z N by
$
'
&2n 2 if x 0
hpxq  2n 1 if x 0
'
if x  0
%
2

where n  |x|
(2) A  Q

Solution:
Let q P Z with q  mn where n P Nzt0u, m P Z and hcfpm, nq  1
$
'
&0 if m  0
m n

f p q  p3 qp2 q, if m is negative
m
n '
5p3n qp2m q, if m is positive
%

It is also injective because

m1
n1
 mn 1 n1  n2 or m1  m2 .
1
m1  m2 
Then clearly, h
n1
 h
n2
. It is however not surjective.
We can restrict the image of h to a subset B of the naturals where B  tl P N : l  hpq q, q P Qu. We then
say that h : A B. h is now a bijection as we made it surjective.
B is an infinite subset of the naturals and thus, there exists a bijection g : B N (the one we defined in
the proof above).
Then, g  h : A N is a bijection (proved that in chapter on function).
Yay! Did it!

There is a less rigorous way of defining the bijection between the naturals and the rationals. It is illustrated in
the formal lecture guide.

Proposition 11.2. Nn is countably infinite.

Remark: Nn  tpl1 , l2 , ...., ln q : li P N where 1 i nu

Proof. It is easy to define a bijective map h : Nn B N by

hpl1 , l2 , ...., ln q  2l1  3l  ....  pln

2 n
product of powers of primes

Then we define B  tb P N : b  hpl1 , l2 , ...., ln qu

Using the same argument as for the rationals, there then exists a bijection g : Nn N.
It is possible to adapt the proposition to N8 . The bijection stays the same, but the bijection is a product of
infinitely many powers of primes. This is possible as we know there are infinitely many primes.

Corollary 11.1. Let X1 , X2 , ...., Xn each be countably infinite sets. Then the set X1  X2  ....  Xn is also
countably infinite.

Proof. We know can map each element xi of the vector px1 , x2 , ...., xn q to a natural number li using a bijection.
So there exists bijection h : X1  X2  ....  Xn Nn .
Then we know there is a bijection g : Nn N. Then the composition g  h : X1  X2  ....  Xn N is also a
bijection.

n

Proposition 11.3. Let X1 , X2 , ...., Xn each be countably infinite sets. Then Xk is also countably infinite.

k 1

60
Sketch Proof. We can use the method used in the formal lecture notes to do that. The idea is to list all elements
of the ith set in an n  8 matrix. We can do that because the set is countable.
Xi  txi1 , xi2 , xi3 , ...u

x11 x12 x13 x14 ...

x21 x22 x23 x24 ...
x31 x32 x33 x34 ...
. . . . ...
. . . . ...
. . . . ...
The we perform the counting by zig-zagging through and skipping any recurrent elements.
Proposition 11.4. The set of algebraic numbers A is countably infinite.

Proof. The algebraic numbers are roots of polynomials with rational coefficients P Rrxs. We will risk contradict-
ing our earlier notation by saying (only in this proof) instead that algebraic numbers are roots of polynomials
P Qrxs. We do this for simplicity of notation. The notion of such numbers is the same as that previously
discussed.
We will say that An is the set of algebraic numbers of polynomials P Qrxsn , that is the set of polynomials of
degree at most n.
There is a bijection f : Qrxsn Qn 1 given by f pP q  pa0 , a1 , ..., an q where P pxq  a0 x a1 x ... an xn .
So the set Qrxsn is countable.
An is the union of the sets of (finite) roots of each polynomial in Qrxsn .
So for each polynomial, we have a set RP of at most n roots. We order the roots in some arbitrary manner and
put them in a vector with n components. p1 , 2 , ..., n q. If there are less than n roots, then repeat the last
root in the vector. p1 , 2 , ..., k , k , k q. Note that repeated roots will have different indices from each other.

An  RP
P r sn
P Qx
8

A An

n 0

Clearly, An is countable and the union of all such An is also countable.

Let us try to count the reals in the interval p0, 1q. Note that x P p0, 1q 0 x 1. They will all have decimal
expansions.

a1  0.a11 a12 a13 ...

a2  0.a21 a22 a23 ...
a3  0.a31 a32 a33 ...
.  ....
.  ....
ak  0.ak1 ak2 ak3 ...
.  ....
.  ....

Note that we will not include a sequence that has an infinite recurrence of 9s at the end. (It converges to 1;
refer to Analysis 1)

It seems that by ordering the digits in the decimal expansion in some way, we are able to count the reals in p0, 1q.

Choose a number ak in the above array. Then change the number at akk (switch it with any number from 1 to
8) and name this new number a1kk . Define a new number

a  0.a111 a122 a133 ...a1kk ...

61
No number in the list has the same number at the ith decimal place. So we did not list all the numbers. We
can add this new number to the list and repeat the step. However many numbers we have, we are always able
to generate a new real number. The interval p0, 1q is not countable.

We just showed that:

Proposition 11.5. The set R is not countable.
Theorem 11.1 (Cantor-Bernstein-Schroeder). Let X and Y be sets. If there exists injections f : X Y
and g : Y X, then there exists a bijection h : X Y .

Remark: Note that there is then also a bijection h
1 : Y X

Proof. We will split each set into 3 parts according to how the elements are linked to those of the other set, or
rather, according to how they get created. We will carry the splitting out on Y . The process is similar for X.
We consider an element y0 (the originator) in Y and look for an element x0 in X for which f px0 q  y0 . Now
for this x0 , we look for an element y1 P Y s.t g py1 q  x0 . Now for this element y1 , we look for x1 P X s.t
f px1 q  y1 .
We continue this process whereby for yn P Y we look for xn P X s.t f pxn q  yn and for this xn , we look for
yn 1 P Y s.t g pyn 1 q  xn .

(i) The process may end in X, that is we find an xn for which there is no yn 1 s.t g pyn 1 q xn . For
originators y with this property, we say that y P YX .

Figure 22: y0 P YX
(ii) The process may end in Y itself. That is we eventually find a yn for which there is no xn s.t f pxn q  yn .
For originators y with this property we say that y P YY
(iii) Remember that we may be dealing with infinite sets, so the process may never end. For such originators,
y P Y8

We can do the exact same thing with the set X. What we have then are 6 sets, YX , YY , Y8 , XX , XY , X8 . Note
that there is then a bijection between YY and XY ; YX and XX ; Y8 and X8 .
We then define our bijection h : X Y as follows:

$
' pq
&f x if x P XX
hpxq  g
1 pxq if x P XY
'
f pxq if x P X8
%

Remark: Understanding the existence of the bijections can be tricky. It is a very good mental exercise for you.
However, if ever you are asked a proof during the exams, the above argument should give you your marks.

62
 Figure 23: y0 P YY

Figure 24: y0 P Y8

11.2 Different Infinities

This is probably the most revealing section of this entire module. We already know that the set N is infinite.
Then we showed that we actually can count the naturals but not the reals. Does that mean that there are more
reals than naturals? There are then different infinities!
Definition 11.3. For two sets X and Y we say that pX q pY q if there is an injection X Y
We can then say that if there is an injection X Y and an injection Y X, then pX q  pY q. The
Cantor-Bernstein-Schroeder Theorem strengthens this argument as we know that there is a bijection X Y .
Moreover if there is a surjection A B, then pAq pB q.

We will show the existence of different infinities using the one we are already familiar with, 0  pNq.
For this we introduce the notion of power sets.
Definition 11.4. For a set X, its power set P pX q is the set of all subsets of X.

Example X  tx1 , x2 , x3 u P pX q  tu, tx1 u, tx2 u, tx3 u, tx1 , x2 u, tx1 , x3 u, tx2 , x3 u, tx1 , x2 , x3 uu
We see that P pX q  8  2n where n  pX q  3.

63
 
Proposition 11.6. Let X be a set with pX q  n, then P pX q  2n
Proof. We will do a combinatoric proof of this. We find the number of possible combinations for one subset.
For each element in X, it is either absent or present in the subset. So there are 2 possibilities for each slot.
There are n such slot (the subset can accommodate up to n elements). Then the number of possibilities for the
entire subset is 2n .

We showed that we can build a bigger set from a smaller set by considering its power set. But that was in
the case of finite sets. It still works for infinite sets. That is, given a countably infinite set B, we have that
P pB q pB q. But we cant use 28 (this does not make sense) to justify that, but we can do it using
functions.
Proposition 11.7. Let X be a set. There exists no surjections (hence no bijections) X P pX q
Proof. We will create a contradiction. Let f : X P pX q. If f were a surjection, there @px P P pX q, Dx P
X s.t f pxq  px .
We need to understand the map f first. f maps elements of x to a subset of X. So f pxq X.
Let us define a set P  tx P X : x R f pxqu

Now the fun part. Clearly, P X. If we let f pxq  P , then we have for x P P x P f pxq. But if x P P , then
x R f pxq .
So we have that Ex P X for which f pxq  P

We just showed that the prospect of having pX q P pX q is simply not possible. Moreover, define f : X
P pX q as f pxq  x. We map x onto the subset containing x only.
This is clearly an injection so we have pX q P pX q. We conclude that pX q P pX q.
We end the guide with the coolest statement that you first years will ever here:

Corollary 11.2. There are infinitely many different infinities as 0  pNq pP pNqq pP pP pNqqq ...

64
Keegans Questions
Introduction

I was studying for my January exam over three years ago, and I was wondering: How was it going to be
like? Will I be able to do the paper? Have I done enough practice? These questions were floating through my
mind, and I had few seniors to help me resolve these questions. Last year, I thought I would alleviate some
questions for the year 1s doing their first ever University exam. And so, I wrote this compilation of past year
exam paper questions, meant as a easy reference guide, and as a means to test yourself - the exam will consist
of questions similar to this standard. Thanks goes to the numerous seniors who have loaned me their exam
papers. Consider this as a Christmas present. Merry Christmas :)

Disclaimer: I do not guarantee that the format of the exam will remain the same, nor that successfully being
able to do all the questions within will net you a First Class, but it will aid you in understanding the concepts
covered during the Foundations and Analysis lectures.

PS. Feel free to distribute this to other Year 1s who want this - they dont need to be a member of the
MORSE Society. However, do encourage them to join if they are not - firms look at societys memberships at
the start of the year, not before socials / printing of revision guides - so how much everyone gets subsidized is re-
ally due to membership levels! However, I do not want this to be uploaded on TSR or any other forums. I am shy.

Yours Sincerely, Keegan Kang

65
12 Foundations

This exam paper usually consists of four questions. You have to do three questions, being Question 1 (which
is compulsory), and another two out of the remaining Questions 2, 3, and 4. The trend in the last three years
focuses on the following three topics for the three non compulsory questions, namely: Polynomials, Equivalence
Relations, and Groups/Isomorphisms. However, all (other) possible questions will be listed here, just in case.

For the last several years, the first question usually comprised of many subquestions, each testing the gen-
eral concepts and understanding of the entire module. Here are questions which have come out over the years.
The same numerical questions will not be repeated, *but* they will be in the same format as those below.
Some subquestions from other questions are also put in here, because it just, *just* might come out. Generally,
definitions may come out here as well, and it is good to know them.

12.1 Basics

Question 12.1. State the Well Ordering Principle.

Question 12.2. State the Fundamental Theorem of Arithmetic
Question 12.3. If a, b, c P N, what does the notation a|b mean? Show that if a|b and b|c then a|c.
Question 12.4. Let h  hcft245, 52u. Find h and write it as an integer linear combination of 245 and 52.
Question 12.5. Find h  hcfp750, 231q and then find a, b P Z such that h  750a 231b.
Question 12.6. Find h  hcfp1900, 2011q, and then find a, b P Z such that h  1900a 2011b.
m
Question 12.7. Express the repeating decimal 0.11232323 . . . as a rational number in the form where
n
m, n P N. (You do not have to simplify your answer.)

12.2 Subgroups

Question 12.8. For n, m P N define the sets nZ and nZ mZ. Which of the following three sets are subgroups
of Z? (i) 5Z X 3Z; (ii) 5Z 3Z; (iii) 5Z Y 3Z. In each case where the set is a subgroup, find an n P Z such that
the subgroup is equal to nZ.
Question 12.9. 7Z X 3Z is an additive subgroup of Z. Which additive subgroup is it?
Question 12.10. 8Z 12Z is an additive subgroup of Z. Which additive subgroup is it? (Give a shorter
answer than 8Z 12Z!q
Question 12.11. 9Z 15Z is an additive subgroup of Z. Which additive subgroup is it? (Give a shorter
answer than 9Z 15Z!q
Question 12.12. What is the smallest positive integer (i) in 7Z 9Z? (ii) 7Z X 9Z?

12.3 Sets and Venn Diagrams

Question 12.13. State (without proof) De Morgans Laws for sets A, B and their complements.
Question 12.14. Draw a Venn diagram illustrating the fact that AzpB zC q  pAzB qzC for some choices of sets
A, B and C.
Question 12.15. The set (AzB qzC is equal, for all sets A, B and C, to one of the following sets; which one?
AzpB X C q; AzpB zC q; AzpB Y C q.
Question 12.16. For all sets A, B and C, the set pAzB q X pAzC q is equal to one of the following sets; which
one? AzpB Y C q; AzpB X C q; A Y pB X C q; A X pB Y C q.
Question 12.17. Draw a truth table for (P ^ Qq _ p P ^ Qq and hence demonstrate that this proposition is
logically equivalent to Q.
Question 12.18. Which if any, of the following three statements are logically equivalent to the statement
P pP Qq?
a) pP P q Q
b) Q P
c) P Q

66
Question 12.19. Which of the propositions (i)-(iv) is logically equivalent to P Q?
(i) pP _ Qq ; (ii) P _ p Qq ; (iii) p P q _ Q ; (iv) p P q _ p Qq.
Question 12.20. Which of the propositions (i)-(iv) is logically equivalent to P Q?
(i) pP _ p Qqq; (ii) P _ Q; (iii) p P q _ p Qq; (iv) p P q _ Q.
Show truth tables to prove that your answer is correct.
Question 12.21. Which of the propositions (i)-(iv) is logically equivalent to p P Qq?
(i) P Q; (ii) P _ Q; (iii) p P q _ p Qq; (iv) p P q ^ p Qq.
Show truth tables to prove that the one you select is equivalent to p P Qq.
Question 12.22. One of the following propositions (i)-(iv) is logically equivalent to pa bq. (i) a ^ b (ii)
b a (iii) a _ b (iv) b a.
Which one is it? Use a truth table to justify your answer.
Question 12.23. One of the following propositions (i) - (iv) is logically equivalent to a b.
(i) a ^ b; (ii) p aq _ b; (iii) p bq _ a; (iv) p aq ^ b.
Which one is it? Use a truth table to justify your answer.

12.4 Mathematical Induction

n
Question 12.24. Prove by induction that k  21 npn 1q for all n P N.

k 1

Question 12.25. Prove by induction 1 3 5 ... p2n
1q  n2 .

12.5 Remainder Theorem

Question 12.26. State the Remainder Theorem for polynomials.

Question 12.27. Which of the factors px 4q, px
2q and px
1q divide the polynomial x4
10x2 21x
12
without remainder?
Question 12.28. Which of the factors px 2q, px
3q and px
1q divide the polynomial x4
5x3
13x2 77x
60
without remainder?
Question 12.29. Which of the factors px 2q, px 3q, and px
2q divide the polynomial x4
3x3
15x2 19x 30
without remainder?
Question 12.30. Which of the factors p2x 1q, p2x
1q and px 4q divide the polynomial 2x4 9x3 2x2
9x
4
without remainder?
Question 12.31. Which of the factors x
2, x 1 and x
3 divides the polynomial P pxq  x3 12x2 9x
2
without remainder?
Question 12.32. For a certain value of the integer a, the polynomial P  2x3
3x2 ax
6 is divisible
(without remainder) by x
1. What is the remainder when P is divided by x
2? [Hint: find the value of a.]

12.6 Pascals Triangle

Question 12.33. Prove, for integers 1 k n, that





n
k
1
n
k
 n
k
1

You may assume that

n
k
 k!pnn!
kq! .
Question 12.34. For n a positive integer and 0 k n, what is the coefficient of xk in the expansion of the
polynomial pa bxqn ? Hence prove that the alternating sum of the numbers in the nth row of Pascals Triangle


n
is zero. [In other words, show that p
1q k  0.]
k n

k 0

Question 12.35. Prove that the sum of the nth row of Pascals triangle is 2n . [In other words, show that
n 

n
k
 2n .]

k 0

67
12.7 Injections and Surjections

Question 12.36. Is the following statement always true? Justify your answer. If f : A B and g : C D
are surjections where Impf q C, then g  f : A D is a surjection.
Question 12.37. For each of the following pairs of sets, give a bijection between them [Note: you do not need
to show that the function you have constructed is a bijection]: (i) the interval p0, 1q to R (ii) N to Z.
Question 12.38. For a function f : A B, define what it means for f to be surjective, injective, and hence
bijective.
Question 12.39. Give examples of well-defined functions f : Z Z which are:
(i) surjective but not injective;
(ii) injective but not surjective;
(iii) neither injective nor surjective;
(iv) bijective
Question 12.40. Give examples of well-defined functions f : N N which are:
(i) surjective but not injective;
(ii) injective but not surjective;
(iii) neither injective nor surjective;
Question 12.41. Define f : Q N (where Q is the positive rationals) by f prq  2m 3n where r  m
n
and
hcftm, nu = 1. Show that f is an injection.
Question 12.42. Give an example of an injection h : N Q .
Question 12.43. Give an example of an injection Q N.
Question 12.44. Give an example of an injection N  N  N N.
Question 12.45. Let A and B be sets. True or false: If there is an injection f : A B and a surjection
g : A B, then there exists a bijection h : A B. Explain your answer.
Question 12.46. Give examples of well-defined functions f : R R which are:
(i) surjective but not injective;
(ii) injective but not surjective;
(iii) neither injective nor surjective;

12.8 Counting

Question 12.47. If A  ta, b, cu, list the elements of P pAq.

Question 12.48. Let A and B be two (possibly infinite) sets. What precisely does it mean to say that
|A| |B |?
Question 12.49. What precisely does it mean to say that the cardinality of two infinite sets, A and B, is
equal?
Question 12.50. What precisely does it mean to say that |N|  |Q| but that |N|  |R|?
Question 12.51. If A and B are two (possibly infinite) sets, and there is a surjection A B, then is there an
injection B A? Answer with a proof or a counterexample.
Question 12.52. For a set S, define what it means for a collection of subsets tAi u to be a partition of S.
Question 12.53. State Schroder-Bernsteins theorem.
Question 12.54. Let A  t1, 2, 3, 4, 5, 6, 7, 8u. How many elements does P pAq (the set of all subsets of A)
have? How many elements of P pAq have 5 elements?

12.9 Samir

Question 12.55. State Fermats Little Theorem.

Question 12.56. If n, n 2, n 4 are all prime, tn, n 2, n 4u is called a prime triple. Prove that there is
only one prime triple.
Question 12.57. Show that if x and y are integers and x3 6  y 2 then x is odd.
Question 12.58. Calculate 1718 mod 19.
Question 12.59. Calculate 1522 mod 11.
Question 12.60. Calculate 1517 mod 17. Briefly explain your method if it is clever.
Question 12.61. Calculate 1012 mod 11. Explain your answer.

68
13 Analysis

This exam paper usually consists of four questions. You have to do three questions, being Question 1 (which
is compulsory), and another two out of the remaining Questions 2, 3, and 4. Unlike Foundations, the optional
three questions (to choose two from) all follow the same style, one on series, one on sequences, and the last on
boundedness, infinum, suprenums and so on.

The first question is comprised of many subquestions in order to test how well you know the Analysis syl-
labus. Be prepared to answer simple T/F questions, and give simple examples, work out toy problems. You
actually should be able to answer most of them instantly at this point in time without having to refer to any
textbooks at all.

13.1 True / False Questions

Say whether each of the following statements is true or false. Give a brief explanation or counter example for
those that are false.
Question 13.1. If pan q has a convergent subsequence then pan q is bounded.
Question 13.2. pan q is strictly increasing pan q 8.
Question 13.3. There exists a sequence pan q for which p|an |q doesnt converge but pan q does.
Question 13.4. Every convergent series is either absolutely convergent or conditionally convergent.
Question 13.5. The following is correct as a definition of pan q is not null:

D  0 s.t. @ N P N, D n N s.t. |an | 

Question 13.6. If a sequence pan q converges to a then we must have:

|an 1
a| |an
a| @ n P N
8
8

Question 13.7. If an 0 for all n P N and an is convergent then an is absolutely convergent.

n 1 
n 1

Question 13.8. If p|an |q is a Cauchy sequence then pan q is bounded.

Question 13.9. If pan q has two different subsequences which both converge to a, then pan q converges to a.
8 1

Question 13.10. If psn q is the sequence of partial sums of , then psn q 0 as n 8.
 n
n 1

8
8
8 a

Question 13.11. If an and bn both converge, and bn 0 for all n P N then n
converges.

n 1 
n 1  bn
n 1

8
8
8
8

Question 13.12. Suppose an bn for all n 1. If bn converges then an converges and an bn .

n 1 
n 1 
n 1 
n 1

Question 13.13. If pan q is increasing and bounded above then p
an q converges.
Question 13.14. pan q converges if and only if p|an |q converges.
Question 13.15. If x P Q and x 0 then p2

?xqp2 ?xq P Q.
Question 13.16. There exists a Cauchy sequence which doesnt converge.


Question 13.17. ?5n 1 is a Cauchy sequence.

Question 13.18. The series 1
? ?1 ?1
?1

1
... can be rearranged to converge to any real number.
2 3 4 5

Question 13.19. If pan q is a sequence such that 1 for every n P N then pan q 0.
an 1
an
Question 13.20. pp3n 5n q n q tends to 3 as n tends to infinity.
1

Question 13.21. pp2n 5n q n q 1

1

69
Question 13.22. If a sequence pan q converges to a then |an
a| |an
a| for all natural numbers n.
1

Question 13.23. If x and y are real numbers such that x y
 for all  0 then x y.

Question 13.24. If pan q is a sequence such that |an 2
an 1 | 21 |an 1
an | for all natural numbers n then
pan q converges.
Question 13.25. If pan q is a sequence such that a2n  0 for all natural numbers n then pan q converges if and
only if pa2n
1 q tends to 0.
8
8
8

Question 13.26. If both an and bn diverge to infinity then an bn diverges to infinity.
n 1   n 1 n 1 
8
8
8

Question 13.27. If an  0 and bn  0 then pan bn q  0.

n 1 n 1  
n 1

Question 13.28. If (an ) is an increasing sequence which is bounded above then pan q tends to suptan | n P Nu
as n tends to infinity.
Question 13.29. The total of the sum:

n
p
1qi 1


i 1
i
depends on the order in which the terms are added.
8 1

Question 13.30. The sequence (sn ) of partial sums of the series is Cauchy.
 n
n 1

Question 13.31. If pan q is an increasing sequence such that for all C 0 there exists a natural number, n,
such that an C
1, then pan q tends to infinity.
Question 13.32. For every real number a there exists a sequence (an ) of irrational numbers such that lim an
n8

a.
8

Question 13.33. Let pan q be a positive decreasing null sequence. The series p
1qn an can always be

n 1
rearranged so that it converges to 1.
Question 13.34. If a sequence pan q does not converge to 0, then there exists  0 such that |an |  for every
n.
8
Question 13.35. Let f be a positive decreasing function such that f pxq dx is finite and let pan q be a
0
sequence such that an f pnq for every n. Then, the series an converges.
Question 13.36. The inequality |x
y | |x| |y | holds for every two real numbers x and y.
Question 13.37. Given any sequence pan q, the sequence sinpan ) contains a convergent subsequence.
Question 13.38. If (1/an q2 0, then an 8.
Question 13.39. A non-monotonic sequence is always bounded.
Question 13.40. The limit of the sequence p0.9, 0.99, 0.999, . . .q is the largest number strictly less than 1.
Question 13.41. If an 1 and an a then a 1.
Question 13.42. Every bounded monotonic sequence has a convergent subsequence.
Question 13.43. Every sequence contains a monotonic subsequence.
Question 13.44. e has a non-recurring decimal expansion
Question 13.45. If for each  0 there exists an N such that |aN
a| , then pan q a.
Question 13.46. Three of the sequences defined as follows tend to infinity:
$ n
" " " &
an  0
n
n odd
n even
an  n
n
n odd
1 n even
an  ?n nn even
n odd
an % 2
n
n odd
n even
4
Question 13.47. If for each  0 there exists an N such that |an
an 1 |  for all n N then pan q converges.

70
Question 13.48. Every Cauchy sequence is bounded.
Question 13.49. Every convergent sequence is a Cauchy sequence.
Question 13.50. Given a non-empty bounded set of real numbers A, inf A a for all a P A.


8n n4
Question 13.51. The sequence
4n4 32n
0.
Question 13.52. Every real number has exactly two distinct decimal expansions.

Question 13.53. If an converges then |an | converges.

Question 13.54. If the sequence pan q converges then an also converges.

Question 13.55. If
an 1
an
0 then an converges.

8
p
?1qn
Question 13.56. The series can be rearranged to sum to 42.

n 1
n

Question 13.57. If an diverges, then p
1qn an also diverges.
n n

Question 13.58. Every decreasing sequence is bounded above.

Question 13.59. If pan q 0 then ( q 8.

1
an
Question 13.60. If a2n 1 for all n and an a then a 1.
Question 13.61. If pan q satisfies an  3 for infinitely many n then pan q 3.
Question 13.62. Every bounded sequence is Cauchy.
Question 13.63. If pan q is unbounded then |an | 8.
Question 13.64. The sequence sinpn!) has a Cauchy subsequence.
Question 13.65. Every increasing sequence which is bounded above is Cauchy.
Question 13.66. If lim pa2n q  lim pa2n q then (an ) converges.
n 8 n 8 7

Question 13.67. If lim suppan q  10 then an 10 eventually.

n 8
Question 13.68. If A is a bounded non-empty set of real numbers such that a 1 for all a P A then inf A 1.
8
8

Question 13.69. If an converges then a2n converges.

n 1 n 1 
8

Question 13.70. If an converges then an 0.
n 1 
8

Question 13.71. If
an 1
12 then an converges.
an n 1 
8

Question 13.72. If lim inf pan q  lim suppan q then an diverges.

n 1

8

Question 13.73. If pan q is strictly contracting then |an 1
an | converges.

n 1

Question 13.74. There can exist a Cauchy sequence pan q in R, which does not converge in R.
Question 13.75. A bounded and monotonic subsequence is always convergent.
8 sin n!
p q is convergent.
Question 13.76. The series
 n 1
10n

8 n!
Question 13.77. The series is convergent.
n1
10n

71
Question 13.78. The series 1
1


1 1 1 1
... . . . can be rearranged to converge to any real number.
2 2 n n
8
8

Question 13.79. The convergence of |an | implies that an is convergent.

n 1 
n 1

Question 13.80. The sequence cos n has a convergent subsequence.

Question 13.81. The sequence psinpnq cospnqq has a convergent subsequence.

Question 13.82. If pan q 0 for all n P N and an a then a 0.

8
8

Question 13.83. If an  1, then an  1.

n 1 
n 2

Question 13.84. There exists a sequence which has no bounded subsequence.

8

Question 13.85. The sequence an 0 if, and only if, the series an converges.

n 1

8

Question 13.86. The series ?1n converges.
n 1
8

Question 13.87. The series an converges if its sequence of partial sums (sn ) is bounded.
n 1
Question 13.88. If A is a non-empty set of real numbers such that a 3 for all a P A then sup A 3.
8
8

Question 13.89. There exists a series an which converges but for which |an | diverges.
n 1  
n 1

13.2 Give an Example Questions

Give an example of each of the following (no proofs needed):

Question 13.90. A bounded sequence which doesnt converge.

Question 13.91. A Cauchy sequence.

8
8
?a
Question 13.92. A series an which converges, but such that 3
n diverges.

n 1 
n 1

8

Question 13.93. A series an with sequence of partial sums psn q such that sn  an for all natural numbers
n 1 
n.
Question 13.94. An unbounded sequence that has a convergent subsequence.

Question 13.95. An unbounded sequence that has no convergent subsequence.

Question 13.96. A non-monotonic, unbounded sequence.

Question 13.97. A null sequence pan q such that the series an does not converge.
Question 13.98. A sequence which is not a Cauchy sequence but has the property that for every  0 and
every N 0 there exists n N and m 2n such that |an
am | .
Question 13.99. Two sequences pan q and pbn q such that the sequence pcn q defined by cn  an bn converges
to 1, but nether pan q nor pbn q converge.
Question 13.100. A sequence pan q that tends to 8 but is neither increasing nor eventually increasing.
Question 13.101. A sequence pan q such that tends to 1 but pan q doesnt converge.
an 1
an

Question 13.102. A bounded sequence pan q such that tends to 1 but pan q doesnt converge.
an 1
an

Question 13.103. A sequence of the form pxn y n q n which converges to 6 as n 8.

1

72
Question 13.104. A sequence of irrationals which has a subsequence converging to 2 and a subsequence
converging to 3.
Question 13.105. A sequence which is not increasing, but tends to infinity.
Question 13.106. A sequence bounded above which does not have a convergent subsequence.
Question 13.107. A sequence which is not bounded above but which doesnt tend to infinity.

Question 13.108. A number with only one decimal expansion.



Question 13.109. A sequence (an ) of non-zero terms such that

an 2
an
1 but (an ) does not converge.
Question

13.110. A sequence (an ) of non-zero terms which is bounded above but such that the sequence
1
is not bounded below.
an
Question 13.111. A sequence pan q which tends to infinity but has an 1 an for infinitely many n.
 
 3n 10 
Question 13.112. A natural number N such that for all n
3 100
1
N,  .
n
 
 n 10 
Question 13.113. An integer N so that 
n

100
1
1whenever n N .

8
8

Question 13.114. A convergent series an with all terms non-zero and an  7 can be rearranged so
n 1  
n 1
that its sum is 7.
8
8

Question 13.115. A convergent series an such that p
1qn 1
an does not converge.
n 1  
n 1

Question 13.116. A sequence an such that, for all m P N, an  m for infinitely many natural numbers n.
1
Question 13.117. A series an , with non-zero terms, which can be rearranged to form 3 .
2

Question 13.118. A series an with an 0 and with sn 8 where psn q is the sequence of partial sums.

Question 13.119. Two series an and bn which diverge to infinity but where an bn converges.

Question 13.120. A sequence pan q which satisfies the following but does not diverge to infinity:

@ C 0, D N s.t. D n N with an C
Question 13.121. A sequence pan q with 0, but an 8.
1
an
Question 13.122. A sequence pan q which satisfies the following but does not converge to 1:

@  0, @ N, D n N with |an
1| 
Question 13.123. A geometric series that neither converges nor diverges to infinity.
Question 13.124. A Cauchy sequence pan q such that an is rational for all n 1, and such that its limit is not
rational.

Question 13.125. A sequence that satisfies the conditions of the Ratio Lemma, and explain your answer.

Question 13.126. A sequence pan q such that 1 for all n 1, and such that an does not tend to 0.
an 1
an
Question 13.127. A set A which has no supremum. Explain why it does not violate the completeness axiom.
8

Question 13.128. Give an example of a diverging series an such that
an 1
1 as n 8.

n 1
an
8

Question 13.129. Give an example of a converging series an such that
an 1
1 as n 8.
n 1 an

73
13.3 Toy Questions

Question 13.130. Define what it means for pan q to be a null sequence.

Question 13.131. Define what it means for a sequence pan q to be bounded.
Question 13.132. For what values of x is the following inequality valid:

|x 2| 5

Question 13.133. Let pan q be a sequence such that |an | 2
n for every n 0. Find a value N such that
8

| an | 10
3 .

n N

Question 13.134. State, without proof, whether the sequence tanpnq is convergent, bounded or monotone.
cos n
2
sin n
Question 13.135. State without proof whether the sequence is convergent, bounded, or monotone.

Question 13.136. Show that every sequence which is not bounded above has an increasing subsequence.
Question 13.137. Show that if lim an
n 8  l and l 0, then an is eventually positive.
Question 13.138. If pan q a and an 0 for all n 0, then is it true that a 0? Justify your answer.
Question 13.139. Show that if pan q 0 then either there is a subsequence of p a1 q which tends to 8, or
n

there is a subsequence of ( q which tends to
8.

1
a n

Question 13.140. If a is irrational and b  0 is an integer, is

a
rational or irrational? Justify your answer.
b
Question 13.141. What is the value of lim p3n 2n q n ?
1
5n
n 8
Question 13.142. Find the value of inf tcos x | 0 x u.
Question 13.143. Compute the limit of the sequence:


2pn!q 3n
4n n!

using any rules of algebra of limits and standard limits.

Question 13.144. Find real numbers a, b such that:

tx P R | x
3, x 5u  ty P R | |y
a| bu
Question 13.145. Two of the following sequences eventually sandwich the third:




1 1 1
n3
,
36n
,
p2nq2
Write down the corresponding inequality in the form an bn cn for all n N and find the smallest value of
N for which it is true.
Question 13.146. If pan q converges to a, and an
1, for all n 1, then show that a 1.
Question 13.147. State without proof the possible limits of the sequence pxn q for x 0.
?
Question 13.148. Consider the recursive sequence an 1  1 an , a1  1. Assuming that the sequence
converges, what is the limit?
b
Question 13.149. Let the sequence pan q be defined according to a1  1 and an  3an
1
a2n
1 for n 2.
Assume that pan q tends to a 0. Find the numerical value of a. Justify your answer.
"  *
1 
Question 13.150. Find inf
tan x  4
x
2
8

Question 13.151. State without proof for which a P R the series an converges.
n 1 

74

cos n 8

Question 13.152. Does there exist a rearrangement pbn q of pan q  such that bn  ?
n 
n 1

M
Question 13.153. Assume that one knows that 0 e
1
pM 2 1q! , for M 1. Find the smallest
n0
n!

N
N 1 such that the error in approximation e  1
is less than
1
.
n0
n! 50

Question 13.154. Assume that pan q is a decreasing sequence. Which of the following statements are true:

1. pan q has a limit.

2. pan q is eventually negative.
3. pan q tends to minus infinity.

4. pan q tends to zero.

Question 13.155. State, without proof, whether the following sequences tends to A) infinity, B) minus infinity,
C) zero, D) a constant c  0 or E) does not tend to either infinity, minus infinity, or any real number.

n2
1. pan q  ,
2n


2n2 sinpnq
2. pan q  ,
5n2 3n 17


3. pan q  10n
2 ?n 1
n
,

4. pan q  pen cospnqq.

Question 13.156. Prove that inf ta2n : n P Nu  0 for any null sequence pan q.
Question 13.157. An increasing sequence pan q satisfies an 4 for every n P N. State, without proof, whether
pan q must converge.
Question 13.158. State the Bolzano - Weierstrass Theorem and use it to decide whether sequence pcosp2n qq
has a convergent subsequence.
 
 n2 
Question 13.159. Find a value N such that  2
n 1

1 10
3 for every value of n N .
8

Question 13.160. Find a series an with every an 0 which converges to 3.

n 1

Question 13.161. Find a sequence pan q that satisfies |an 1
an | |an
an
1 | for every n 1 but pan q is
not Cauchy.
Question 13.162. Prove that every Cauchy sequence is bounded.
8

Question 13.163. Compute n
.
n 1  3
8 n
sin2 n
Question 13.164. Is finite or infinite? Justify your answer.
 n 1
n2 cos n

8 p2nq!
Question 13.165. Is finite or infinite? Justify your answer.
n1
n2n

75
14 Conclusion

If you can do all of the questions listed above, as well as successfully writing out the proofs, you shouldnt have
a problem passing the exam, and might *just* might, achieve your First. Good luck! The best is yet to be :)

76