Introduction to

Algebraic Structures
Lecture Notes 2019

Janko Böhm, Magdaleen Marais

November 6, 2019
Contents

0 Introduction 1

1 Fundamental constructions 4
1.1 Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Mathematical induction . . . . . . . . . . . . . . . . 8
1.3 Relations . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.5 Partial orderings and equivalence relations . . . . 18
1.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 21

2 Numbers 27
2.1 The integers and division with remainder . . . . . 27
2.2 Fundamental theorem of arithmetic . . . . . . . . . 32
2.3 Greatest common divisor and Euclidean algorithm 35
2.4 The Chinese remainder theorem . . . . . . . . . . . 38
2.5 Prime factorization . . . . . . . . . . . . . . . . . . . 41
2.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 43

3 Groups 47
3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2 Groups and actions . . . . . . . . . . . . . . . . . . 49
3.2.1 Basics . . . . . . . . . . . . . . . . . . . . . . 49
3.2.2 Group actions . . . . . . . . . . . . . . . . . 61
3.2.3 Action by translation . . . . . . . . . . . . . 73
3.3 Normal subgroups . . . . . . . . . . . . . . . . . . . 79
3.3.1 Normal subgroups and the quotient group 79
3.3.2 Homomorphism theorem . . . . . . . . . . . 83
3.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 85

1
CONTENTS 2

4 Rings and fields 93

4.1 Basics . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.2 The group of units of Z/n . . . . . . . . . . . . . . 96
4.3 Ideals and quotient rings . . . . . . . . . . . . . . . 99
4.4 Integral domains and fields . . . . . . . . . . . . . . 102
4.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 104

5 Vector spaces 107

5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.2 Gauß algorithm . . . . . . . . . . . . . . . . . . . . . 108
5.3 Vector spaces and bases . . . . . . . . . . . . . . . . 113
5.4 Dimension . . . . . . . . . . . . . . . . . . . . . . . . 119
5.5 Vector space homomorphisms . . . . . . . . . . . . 121
5.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 126
List of Figures

1 Knot . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Four points. . . . . . . . . . . . . . . . . . . . . . . . 2

1.1 Complement . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 Union . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Intersection . . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Graph of parabola. . . . . . . . . . . . . . . . . . . . 13
1.5 Hyperbola . . . . . . . . . . . . . . . . . . . . . . . . 14
1.6 A non-injective map . . . . . . . . . . . . . . . . . . 14
1.7 A bijektive map and its inverse map. . . . . . . . . 16
1.8 Square root . . . . . . . . . . . . . . . . . . . . . . . 16
1.9 Identity map R → R . . . . . . . . . . . . . . . . . . 18
1.10 Equivalence classes . . . . . . . . . . . . . . . . . . . 21
1.11 Towers of Hanoi . . . . . . . . . . . . . . . . . . . . 24
1.12 How many shortest paths are there from A to B. 24

2.1 Two configurations of gearwheels. . . . . . . . . . . 46

3.1 The Platonic solids . . . . . . . . . . . . . . . . . . . 48

3.2 Composition of two symmetries of the tetrahedron. 48
3.3 Rotational symmetry of the tetrahedron . . . . . . 52
3.4 Reflection symmetry of the tetrahedron . . . . . . 53
3.5 residue classes modulo 3 . . . . . . . . . . . . . . . 55
3.6 Exponential function . . . . . . . . . . . . . . . . . . 58
3.7 Example of a motion of R2 . . . . . . . . . . . . . . . 63
3.8 Tetraedron . . . . . . . . . . . . . . . . . . . . . . . . 88
3.9 Regular 5-gon . . . . . . . . . . . . . . . . . . . . . . 89
3.10 Tetrahedron with diagonals connecting mid-points
of the edges . . . . . . . . . . . . . . . . . . . . . . . 91
3.11 Icosahedron with numbering of the vertices . . . . 92

3
LIST OF FIGURES i

5.1 Cubic polynomials with zeroes −1 and 2 and in-

flection point at 0. . . . . . . . . . . . . . . . . . . . 112
5.2 Half plane . . . . . . . . . . . . . . . . . . . . . . . . 116
5.3 Parabola . . . . . . . . . . . . . . . . . . . . . . . . . 117
List of Symbols

m∉M m is not an element of M . . . . . . . . . 4

m∈M m is an element of M . . . . . . . . . . . . 4
N natural numbers . . . . . . . . . . . . . . . 5
N0 natural numbers including 0 . . . . . . . . 5
Z integers . . . . . . . . . . . . . . . . . . . . 5
Q rational numbers . . . . . . . . . . . . . . . 5
∣ with . . . . . . . . . . . . . . . . . . . . . . 5
⇒ from which it follows that . . . . . . . . . 5
⇔ if and only if . . . . . . . . . . . . . . . . . 5
M /N complement of N in M . . . . . . . . . . . 5
M ∪N union of N and M . . . . . . . . . . . . . . 5
M ∩N intersection of N and M . . . . . . . . . . 5
∀ for all . . . . . . . . . . . . . . . . . . . . . 7
∃ there exists . . . . . . . . . . . . . . . . . . 7
∣M ∣ number of elements of M . . . . . . . . . . 7
M ×N cartesian product of M and N . . . . . . 7
2M power set of M . . . . . . . . . . . . . . . . 8
n
∑k=1 ak sum . . . . . . . . . . . . . . . . . . . . . . . 9
n
∏k=1 ak product . . . . . . . . . . . . . . . . . . . . 9
f (A) image of A under the map f . . . . . . . . 12
f −1 (B) preimage of B under the map f . . . . . . 12
Graph(f ) graph of f . . . . . . . . . . . . . . . . . . . 12
f −1 inverse of the bijective map f . . . . . . . 15
∃1 there exists a unique . . . . . . . . . . . . 15
(nk) binomial coefficient . . . . . . . . . . . . . 23
b∣a b divides a . . . . . . . . . . . . . . . . . . . 31
a ≡ b mod m a congruent to b modulo m . . . . . . . . 32
π (x) number or prime numbers less or equal to x 34
ggT greatest common divisor . . . . . . . . . . 35

ii
LIST OF SYMBOLS iii

lcm least common multiple . . . . . . . . . . . 35

S (X) group of self-mappings of X . . . . . . . . 51
Sn symmetric group . . . . . . . . . . . . . . . 51
G1 × G2 cartesian product of G1 and G2 . . . . . . 52
Z/n group of residue classes . . . . . . . . . . . 54
Zn group of residue classes . . . . . . . . . . . 54
ker ϕ kernel of ϕ . . . . . . . . . . . . . . . . . . 56
Im ϕ Image of ϕ . . . . . . . . . . . . . . . . . . 56
⟨E⟩ subgroup generated by E . . . . . . . . . . 60
ord (g) order of g . . . . . . . . . . . . . . . . . . . 61
E (n) group of Euklidean motions . . . . . . . . 63
Sym (M ) symmetry group . . . . . . . . . . . . . . . 63
Gm orbit of m under the action of G . . . . . 66
Stab (N ) stabilizer of the set N unter the action of
G . . . . . . . . . . . . . . . . . . . . . . . . 66
Stab (m) stabilizer of m under the action of G . . . 66
[G ∶ H] index of the subgroup H ⊂ G . . . . . . . . 77
deg (f ) degree of the polynomial f . . . . . . . . . 95
ϕ(n) Euler Phi-function, n ∈ N . . . . . . . . . . 97
char (K) characteristic of K . . . . . . . . . . . . . . 104
ϕΩ linear combination map with respect to Ω 119
dim V dimension of V . . . . . . . . . . . . . . . . 120
M∆Ω (F ) Darstellende Matrix von F bezüglich der
Basen Ω und ∆ . . . . . . . . . . . . . . . . 125
LIST OF SYMBOLS iv
0

Introduction

In this course, we will focus on the foundations of algebra, in-

cluding linear algebra. We will also discuss some very simple,
but nevertheless fundamental facts from number theory. Algebra
and number theory are very closely related areas of pure math-
ematics, complementing analysis, combinatorics1 , geometry and
topology2 .
What is number theory? As the name says, number theory is
studying the properties of the integer numbers (..., −1, 0, 1, 2, 3, ...),
in particular the relation of addition and multiplication. Many
number theory problems are easy to formulate, but very difficult
to solve. The most prominent example is Fermat’s last theorem
from 1637: For n ≥ 3 there are no (non-trivial) integer solutions
1
Using combinatorics one can, for example, compute that in the standard
lotto game there are (49
6
) ≈ 14 000 000 possible results.
2
In topology one can see, for example, that the knot in Figure 1 cannot
be untangled without cutting the string.

Figure 1: Knot

1
0. INTRODUCTION 2

of the equation
xn + y n = z n .
Fermat’s last theorem was only proven 1995 (by A. Wiles) af-
ter 350 years of work of many mathematicians, which involved
introducing various new concepts in mathematics. Today, there
are close connections of number theory to, for example, algebraic
geometry, combinatorics, cryptography and coding theory.
What is algebra? Algebra is a very diverse area of mathemat-
ics, which discusses basic structures which are of key importance
in all fields of mathematics, like groups rings and fields. That is,
algebra studies the question, how one can introduce operations
on sets, like the addition and multiplication of integer numbers.
By combining methods from algebra and number theory, one
can construct, for example, public key cryptosystems. Another
connection of algebra and number theory arises from algebraic
geometry, which studies solution sets of polynomial systems of
equations in several variables3 .
The simplest (but in practice the most important) special
case are linear systems of equations over a field K (for example,
K = Q, R, C the field of rational, real or complex numbers), the
For example, the common solution set of x2 + 2y 2 = 3 and 2x2 +
3

y = 3, that is, the intersection of two ellipses, consists out of 4 points

2

(1, 1), (−1, 1), (1, −1), (−1, −1), see Figure 2.

y 0

–1

–2 –1 0 1 2
x

Figure 2: Four points.

0. INTRODUCTION 3

core topic of linear algebra. Here, we solve systems

a1,1 x1 + ... + a1,m xm = b1

⋮
an,1 x1 + ... + an,m xm = bn

with aij ∈ K, bi ∈ K for xj ∈ K (with i = 1, ..., n and j = 1, ...m).

As an application of linear algebra, we will discuss the Google
page rank algorithm.
Let us also comment on an other special case, which however
goes beyond the topics discussed in this course, that is, poly-
nomial equations of higher degree in a single variable x. For
example, one can ask for the solution set of the quadratic equa-
tion
ax2 + bx + c = 0.
The solutions can be described, using radicals, as
√
−b ± b2 − 4ac
x= .
2a
In a similar way, one can find expressions in terms of radicals
for equations of degree d = 3 (Tartaglia 1535, Cardano 1545) and
d = 4 (Ferrari 1522), for d ≥ 5 the solutions can, in general, not
be written in terms of radical any more. An important subtopic
of algebra, the Galois theory, discusses when this is possible.
1

Fundamental constructions

In this section, we discuss fundamental construction, which are

used to construct from given mathematical objects new math-
ematical objects. Starting out with the notation of a set, we
discuss how two given sets can be related to each other. In par-
ticular, we will discuss maps and equivalence relations.

1.1 Sets
Definition 1.1.1 (Cantor) A set is a collection of definite,
distinct objects m, concrete or imaginary, thus forming a new
object M .

If m is an element of M , we write

m ∈ M,

if not, then m ∉ M . We write the set M with elements m1 , m2 , ...

as
M = {m1 , m2 , ...} .
We call the set with no elements the empty set ∅ = { }.

Remark 1.1.2 The definition we interpret as follows: Objects

are mathematical objects and the collection of objects into a set,
is a new mathematical object. By the term distinct, we mean
that we can decide for any two objects in the set, whether they
are equal or not.

4
1. FUNDAMENTAL CONSTRUCTIONS 5

Example 1.1.3 The following sets of numbers are examples of

sets: The numerical digits

{0, 1, 2, ..., 9} ,

the natural numbers

N = {1, 2, 3, ...}
N0 = {0, 1, 2, 3, ...} ,

the integers
Z = {0, 1, −1, 2, −2, ...} ,
the rational numbers
a
Q = { ∣ a, b ∈ Z, b ≠ 0} .
b
The Symbol ∣ is written in place of with.

Definition 1.1.4 If every element of the set N is also an ele-

ment of the set M (that is, m ∈ N ⇒ m ∈ M ), then N is called
a subset of M (we write N ⊂ M or N ⊆ M ). The symbol ⇒ is
used in place of from which it follows that.
Two sets M1 and M2 are equal (we write M1 = M2 ), when
M1 ⊂ M2 and M2 ⊂ M1 . That means m ∈ M1 ⇔ m ∈ M2 . Here
the symbol ⇔ is written instead of if and only if, that is both
⇒ as well as ⇐ holds.

Example 1.1.5 {0, ..., 9} ⊂ N0 .

Definition 1.1.6 Let M, N be sets. Then the complement of

N in M is,
M /N = {m ∈ M ∣ m ∉ N } .
The complement M /N can be seen in terms of a so-called Venn-
Diagram in Figure 1.1. Further we call

M ∪ N = {m ∣ m ∈ M or m ∈ N }

the union of M and N , see Figure 1.2, and

M ∩ N = {m ∣ m ∈ M und m ∈ N }

intersection of M and N , see Figure 1.3.

1. FUNDAMENTAL CONSTRUCTIONS 6

Figure 1.1: Complement

Figure 1.2: Union

Example 1.1.7 N0 = N ∪ {0}.

Notation 1.1.8 For an index set I ≠ ∅ and sets Mi , i ∈ I, we

write
⋂ Mi = {m ∣ m ∈ Mi for all i ∈ I}
i∈I

for the intersection of the Mi , i ∈ I, and

⋃ Mi = {m ∣ there exists i ∈ I with m ∈ Mi }

i∈I

Figure 1.3: Intersection

1. FUNDAMENTAL CONSTRUCTIONS 7

for the union of the Mi , i ∈ I.

We abbreviate for all by ∀, and there exists by ∃.

Example 1.1.9 For I = {1, 2} and given sets M1 and M2

⋂ Mi = M1 ∩ M2 .
i∈I

Definition 1.1.10 We write ∣M ∣ or #M for the number of

elements of a finite set M and, ∣M ∣ = ∞, if M has infinite
many elements.

Example 1.1.11 That is ∣∅∣ = 0, ∣{0, ..., 9}∣ = 10 and ∣{0}∣ = 1.

Definition 1.1.12 Let M1 , ..., Mn be sets. Then the set

M1 × ... × Mn = {(m1 , ... , mn ) ∣ mi ∈ Mi ∀i = 1, ..., n}

of ordered tuples constructed from elements of M1 , ..., Mn , is

called the cartesian product of M1 , ..., Mn . For n ∈ N we write

M n = M × ... × M .
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
n-times

The elements of M n are lists (m1 , ..., mn ) of length n with entries

in M .

Example 1.1.13 We have

{1, 2, 3} × {3, 4} = {(1, 3), (1, 4), (2, 3), (2, 4), (3, 3), (3, 4)} .

The chess board is the cartesian product

{a, ... , h} × {1, ... , 8} = {(a, 1), ...} ,

the 3-dimensional space is

R3 = R × R × R,

and the set of 8-bit numbers is

8
{0, 1} = {(0, ... , 0, 0), (0, ... , 0, 1), ... , (1, ... , 1, 1)} .
1. FUNDAMENTAL CONSTRUCTIONS 8

Definition 1.1.14 Let M be a set. The power set of M is

2M = P(M ) = {A ∣ A ⊂ M } .

Theorem 1.1.15 Let M be a finite set. Then

∣2M ∣ = 2∣M ∣ .

Example 1.1.16 Power sets:

2∅ = {∅}
2{1} = {∅, {1}}
2{1,2} = {∅, {1}, {2}, {1, 2}} .

We use the following general principle for proofs to prove, for

example, Theorem 1.1.15.

1.2 Mathematical induction

Suppose we have for every n ∈ N0 a given claim A(n), and fur-
thermore it is given that:

1) Base case: A(0) is true.

2) Induction step: it follows for every n > 0 that

A(n − 1) is true ⇒ A(n) is true.

Then A(n) is true for all n ∈ N0 . In fact we have the following

chain of conclusions:

A(0) true ⇒ A(1) true ⇒ A(2) true ⇒ ...

Remark 1.2.1 Analogously, one can of course proceed to prove

statements A(n) for n ≥ n0 with n0 ∈ Z. One only has to make
sure that the initial step A(n0 ) and all subsequent arrows used
in the chain of conclusions

A(n0 ) true ⇒ A(n0 + 1) true ⇒ A(n0 + 2) true ⇒ ...

are proven.
1. FUNDAMENTAL CONSTRUCTIONS 9

Using induction, we now prove Theorem 1.1.15:

Proof. By numbering the elements of M we can assume without
loss of generality (written in short WLOG) that M = {1, ... , n},
using the convention that {1, ... , 0} = ∅. So we have to show that
the statement
∣2{1,...,n} ∣ = 2n
hold true for all n ∈ N0 .
Initial step n = 0: It is 2∅ = {∅}, and hence ∣2∅ ∣ = 1 = 20 .
Inductive step n − 1 to n: The union
⋅
2{1,...,n} = {A ⊂ {1, ..., n} ∣ n ∉ A} ∪
{A ⊂ {1, ..., n} ∣ n ∈ A}
⋅
= {A ∣ A ⊂ {1, ..., n − 1}} ∪ {A′ ∪ {n} ∣ A′ ⊂ {1, ..., n − 1}}

is disjoint, and therefore it follows from the induction hypoth-

esis
∣2{1,...,n−1} ∣ = 2n−1 ,
that
∣2{1,...,n} ∣ = 2n−1 + 2n−1 = 2n .

Next we discuss another typical example of a proof using

induction.

Notation 1.2.2 We write

n
∑ ak = a1 + ... + an
k=1

for the sum of the numbers a1 , ..., an .

Similar we use n
∏ ak = a1 ⋅ ... ⋅ an
k=1

for their product.

When the numbers ak are represented by the elements k of
the set I, we write
∑ ak
k∈I

for their sum and, analoguesly ∏k∈I ak for their product.

1. FUNDAMENTAL CONSTRUCTIONS 10

Remark 1.2.3 Given the list a = (a1 , ..., an ) the following com-
puter program computes the sum s = ∑nk=1 ak :

s:=0;
for k from 1 to n do
s:=s+a[k];
od;

We use the syntax of Maple, see [21], but the code will be sim-
ilar in most programming language. See also Exercise 1.3.

Using induction, we can prove the following general formula

for the sum ∑nk=1 k, which allows a much more efficient calculation
for this specific sum:

Theorem 1.2.4 For all n ∈ N0 ,

n
n(n + 1)
∑k = .
k=0 2

Proof. Initial step n = 0: We have

0
0 ⋅ (0 + 1)
∑k =0= .
k=0 2

Inductive step n to n + 1: We have

n+1 n
∑ k = ∑ k + (n + 1),
k=1 k=1

and hence it follows from the induction hypothesis that

n+1
n(n + 1)
∑k= + (n + 1)
k=0 2
n(n + 1) + 2(n + 1)
=
2
(n + 1)(n + 2)
= .
2

For a further example, see Exercise 1.4.

1. FUNDAMENTAL CONSTRUCTIONS 11

Remark 1.2.5 The analogue to a proof by induction is in com-

puter science the concept of a recursive algorithm. For example,
the following recursive function calculates the sum ∑nk=0 k:
sumints:=proc(n)
if n=0 then return(0);fi;
return(sumints(n-1)+n);
end proc;
We can also write a recursive function that determines all
subsets of {1, ..., n} from the proof of Theorem 1.1.15. For the
implementation thereof, see Exercise 1.8. Another proof by in-
duction, which provides a recursive algorithm, is discussed in
Exercises 1.10 and 1.11.
For further examples of induction, see the Exercises 1.5, 1.6,
1.7 and 1.12.

1.3 Relations
In the following way we can describe relations between two sets:
Definition 1.3.1 A relation relation between sets M and N is
given by the subset R ⊂ M × N .
Example 1.3.2 For M = {2, 3, 7}, N = {4, 5, 6} and
R = {(m, n) ∈ M × N ∣ m divides n}
we have
R = {(2, 4), (2, 6), (3, 6)} .
The most important role is played by relations in which each
element of M gets assigned exactly one element of N :

1.4 Maps
Definition 1.4.1 A map f ∶ M → N is a relation R ⊂ M × N ,
such that for every m ∈ M there is a unique element f (m) ∈ N
with (m, f (m)) ∈ R. We write
f∶ M → N
m ↦ f (m).
1. FUNDAMENTAL CONSTRUCTIONS 12

We call M the source and N the target of f .

For a subset A ⊂ M
f (A) = {f (m) ∣ m ∈ A} ⊂ N
is called the image of A under f , and
Image(f ) ∶= f (M )
is called the image of f .
For B ⊂ N
f −1 (B) = {m ∈ M ∣ f (m) ∈ B} ⊂ M
is called the preimage of B under f .
Remark 1.4.2 If a map is given by a mapping rule f ∶ M → N ,
m ↦ f (m), the representation of f as a relation is nothing else
than the graph
R = Graph(f ) = {(m, f (m)) ∣ m ∈ M } ⊂ M × N
of f .
Example 1.4.3 For
f∶ R → R
x ↦ f (x) = x2
we have
R = Graph(f ) = {(x, x2 ) ∣ x ∈ R} ,
see Figure 1.4. The image of f is
f (R) = R≥0
and we have
√ √
f −1 ({1, 2}) = {−1, 1, − 2, 2}.
Definition 1.4.4 A map f ∶ M → N is surjective, if for the
image of f we have
f (M ) = N .
If for all m1 , m2 ∈ M we have, that
f (m1 ) = f (m2 ) Ô⇒ m1 = m2 ,
then f is injective.
A map that is both injective and surjective is bijective.
1. FUNDAMENTAL CONSTRUCTIONS 13

–1

–2 –1 0 1 2

Figure 1.4: Graph of parabola.

Example 1.4.5 The parabola function

R → R, x ↦ x2

in Example 1.4.3 is neither injective or surjective. As a map

onto its image
R → R≥0 , x ↦ x2
it becomes surjective. The map

R≥0 → R≥0 , x ↦ x2

in which we also restrict the source is bijective. The hyperbola

1
R/{0} → R, x ↦
x
is injective, but not surjective (see Figure 1.5).

Theorem 1.4.6 (Pigeonhole principle) If M, N are finite sets

and if f ∶ M → N is an injective mapping, then ∣M ∣ ≤ ∣N ∣.

Example 1.4.7 Let M = {1, 2, 3} and N = {1, 2}. Since ∣M ∣ >

∣N ∣ there is no injective mapping f ∶ M → N . Figure 1.6 shows
an example of a map f ∶ M → N . Since f (1) = f (3), it follows
that f is not injective.
1. FUNDAMENTAL CONSTRUCTIONS 14

Figure 1.5: Hyperbola

1 a
3 b
2

Figure 1.6: A non-injective map

We now prove Theorem 1.4.6:

Proof. We have
∣N ∣ = ∑ 1 ≥ ∑ ∣f −1 ({n})∣ = ∣M ∣ ,
n∈N n∈N

since f −1 ({n}) has exactly 1 element, if n lies in the image of

f (since f is injektive), and is empty otherwise. The second
equality holds true, since M is the disjoint union
⋅
M = ⋃ f −1 ({n})
n∈N

of the level sets f −1 ({n}) of the map f (which is analogous to the

contour lines on a map giving the height of the respective point):
A map assigns to each element m exactly one value n. We will
come back to this idea in the context of equivalence relations.
Since a map assigns exactly one image element to each ele-
ment of the source, the corresponding statement for surjective
maps follows directly:
1. FUNDAMENTAL CONSTRUCTIONS 15

Theorem 1.4.8 If M, N are finite sets and f ∶ M → N a sur-

jective map, then ∣M ∣ ≥ ∣N ∣.

Combining Theorem 1.4.6 and Theorem 1.4.8, we have:

Corollary 1.4.9 If M, N are finite sets and f ∶ M → N a bijec-

tive map, then ∣M ∣ = ∣N ∣.

Definition and Theorem 1.4.10 If f ∶ M → N is bijective,

then there is a unique inverse map

f −1 ∶ N → M , y ↦ x if f (x) = y.

We have

f −1 (f (x)) = x and f (f −1 (y)) = y

for all x ∈ M and y ∈ N , respectively. Moreover, f −1 is bijective.

Proof. The inverse map is well-defined (that is, its definition

assignes to every element of the source a unique element of the
target): For every y ∈ N there is exactly one x ∈ M with f (x) =
y. Since f is surjective there exists such an x, and since f is
injective, this x is unique.
The two equalities are clear by definition of f −1 . On the
bijectivity of f −1 : Since for every x ∈ M we have f −1 (f (x)) = x,
the map f −1 is surjective. For the injectivity, we use that f is a
map: If y1 , y2 ∈ N and xi ∈ M with yi = f (xi ), then from

x1 = f −1 (y1 ) = f −1 (y2 ) = x2 ,

it follows, that y1 = f (x1 ) = f (x2 ) = y2 .

For the term “there is a unique” used above, we also write
the symbol ∃1 .

Remark 1.4.11 The inverse map f −1 is the relation

{(f (x), x) ∣ x ∈ M } ⊂ N × M .
1. FUNDAMENTAL CONSTRUCTIONS 16

1 a
3 b
2 c

Figure 1.7: A bijektive map and its inverse map.

2

0 1 2

Figure 1.8: Square root

Example 1.4.12 Figure 1.7 shows a bijective map f ∶ {1, 2, 3} →

{a, b, c} and its inverse map f −1 .The inverse of the bijective map

R≥0 → R≥0 , x ↦ x2

is
√
R≥0 → R≥0 , y ↦ y
as shown in Figure 1.8.

See also the Exercises 1.14, 1.17, 1.18 and 1.19.

Definition 1.4.13 Let f ∶ M → N and g ∶ N → L be maps.

Then the composition of f and g is defined as

g○f ∶ M → L
m ↦ g(f (m))
1. FUNDAMENTAL CONSTRUCTIONS 17

Lemma 1.4.14 The composition of maps is associative, that is,

for maps
f g h
M →N →L→K
we have
h ○ (g ○ f ) = (h ○ g) ○ f .

For the proof see Exercise 1.15.

Example 1.4.15 Even if f ∶ M → M and g ∶ M → M , in general

f ○ g ≠ g ○ f . For example for

f ∶ R2 → R2 , (x, y) ↦ (x + y, y)
g ∶ R2 → R2 , (x, y) ↦ (x, x + y)

we get

f ○ g ∶ R2 → R2 , (x, y) ↦ (2x + y, x + y)
g ○ f ∶ R2 → R2 , (x, y) ↦ (x + y, x + 2y).

Definition 1.4.16 Let M be a set. The identity map on M

is
idM ∶ M → M
m ↦ m

Example 1.4.17 Figure 1.9 shows the graph of idR .

The equations in Definition and Theorem 1.4.10 can then be

wirtten as follows:

Theorem 1.4.18 If f ∶ M → N is a bijective map, then

f −1 ○ f = idM f ○ f −1 = idN

Indeed, these equalities characterize the property bijective

and uniquely define the inverse map. For this see Exercise 1.16.
1. FUNDAMENTAL CONSTRUCTIONS 18

–2 0 2

Figure 1.9: Identity map R → R

1.5 Partial orderings and equivalence re-

lations
Definition 1.5.1 A relation R ⊂ M × M on a set M is called

• reflexive, if (m, m) ∈ R for all m ∈ M ,

• transitive, if

(l, m) ∈ R and (m, n) ∈ R Ô⇒ (l, n) ∈ R,

• anti-symmetric, if

(n, m) ∈ R and (m, n) ∈ R Ô⇒ m = n.

If R is reflexive, transitive, and anti-symmetric, then R is

called a partial ordering. If, in addition, for all m, n ∈ M we
have (m, n) ∈ R or (n, m) ∈ R, then R is called a total ordering.

Example 1.5.2 1) The inclusion ⊂ of subsets of a set M is a

partial ordering on the power set 2M : For all A, B, C ⊂ M
we have

• A ⊂ A (reflexive)
1. FUNDAMENTAL CONSTRUCTIONS 19

• A ⊂ B and B ⊂ C Ô⇒ A ⊂ C (transitive)
• A ⊂ B and B ⊂ A Ô⇒ A = B (anti-symmetric).

In general ⊂ is not a total ordering, for example, for M =

{1, 2} neither {1} ⊂ {2} nor {2} ⊂ {1}.

2) In contrast, ≤ on R is a total ordering.

The concept of an equivalence relation relaxes the concept of

equality.

Definition 1.5.3 Let M be a set and R ⊂ M × M a reflexive

and transitive relation. If R is in addition symmetric, that is,

(m, n) ∈ R ⇒ (n, m) ∈ R,

then R is called an equivalence relation.

If we write m ∼ n for (m, n) ∈ R, then

• reflexive means, that m ∼ m for all m ∈ M ,

• transitive means, that m ∼ l and l ∼ n ⇒ m ∼ n for all

m, l, n ∈ M and

• symmetric means, that m ∼ n ⇒ n ∼ m for all m, n ∈ M .

Example 1.5.4 Equality is an equivalence relation.

The property of two persons to be the same height, is an equiv-
alence relation (in contrast, the property of being the same hight
up to 1cm difference is not an equivalence relation, since it is
not transitive).
More generally: Let f ∶ M → N be a map. Then

m1 ∼ m2 ⇐⇒ f (m1 ) = f (m2 )

defines an equivalence relation on M .

Definition 1.5.5 If M is a set and ∼ is an equivalence relation

and m ∈ M , then

[m] = {n ∈ M ∣ m ∼ n} ⊂ M
1. FUNDAMENTAL CONSTRUCTIONS 20

is called the equivalence class of m. Every n ∈ [m] is called a

representative of [m].
Moreover, we write

M / ∼ = {[m] ∣ m ∈ M } ⊂ 2M

for the set of equivalence classes of ∼ and

π∶ M → M/ ∼
m ↦ [m]

for the canonical map.

Theorem 1.5.6 Given an equivalence relation, any two equiva-

lence classes are equal or disjoint.

Proof. Let [m] ∩ [n] ≠ ∅. We have to show that [m] = [n]. If

a ∈ [m] ∩ [n], that is, a ∼ m and a ∼ n, then, using symmetry and
transitivity, it follows that m ∼ n, that is, m ∈ [n]. Let a ∈ [m]
be any element. Then a ∼ m and m ∼ n, hence a ∼ n, that is,
a ∈ [n]. So we have seen, that [m] ⊂ [n]. The other inclusion
follows in the same way.
An equivalence relation paritions (subdivides) M in the equiv-
alence classes.

Remark 1.5.7 We have

m1 ∼ m2 ⇐⇒ [m1 ] = [m2 ],

that is, equivalence translates into equality of equivalence classes.

Example 1.5.8 The equivalence classes under the equivalence

relation of being the same hight on a set M of persons (see Ex-
ample 1.5.4) are the subsets of all persons with the same hight.
So the set of equivalence classes M / ∼ is in bijection to the set of
all occuring hights of persons. A cloths sales person is interested
mainly in [m] not in m.

Example 1.5.9 Consider the equivalence relation ∼ on R2 given

by
(x1 , y1 ) ∼ (x2 , y2 ) ⇐⇒ f (x1 , y1 ) = f (x2 , y2 )
1. FUNDAMENTAL CONSTRUCTIONS 21

with
f (x, y) = x2 + y 2 .
The equivalence classes are the concentric circles

Ks = {(x, y) ∈ R2 ∣ x2 + y 2 = s}

for s ∈ R≥0 , and the point (0, 0), which is a degenerate form of a
circle, the circle with radius 0. For example,

[(1, 2)] = {(x, y) ∈ R2 ∣ x2 + y 2 = 5} .

Hence
M / ∼ = {Ks ∣ s ∈ R≥0 } ,
and the map R≥0 → M / ∼, s ↦ Ks is bijective. See Figure 1.10.
3

–1

–2

–3 –2 –1 0 1 2 3

Figure 1.10: Equivalence classes

1.6 Exercises
Exercise 1.1 Let M be a set. Show that for subsets A, B, C ⊂ M
(using, for example, Venn diagrams):

1) For ∩ we have:

(a) Commutativity A ∩ B = B ∩ A,
(b) Identity A ∩ M = A,
(c) Assoziativity A ∩ (B ∩ C) = (A ∩ B) ∩ C.
1. FUNDAMENTAL CONSTRUCTIONS 22

2) For ∪ we have:

(a) Commutativity A ∪ B = B ∪ A,
(b) Identity A ∪ ∅ = A,
(c) Assoziativity A ∪ (B ∪ C) = (A ∪ B) ∪ C.

3) For ∩ and ∪ the distributive laws hold:

A ∩ (B ∪ C) = (A ∩ B) ∪ (A ∩ C)
A ∪ (B ∩ C) = (A ∪ B) ∩ (A ∪ C)

4) Compare with the arithmetic rules for integers.

Exercise 1.2 Show for finite sets M and N that

∣M ∪ N ∣ = ∣M ∣ + ∣N ∣ − ∣M ∩ N ∣

and
∣M × N ∣ = ∣M ∣ ⋅ ∣N ∣ .

Exercise 1.3 Write a program, that computes for a list a =

(a1 , ..., an ) ∈ Zn the sum
n
∑ ak .
k=1

Exercise 1.4 Prove, using induction, that

n
n (n + 1) (2n + 1)
∑ k2 =
k=1 6

for alle n ∈ N.

Exercise 1.5 Find a formula for

n
∑ (2k − 1)
k=1

and prove your claim using induction.

1. FUNDAMENTAL CONSTRUCTIONS 23

Exercise 1.6 Find a formula for

n
∑ k3
k=1

and prove your claim using induction.

Exercise 1.7 Show, using induction, that for q ∈ R, q ≠ 1 we

have n
1 − q n+1
∑ qk =
k=0 1−q

Exercise 1.8 Implement a function, which enumerates recur-

sively all subsets of {1, ..., n}.

Exercise 1.9 Let 0 ≤ k ≤ n. Prove that for the number (nk) of

k-element subsets of an n-element set we have
n n!
( )=
k k!(n − k)!
with n! = 1 ⋅ 2 ⋅ ... ⋅ n.

Exercise 1.10 The game ”Towers of Hanoi” consists out of 3

fields, on which n discs with pairwise different radius can be
stacked (see Figure 1.11). At the start of the game all discs
are stacked on one field, sorted by increasing size, thus forming
a tower. The goal of the game is to move the original stack to a
different field. To do so, in every move of the game, one can shift
the top disc on an arbitrary tower to any other tower, provided
this tower does not contain a smaller disc.
Describe an algorithm, which solves the game, find a formula
for the required number of moves, and prove this formula using
induction.

Exercise 1.11 Write a recursive program, which solve the game

”Towers of Hanoi”.

Exercise 1.12 In an american city map with n avenues and m

streets (see Figure 1.12) we want to move from point A to point
B. How many shortest paths are there?
Prove your formula using induction on n + m.
1. FUNDAMENTAL CONSTRUCTIONS 24

Figure 1.11: Towers of Hanoi

Figure 1.12: How many shortest paths are there from A to B.

Exercise 1.13 Give an example for a map N → N, which is

1) injective but not surjective,

2) surjective but not injective.

Exercise 1.14 At a party n persons meet. Prove that two of

them know the same number of persons at the party.

Exercise 1.15 The composition of maps is associative, that is,

for maps
f g h
M →N →L→K
we have
h ○ (g ○ f ) = (h ○ g) ○ f .

Exercise 1.16 Let f ∶ M → N be a map. Prove:

1) f is injective if and only if there is a map g ∶ f (M ) → M

with g ○ f = idM .

2) f is surjective if and only if there is a map g ∶ N → M with

f ○ g = idN .
1. FUNDAMENTAL CONSTRUCTIONS 25

3) f ist bijective if and only if there is a map g ∶ N → M with

g ○ f = idM and f ○ g = idN .
Moreover, g = f −1 is the inverse map.

Exercise 1.17 Let M, N be finite sets with ∣M ∣ = ∣N ∣ and f ∶

M → N a map. Prove that the following are equivalent:

1) f is bijective,

2) f is injective,

3) f is surjective.

Exercise 1.18 Suppose the numbers 1, ..., 101 are given in any
order. Prove that 11 of them (not necessarily consecutive) are in
an increasing or decreasing order.
Hint: Consider a suitable set of paris and use the pigeon hole
principle.

Exercise 1.19 Let n ∈ N and suppose there are n2 + 1 points

given in the square

{(x, y) ∣ 0 ≤ x < n, 0 ≤ y < n} ⊂ R2 .

Show
√ that among these points there are two which have distance
≤ 2.

Exercise 1.20 Let M be an infinite set. Prove:

1) There is no surjective map ϕ ∶ M → 2M .

2) There is no injective map ψ ∶ 2M → M .

Exercise 1.21 Let M ∶= R2 / {(0, 0)} be the set of points in the

real plane without the 0-point. On M define (x, y) ∼ (x′ , y ′ ) if
1. FUNDAMENTAL CONSTRUCTIONS 26

and only if there is a line through (0, 0) ∈ R2 which contains both

the point (x, y) and the point (x′ , y ′ ).

1) Prove that ∼ is an equivalence relation.

2) Find a geometric representation of M / ∼ by assigning to

any equivalence class a suitable representative.
2

Numbers

In this section, as a start into algebra, we discuss the key prop-

erties of numbers. All these properties serve as role models for
more general classes of rings.

2.1 The integers and division with re-

mainder
On the natural numbers N0 = {0, 1, 2, 3, ...} there are operations
+ and ⋅, which obey the associative laws

a + (b + c) = (a + b) + c
a ⋅ (b ⋅ c) = (a ⋅ b) ⋅ c

the commutative laws

a+b=b+a
a⋅b=b⋅a

and the distributive law

a ⋅ (b + c) = a ⋅ b + a ⋅ c

for alle a, b, c ∈ N0 . We will not discuss the axiomatic definition

of the natural numbers, but we remark that their key property
is that for any number there is a number which is larger by one.
As an exercise read up in a book or seach engine of your choice
about the Peano axioms.

27
2. NUMBERS 28

In N0 there is no number a with

1 + a = 0.

In practice this means: We can describe assets on an account,

but no debt.
From the natural numbers, one constructs the integers Z =
{0, 1, −1, 2, −2, ...} as follows:

Remark 2.1.1 The basic idea of the construction is: The net
worth of a bank account one can write as the difference of assets
and depts. Different tuples of (assets, debt) lead to the same net
worth of the account, for example,

5 − 1 = 1000006 − 1000002

that is, the net worth of an account with 5 e assets and 1 e debt
is the same as that of one with 1000006 e assets and 1000002 e
debt. The represent the net worth, we have to consider equiva-
lence classes with respect to an appropriate equivalence relation.
The account in the example have the same net worth, since

5 + 1000002 = 1000006 + 1.

One hence defines

Z ∶= (N0 × N0 ) / ∼

with the equivalence relation

(a, b) ∼ (c, d) ⇔ a + d = b + c,

and considers the equivalence classes

[(a, b)] = {(c, d) ∣ (c, d) ∼ (a, b)} .

We think of [(a, b)] as the integer a − b. This motivates the

following (well defined) operations + and ⋅ on Z

[(a, b)] + [(c, d)] ∶= [(a + c, b + d)]

[(a, b)] ⋅ [(c, d)] ∶= [(a ⋅ c + b ⋅ d, a ⋅ d + b ⋅ c)] ,
2. NUMBERS 29

which obey the associative, commutative and distributive laws

(see also Exercise 2.2). We then have

[(a, b)] + [(b, a)] = [(0, 0)]

for all [(a, b)] ∈ Z, in particular,

[(1, 0)] + [(0, 1)] = [(0, 0)] .

Moreover

[(0, 0)] + [(a, b)] = [(a, b)]

[(1, 0)] ⋅ [(a, b)] = [(a, b)] .

A set with such operations is called a commutative ring with 1.

The integers are sorted by the total ordering ≤.
Every account [(a, b)] is equivalent to a unique account with
either no assets or no debt: For a ≥ b let c ∈ N0 with a = b + c.
Then (a, b) ∼ (c, 0). For a < b let c ∈ N with b = a + c. Then
(a, b) ∼ (0, c). We write short

c ∶= [(c, 0)]

and
−c ∶= [(0, c)] .
We then have
c + (−c) = 0
for all c ∈ Z/{0}, since c + (−c) = [(c, c)] = [(0, 0)] = 0.
In a similar way, one can construct Q from Z as

Q = (Z × Z/{0}) / ∼

with the equivalence relation

(a, b) ∼ (c, d) ⇔ ad = bc,

where we write the equivalence classes as

a
∶= [(a, b)].
b
The real numbers R one can again construct from Q using a
suitable equivalence relation.
2. NUMBERS 30

In Q every number a can be divided by any number b ≠ 0.

In many problems of daily life and mathematics this does not
make sense, since the smallest useable unit is 1. If we want to
distribute 1000 passengers uniformly on 3 airplanes, then 1000
3 is
not a resonable solution, we rather want

1000 = 3 ⋅ 333 + 1,

so one person has to stay behind (or one of the airplanes will
have to take the extra passenger). This process is called division
with remainder (then 1 left-over passenger is the remainder):

Lemma 2.1.2 (Division with remainder) Given a, b ∈ Z, b ≠

0, there are uniquely defined q, r ∈ Z with

a=b⋅q+r

and 0 ≤ r < ∣b∣.

Example 2.1.3 In the above example, a = 1000 and b = 3, and

we have
1000 = 3 ⋅ 333 + 1,
that is, q = 333 and r = 1.

We prove Lemma 2.1.2:

Proof. Existence: Without loss of generality b > 0. The set

{w ∈ Z ∣ b ⋅ w > a} ≠ ∅

has a smallest element w. We set

q ∶= w − 1 r ∶= a − qb.

Obviously a = qb + r, moreover qb + b > a, hence,

r<b

and, since w was chosen minimal, also bq ≤ a, hence,

r ≥ 0.
2. NUMBERS 31

Uniqueness: If we have two representations

b ⋅ q1 + r1 = a = b ⋅ q2 + r2

and without loss of generality r2 ≤ r1 , then

0 ≤ r1 − r2 < ∣b∣ ,
´¹¹ ¹ ¸¹ ¹ ¹¶
b⋅(q2 −q1 )

hence, q1 = q2 and r1 = r2 .
The proof gives an explicit (but very inefficient) algorithm
for division with remainder (scan through the w, starting with
a random number, decreasing w iteratively by 1 if b ⋅ w > a,
and increasing w iteratively by 1 if b ⋅ w < a until the respective
condition is not satisfied any more). In practice, one rather
proceeds as follows:

Remark 2.1.4 School book division without digits after the dec-
imal point iteratively determines q (starting with the largest digit),
thus yielding an algorithm for division with remainder.

Example 2.1.5 For a = 2225 and b = 7 write

2225 = 7 ⋅ 317 + 6
−21
12
−7
55
−49
6

hence q = 317 and r = 6.

Using division with remainder, we can algorithmically decide

divisibility.

Definition 2.1.6 Let a, b ∈ Z. We say that b divides a

b∣a

if there is a q ∈ Z with a = b ⋅ q. This means the division of a by

b yields remainder r = 0.
2. NUMBERS 32

Two numbers a, b ∈ Z are called coprime, if for any t ∈ N

with t ∣ a and t ∣ b it follows that t = 1.
Let m ∈ N and a, b ∈ Z. Then a is congruent to b modulo m

a ≡ b mod m

if m ∣ (a − b).

Example 2.1.7 1 ≡ 7 mod 3.

Being congruent modulo m is an equivalence relation, see

Exercise 2.3. There we will also implement a function, which
decides congruence modulo m using division with remainder.
For fixed m we write the equivalence class (called residue
class) of a as

a = {b ∈ Z ∣ a ≡ b mod m}
= {a + k ⋅ m ∣ k ∈ Z} .

Hence, a ≡ b mod m if and only if a = b.

Example 2.1.8 Congruence modulo 3 partitions Z in 3 residue

classes

0 = {..., −3, 0, 3, 6, ...}

1 = {..., −2, 1, 4, 7, ...}
2 = {..., −1, 2, 5, 8, ...},

since division with remainder by 3 can give the remainders 0, 1

and 2.

Residue classes play an important role in many public-key

crypto systems.

2.2 Fundamental theorem of arithmetic

Definition 2.2.1 An element p ∈ N, p ≥ 2 is called a prime
number, if p = a ⋅ b, a, b ∈ N implies a = 1 or b = 1.

Example 2.2.2 2, 3, 5, 7, 11, 13, 17, 19, 23...

2. NUMBERS 33

How to find all prime numbers up to a given bound will be

discussed in the next section.
Theorem 2.2.3 (Fundamental theorem of arithmetic) Every
number n ∈ Z/{0, −1, 1} has a unique representation
n = ± pr11 ⋅ ... ⋅ prss
with prime numbers p1 < ... < ps and ri ∈ N. The pi are called the
prime factors of n.
Proof. Existence of prime factorization using induction on n:
n = 2 is a prime number. If n > 2 and not prime, then
n = a ⋅ b with a, b ≠ 1. Since a, b < n, both a and b by the
induction hypothesis have factorizations, and by sorting by the
prime factors we obtain a prime factorization of n = a ⋅ b.
Proof of uniqueness, using induction on n:
For n = 2, then n is prime, and the claim is clear. Suppose
n > 2 and
n = p1 ⋅ ... ⋅ ps = q1 ⋅ ... ⋅ qt
with p1 ≤ ... ≤ ps and q1 ≤ ... ≤ qt . If s = 1 or t = 1, then again n
is prime, and the claim is clear. So suppose now that s, t ≥ 2.
If p1 = q1 then
p2 ⋅ ... ⋅ ps = q2 ⋅ ... ⋅ qt < n
has, by the induction hypothesis, a unique prime factorization
and die claim follows.
Assume now that p1 < q1 . Then
n > p1 ⋅ (p2 ⋅ ... ⋅ ps − q2 ⋅ ... ⋅ qt ) = (q1 − p1 ) ⋅ q2 ⋅ ... ⋅ qt ≥ 2,
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
=∶N1 =∶N2

so N1 = N2 , by the induction hypothesis, has a unique prime

factorization. Since p1 < q1 ≤ ... ≤ qt we have p1 ≠ qi for i ≥ 1.
Moreover, p1 is not a divisor of q1 − p1 , since otherwiese p1 would
divide the prime q1 . Hence p1 is a prime factor of N1 , but not
one of N2 , a contradiction.
Example 2.2.4 24 = 23 ⋅ 3.
In Maple we can compute a prime factorization by:
ifactor(24);
(2)3 (3)
2. NUMBERS 34

The proof of the fundamental theorem only shows the exis-

tence of a (unique) prime factorization. We will come back to
the question, how to compute such a factorization.
From the fundamental theorem we conclude:
Corollary 2.2.5 (Euklid’s first theorem) If p ∈ Z is prime
and a, b ∈ Z with p ∣ ab, then p ∣ a or p ∣ b.
Proof. The prime factorization of ab is obtained by combining
the factorizations of a and b.
Corollary 2.2.6 (Euklid’s second theorem) There are infinitely
many prime numbers.
Proof. Let M = {p1 , ..., pr } be a finite set of prime numbers.
We show, that there is a prime number, which is not in M . The
number N = p1 ⋅ ... ⋅ pr + 1 is not divisible by any of the primes pi ,
since otherwise also 1 would be divisible by pi . Hence, there is a
prime factor p of N , which is not in M .
Without proof we mention the following theorem on the den-
sity of the prime numbers:
Theorem 2.2.7 (Prime number theorem) Setting
π(x) = ∣{p ≤ x ∣ p ∈ N prime}∣ .
for x ∈ R>0 , we have
π (x)
lim x = 1.
x→∞
ln(x)

Example 2.2.8 The following program (in the syntax of Maple)

computes π(x):
pi:=proc(x)
local p,N;
p:=2;
N:=0:
while p<=x do
p:=nextprime(p);
N:=N+1;
od;
return(N);
end proc:
2. NUMBERS 35

We obtain, for example,

pi(100000);
9592
So about 10% of the numbers ≤ 100000 are prime numbers.
See also Exercise 2.5.

2.3 Greatest common divisor and Eu-

clidean algorithm
Definition 2.3.1 If a1 , ..., at ∈ Z, then d ∈ N is called greatest
common divisor of a1 , ..., at , written d = gcd (a1 , ..., at ), if we
have
1) d ∣ aj for all j = 1, ..., t, that is, d is a divisor of all aj , and
2) if d˜ ∈ Z is a divisor of all aj , that is d˜ ∣ aj for all j = 1, ..., t,
then d˜ ∣ d.
Furthermore the number m ∈ N is called least common
multiple of a1 , ..., at , written m = lcm (a1 , ..., at ), if we have
1) aj ∣ m for all j = 1, ..., t, that is, m is a multiple of all aj ,
and
2) if m̃ ∈ Z is a multiple of all aj , that is, aj ∣ m̃ for all
j = 1, ..., t, then m ∣ m̃.
Example 2.3.2 The common divisors of 18 = 2 ⋅ 32 and 66 =
2 ⋅ 3 ⋅ 11 are 1, 2, 3 and 6, hence
gcd(18, 66) = 6.
Remark 2.3.3 If we write
r
aj = ±1 ⋅ ∏si=1 pi ji
with pi prime and rji ≥ 0, then
min{rji ∣j}
gcd (a1 , ..., at ) = ∏si=1 pi (2.1)
(and for lcm the corresponding formula replacing the minimum
with the maximum holds true). Using these formulas, we get:
2. NUMBERS 36

1) Two numbers a, b ∈ Z are coprime if and only if

gcd (a, b) = 1.

2) For a, b ∈ N,

gcd (a, b) ⋅ lcm(a, b) = a ⋅ b.

Example 2.3.4 For 18 = 2 ⋅ 32 and 66 = 2 ⋅ 3 ⋅ 11 we have

gcd(18, 66) = 6.

A much more efficient way for determining the greatest com-

mon divisor (and thus also the least common multiple) is pro-
vided by the Euclidean algorithm:

Theorem 2.3.5 (Euclidean algorithm) Let a1 , a2 ∈ Z/{0}. Then

the successive division with remainder

a1 = q 1 a2 + a3
⋮
aj = qj aj+1 + aj+2
⋮
an−2 = qn−2 an−1 + an
an−1 = qn−1 an + 0

terminates with remainder zero, and

gcd (a1 , a2 ) = an .

Backsubstiution of these equalities

an = an−2 − qn−2 an−1

⋮
a3 = a1 − q 1 a2

yields a representation of the greatest common divisor as

gcd (a1 , a2 ) = u ⋅ a1 + v ⋅ a2

with u, v ∈ Z. Computing this representation is referred to as the

extended Euclidean algorithm.
2. NUMBERS 37

Proof. We have ∣ai+1 ∣ < ∣ai ∣ for all i ≥ 2, hence, after finitely
many iterations, ai = 0. Then an is a divisor of an−1 , hence also
of an−2 = qn−2 an−1 + an and inductively of an−1 , ..., a1 . If t is an
arbitrary divisor of a1 and a2 , then also of a3 = a1 − q1 a2 and
inductively of a1 , ..., an .
Example 2.3.6 We compute the gcd of 66 and 18 using the Eu-
clidean algorithm, that is, by successive division with remainder:
66 = 3 ⋅ 18 + 12
18 = 1 ⋅ 12 + 6
12 = 2 ⋅ 6 + 0
Hence gcd (66, 18) = 6, since reading the equaltities backwards,
we have
6 ∣ 12 hence 6 ∣ 18 hence 6 ∣ 66
and reading them top-to-bottom, if t is a divisor of 66 and 18,
then
t ∣ 12 hence t ∣ 6.
Moreover, we obtain a representation of gcd (36, 15) as a Z-
linear combination of 66 and 18
6 = 18 − 1 ⋅ 12 = 18 − 1 ⋅ (66 − 3 ⋅ 18) = 4 ⋅ 18 + (−1) ⋅ 66.
In Maple we can execute the extended Euclidean algorithm by:
igcdex(66,18,’u’,’v’);
6
Here the command stores in x and y the coefficients in the
representation of the ggT as a linear combintation:
u;
-1
v;
4
u*66+v*18;
6
Note that u and v are not unique. We could also choose, for
example, u = −19 and v = 70.
One key applicaiton of the representation of 1 as a Z-linear
combination of two coprime numbers is solving of simultaneous
congruences. This will be addressed in the next section on the
Chinese remainder theorem.
2. NUMBERS 38

2.4 The Chinese remainder theorem

Theorem 2.4.1 (Chinese remainder theorem in Z) If n1 , ..., nr ∈
N are pairwise coprime and a1 , ..., ar ∈ Z, then the simultaneous
congruences

x ≡ a1 mod n1
⋮
x ≡ ar mod nr

have a solution. The solution is unique up to multiples of n =

n1 ⋅ ... ⋅ nr .

Proof. Set
n
n̂i =
ni
and find, using the extended Euclidean algorithm, xi , yi ∈ Z with

1 = gcd (ni , n̂i ) = xi ni + yi n̂i .

Then

yi n̂i ≡ 0 mod nj ∀j ≠ i
yi n̂i ≡ 1 mod ni .

hence,
r
z = ∑ ai yi n̂i
i=1

satisfies the congruences, similarly all z + k ⋅ n with k ∈ Z do.

If x and x′ are solutions, then ni ∣ (x − x′ ) for all i. Hence,
also lcm(n1 , ... , nr ) ∣ (x − x′ ). Since the ni are pairwise coprime,
lcm(n1 , ... , nr ) = n1 ⋅ ... ⋅ nr , that is,

n ∣ (x − x′ ) .

The Chinese remainder theorem allows us, to replace an ar-

bitrary number of congruences by a single congruence. We can
also iteratively combine pairs of two congruences into one. We
hence formulate the solution algorithm in the special case r = 2:
2. NUMBERS 39

Remark 2.4.2 Given coprime numbers n1 , n2 ∈ N and a1 , a2 ∈

Z, we find a solution of the simultaneous congruences

x ≡ a1 mod n1
x ≡ a2 mod n2

The extended Euclidean algorithm finds u, v ∈ Z with

1 = gcd(n1 , n2 ) = u ⋅ n1 + v ⋅ n2

Since

un1 ≡ 0 mod n1
un1 ≡ 1 mod n2
vn2 ≡ 1 mod n1
vn2 ≡ 0 mod n2

for
z ∶= a2 ⋅ u ⋅ n1 + a1 ⋅ v ⋅ n2
we have

z ≡ a1 mod n1
z ≡ a2 mod n2

If x another solution, then ni ∣ (x − z) for i = 1, 2, and hence

n1 n2 ∣ (x − z).
So we obtain
x ≡ a1 mod n1
} ⇐⇒ x ≡ z mod n1 n2
x ≡ a2 mod n2

If we apply this method iteratively, we get another proof of

Theorem 2.4.1.

We note that the Chinese remainder theorem can be formu-

lated in a much more general setting.

Example 2.4.3 We solve the simultaneous congruences

x ≡ −28 mod 30
x ≡ 5 mod 7.
2. NUMBERS 40

Since gcd (30, 7) = 1, it follows there is a solution. With the

extended Euclidean algorithm, we find u and v with
u ⋅ 30 + v ⋅ 7 = 1,
for example, u = −3, v = 13. Then we have
(−3) ⋅ 30 ≡ 0 mod 30
(−3) ⋅ 30 ≡ 1 mod 7
13 ⋅ 7 ≡ 1 mod 30
13 ⋅ 7 ≡ 0 mod 7
hence
z = (−28) ⋅ (13 ⋅ 7) + 5 ⋅ (−3 ⋅ 30) = −2998
is a solution (which is unique modulo 210). The Chinese re-
mainder theorem thus replaces the two congruences by a single
one:
x ≡ −28 mod 30
} ⇔ x ≡ −2998 ≡ 152 mod 210.
x ≡ 5 mod 7
For the single congruence, it is easy to write down the solution
set
152 + 210 ⋅ Z = {152 + k ⋅ 210 ∣ k ∈ Z} ,
which indeed is the residue class 152.
If the moduli ni are not coprime, we can find a similar so-
lution formula, however, simultaneous congruences may not be
solveable. The following theorem gives a criterion:
Theorem 2.4.4 Let a1 , a2 ∈ Z and n1 , n2 ∈ N. Then the simul-
taneous congruences
x ≡ a1 mod n1
x ≡ a2 mod n2
are solveable if and only if
a1 − a2 ≡ 0 mod gcd (n1 , n2 ) .
The solutions are unique up to the addition of multiples of lcm (n1 , n2 ).
The proof is Exercise 2.12, where we also describe a method
to find the solutions.
2. NUMBERS 41

2.5 Prime factorization

There is a very straight-forward method to prime factorization:
trial division Although there are much more powerful algorithms,
for rather small numbers up to 106 trial division is the best
choice.

Algorithm 2.5.1 (Trial division) Let n ∈ N be composite (that

is, not prime). The smallest prime divisor p of n satisfies
√
p ≤ m ∶= ⌊ n⌋ .

If we know all prime numbers p ≤ m, then we test p ∣ n using

division with remainder. In this way, we can factorize n.
√
Proof. Write n = p ⋅ q. Then√ p ≤ p ⋅ q = n, hence p ≤
2 n. Since
p ∈ N, it follows that p ≤ ⌊ n⌋.

Example 2.5.2 In order to factorize 234 via trial division,

√ we
first test, whether n is divisible by a prime number p ≤ ⌊ 234⌋ =
15. We find
234 = 2 ⋅ 117.
If√117 is not prime, then it must have a prime divisor p ≤
⌊ 117⌋ = 10. We find that

117 = 3 ⋅ 39.
√
If 39 is not prime, then it must have a prime divisor p ≤ ⌊ 39⌋ =
6. We find that
39 = 3 ⋅ 13.
Finally we observe that √
13 is prime, since 13 is not divisible by
any prime number p ≤ ⌊ 13⌋ = 3.

Trial division allows us to enumerate all prime numbers

√ ≤n
in an inductive way: If we know all prime numbers p ≤ ⌊ n⌋ < n,
then we can decide whether n is prime.
2. NUMBERS 42

Example 2.5.3 We enumerate all prime numbers ≤ 11. Since

for the smallest prime divisor of n we have p ≤ m, wa can conlude
that
n m
2 1 ⇒ 2 prime
3 1 ⇒ 3 prime
4 2 4=2⋅2 ⇒ 4 not prime
5 2 2∤5 ⇒ 5 prime
6 2 6=2⋅3 ⇒ 6 not prime
7 2 2∤7 ⇒ 7 prime
8 2 8=2⋅4 ⇒ 8 not prime
9 3 9=3⋅3 ⇒ 9 not prime
10 3 10 = 2 ⋅ 5 ⇒ 10 not prime
11 3 2 ∤ 11 and 3 ∤ 11 ⇒ 11 prime

Practically one proceeds the opposite way, and rules out mul-
tiples of prime numbers, which already have been conputed.:

Algorithm 2.5.4 (Sieve of Eratosthenes) We obtain a list

of all primes up to N ∈ N, N ≥ 4 as follows:

1) Make a boolean list L with one entry for every number

2, ..., N . Mark all numbers as prime (true). Set p = 2.

2) Mark all j ⋅ p with j ≥ p as not prime (f alse).

3) Find the
√ smallest q > p, which ist marked as prime (true).
If q > N return L. Set p ∶= q, goto (2).

Proof. In step (2) all j ⋅ p with 2 ≤ j < p are already marked

as f alse from previous steps, since they have a prime divisor
< p. Hence all true multiples of p are marked as f alse. The
number q in step (3) is always prime, since p is the largest prime
number < q, and hence from previous steps all multiples j ⋅x of all
prime numbers x < q are marked as f alse. Once the algorithm
terminates, all
√ numbers are marked as f alse, which have a prime
number p ≤ N as a true divisor, that is which are not prime.
2. NUMBERS 43

Example 2.5.5 We find all primes ≤ 15 and give in ever itera-

tion the list of all j with Lj = true:

2 3 4 5 6 7 8 9 10 11 12 13 14 15
p=2 2 3 5 7 9 11 13 15
p=3 2 3 5 7 11 13

In the first iteration, we delete all multiples of 2, in the second

iteration all √multiples of 3. All remaining numbers are prime,
since p = 5 > 15.

For large numbers, there are much more efficient ways than
trial division to find a prime divisor of a given number.

2.6 Exercises
Exercise 2.1 Let n ∈ N and M ⊂ {1, ... , 2n} a set of integers
with ∣M ∣ = n+1 elements. Show that in M there are two different
integers such that the one divides the other.

Exercise 2.2 Show:

1) On M = N0 × N0 by

(a, b) ∼ (c, d) ⇔ a + d = b + c

we can define an equivalence relation.

2) The operations addition and multiplication

[(a, b)] + [(c, d)] = [(a + c, b + d)]

[(a, b)] ⋅ [(c, d)] = [(a ⋅ c + b ⋅ d, a ⋅ d + b ⋅ c)]

on
Z = (N0 × N0 ) / ∼
are well-defined, associative, commutative and distributiv.

More generally, these properties play an important role in the

context of groups and rings.
2. NUMBERS 44

Exercise 2.3 1) Let m ∈ N and a, b ∈ Z. Then we say that a

is congruent to b modulo m

a ≡ b mod m

if m ∣ (a − b). Prove that ”being congruent modulo m” is

an equivalence relation.

2) Write a function, which decides a ≡ b mod m.

Exercise 2.4 Show:

1) If r ∈ N and p = 2r − 1 is prime, then r is prime.

2) If r ∈ N and p = 2r + 1 is prime, then r = 2k with k ∈ N0 .

Exercise 2.5 Test the prime number theorem experimentally in

Maple:

1) Write a procedure, which computes

π (x) = ∣{p ≤ x ∣ p ∈ N prime}∣

for x > 0.

2) Compare the function π(x)x with ln(x)−a for a ∈ Z≥0 , in par-

1

ticular for large x. For which a do you get the best approx-
imation?

3) Visualize your observations using Maple.

Hint: Use the Maple-function nextprime.

Exercise 2.6 Let PN be the probability, that two randomly chose

natural numbers n, m ≤ N are coprime. Determine PN for N =
106 , 1012 and 1018 approximatively by samples of 102 , 104 and 106
generated with a computer algebra system. Check experimentally,
that PN for large values of N takes the value
6
≈ 60.7%.
π2
2. NUMBERS 45

Exercise 2.7 Implement the extended Euclidean algorithm. Test

your implementation at examples.

Exercise 2.8 Reduce

90297278063
18261358091
to smallest terms.

Exercise 2.9 Into an originally empty accout there regularily

get paid in 2809 e, and occasionally there get drawn 10403 e
from the account. Is it possible that the account at some point
has the balance of 1 e?

Exercise 2.10 Implement

1) the sieve of Erathosthenes.

2) the factorization of integers via trial division.

3) Find the prime factorization of

116338867864982351.

Exercise 2.11 Find the set L ⊂ Z of all solutions x of the si-

multaneous congruences

x ≡ 2 mod 3
x ≡ 2 mod 7
x ≡ 3 mod 10

Exercise 2.12 Let a1 , a2 ∈ Z and n1 , n2 ∈ Z>0 . Show that the

simultaneous congruences

x ≡ a1 mod n1
x ≡ a2 mod n2

are solveable if and only if

a1 − a2 ≡ 0 mod gcd (n1 , n2 ) .

Show that the solution is unique modulo lcm (n1 , n2 ).

2. NUMBERS 46

Figure 2.1: Two configurations of gearwheels.

Exercise 2.13 Is it possible for the two configurations of gear-

wheels in Figure 2.1 to transform the one into the other by rota-
tion? If possible, by how many ticks we have to turn?

Exercise 2.14 Find the set L ⊂ Z of all solutions x of the si-

multaneous congruences

x ≡ 1 mod 108
x ≡ 25 mod 80

Exercise 2.15 Using your implementation of the extended Eu-

clidean algorithm (or any other available implementation like the
Maple-function igcdex) write a procedure, which determines the
solution set of the simultaneous congruences

x ≡ a1 mod n1
x ≡ a2 mod n2

for given a1 , a2 ∈ Z and n1 , n2 ∈ Z>0 with gcd(n1 , n2 ) = 1. Com-

pare with the Maple-function chrem.
Extend your implementation such that it works correctly also
for n1 , n2 not coprime.
3

Groups

3.1 Overview
In this chapter we discuss the foundations of group theory, which
will have various applications in the chapters of rings, fields and
vector spaces. As and example for groups we consider symme-
try groups of subsets of Rn , for example, the sets of rotations
and (roto-) reflections, which map a Platonic solid (tetrahedron,
cube, octahedron, dodecahedron and icosahedron) to itself (see
Figure 3.1).The group property arises here from the fact, that
the composition of two symmetries is again a symmetry and any
symmetry can be undone by a symmetry. For example in the
symmetry group of the tetrahedron the 120○ rotation is equal to
the composition of two reflections, see Figure 3.2.
In general, we have: The composition of two symmetries is
again a symmetry. For every symmetry there is an inverse sym-
metry, such that the composition gives the identity map.
In the context of symmetry groups, the concept of an action
of a group G on a set M plays an important role. For example,
we can consider for G the symmetry group of the tetrahedron
and for M the tetrahedron or the sets of vertices or edges or faces
of the tetrahedron. A group action is then a map (satisfying a
couple of obvious additional conditions)

G×M Ð→ M
(g, m) z→ g⋅m

that is a group element g maps an element m ∈ M to another

47
3. GROUPS 48

Figure 3.1: The Platonic solids

Figure 3.2: Composition of two symmetries of the tetrahedron.

element of M , which we call g ⋅ m. If we start with some m

and apply all elements of G, then we obtain the so-called orbit
of m, for example, we can map any vertex of a tetrahedron to
any other vertex by applying a symmetry. In general, M will
decompose into disjoint orbit. One of the key theorems in this
context is the orbit counting formula.
The two most important examples of operations for the con-
struction and classification of groups is that of a subgroup H ⊂ G
by
H × G Ð→ G
(h, g) z→ hg
3. GROUPS 49

and of G on itself by conjugation

G × G Ð→ G
(a, b) z→ aba−1
The translation we will study in detail, the conjugation will be
discussed in the exercises.

3.2 Groups and actions

3.2.1 Basics
Definition 3.2.1 A group (G, ○) is a set G together with a map
○ ∶ G × G Ð→ G
(a, b) ↦ a○b
called operation, which satisfies the following axioms:
(G1) Associativity
a ○ (b ○ c) = (a ○ b) ○ c ∀a, b, c ∈ G

(G2) There exist a neutral element , that is an

e∈G
with
e ○ a = a ○ e = a ∀a ∈ G
(G3) Existence of inverses, that is ∀a ∈ G ∃a−1 ∈ G with
a−1 ○ a = a ○ a−1 = e

If in addition the commutative law

a ○ b = b ○ a ∀a, b ∈ G,
is obeyed, we call G abelian.
A set G together with an operation
○ ∶ G × G Ð→ G
which obeys (G1), is called a semigroup.
(G, ○) with (G1) and (G2) is called a monoid.
The number of elements ∣G∣ of G is called the order of G
(which can be ∞).
3. GROUPS 50

Remark 3.2.2 If in the definition of a group G we only require

the existence of a left-neutral element e ∈ G with e ○ a = a ∀a ∈ G
and left-inverse elements a−1 for every a ∈ G with a−1 ○ a = e,
then e is also right-neutral and the elements a−1 right-inverses:

1) For a, b ∈ G we have: If a ○ b = e, then also b ○ a = e.

2) We have that a ○ e = a for all a ∈ G.

Remark 3.2.3 If G is a group then:

1) The neutral element of G is unique.

2) The inverses of the elements of G are unique.

3) For a, b ∈ G we have (a ○ b) = b−1 ○ a−1 .

−1

4) For a ∈ G we have (a−1 )

−1
= a.

These statements are shown in Exercise 3.2.

Appart from the symmetry groups mentioned in Section 3.1,
we discuss the following central examples of groups:

Example 3.2.4 1) The set of integers with addition

(Z, +)

is a group. The neutral element is 0.

2) The set of integers with multiplication

(Z, ⋅)

is a monoid. The neutral element is 1.

3) The set of non-zero rational numbers together with multi-

plication
(Q/{0}, ⋅)
is a group.
3. GROUPS 51

4) Let X be an arbitrary set. The set of self-mappings of X

S (X) = {f ∶ X Ð→ X ∣ f bijective}

together with composition is a group.

In particular, for
X = {1, ..., n} ,
the set of permutations of n elements

Sn ∶= S ({1, ..., n})

is called the symmetric group. Obviously, we have that

∣Sn ∣ = n ⋅ (n − 1) ⋅ ... ⋅ 2 ⋅ 1 = n!

For σ ∈ Sn we also write

1 ⋯ n
σ=( )
σ (1) ⋯ σ (n)

An element of Sn is called a transposition, if it inter-

changes two unique elements.
Through numbering the corners of the tetrahedron in Fig-
ure 3.3 the rotation in Figure 3.3 can be identified by the
permutation
1 2 3 4
( ) ∈ S4
1 3 4 2
and the reflexion in Figure 3.4 with the transposisiton

1 2 3 4
( ) ∈ S4 .
1 3 2 4

5) Let
A = {α, β, γ, ...}
be a finite set. A word over the alphabet A is a finite
sequence
w = b1 b2 ...bn
3. GROUPS 52

Figure 3.3: Rotational symmetry of the tetrahedron

with bi ∈ A. Given another word v = a1 ...am , we define the

operation "writtenaftereachother" by
w ○ v = b1 ...bn a1 ...am .
The set
G = {w ∣ w a word over A}
together with ○ form a semigroup.
If we allow the empty word e in G, then (G, ○) becomes a
monoid.
6) If we add the letters α−1 , β −1 , ... with the calculation rule
αα−1 = α−1 α = e,
then we get the free group generated by A.
7) If G1 , G2 are groups, then the cartesian product G1 × G2
of G1 and G2 with the operation
(a1 , b1 ) ○ (a2 , b2 ) ∶= (a1 ○ a2 , b1 ○ b2 )
is again a group.

Definition and Theorem 3.2.5 (subgroup criterion) Let (G, ○)

be a group. A subset H ⊂ G is called a subgroup, if the following
two equivalent conditions are satisfied:
3. GROUPS 53

Figure 3.4: Reflection symmetry of the tetrahedron

1) (H, ○) is a group (that is, e ∈ H and a, b ∈ H Ô⇒ a ○ b ∈ H,

b−1 ∈ H)

2) H ≠ ∅, and a, b ∈ H Ô⇒ a ○ b−1 ∈ H.

Proof. (1) ⇒ (2) is obvious. Is on the other hand H ≠ ∅, then

there is an a ∈ H. For this element we have e = a ○ a−1 ∈ H, and
thus for all a ∈ H, that a−1 = e ○ a−1 ∈ H. Also for all a, b ∈ H we
have b−1 ∈ H, and hence
−1
a ○ b = a ○ (b−1 ) ∈ H.

Example 3.2.6 Let G be the symmetry group of the tetraederon,

r120 the rotation in Figure 3.3 and s23 the reflection in Figure 3.4.
Then

{id, r120 , (r120 )2 } ⊂ G

{id, s23 } ⊂ G

are subgroups.

Example 3.2.7 The subgroups of (Z, +) are of the form

nZ ∶= {n ⋅ k ∣ k ∈ Z}

where n ∈ Z≥0 .
3. GROUPS 54

Proof. Using the subgroup criterion, we see that nZ ⊂ Z is.a

subgroup. Suppose, on the other hand, that H ⊂ Z is a subgroup.
Then either H = {0} or there is a smallest element n > 0 in H.
We show that then H = nZ: Let m ∈ H. Division with remainder
yields a representation of m
m = qn + r
with 0 ≤ r < n and r ∈ H. By definition of n it follows that r = 0,
hence m ∈ nZ.
Example 3.2.8 Let n ∈ N and a ∈ Z. The equivalence class
(residue class) of a modulo n can be expressed by using the sub-
group nZ ⊂ Z as
a = {b ∈ Z ∣ a ≡ b mod n}
= a + nZ ∶= {a + b ∣ b ∈ nZ} = {a + k ⋅ n ∣ k ∈ Z} ⊂ Z
(see also Exercise 2.3).
The set of residue classes
Zn ∶= Z/n ∶= {0, 1, 2, ..., n − 1}
is together with the operation
a + b ∶= a + b
a group, the group of residue classes modulo n (with neutral
element 0 and inverse −a = −a of a ∈ Z/n).
Since a + b ∶= a + b is not defined in terms of a and b, but in
terms of representatives a and b, we have to show that a+b is well-
defined, that is, does not depend on the choice of represenatives
of a and b:
If a1 = a2 and b1 = b2 , that is, a1 − a2 = n ⋅ k1 and b1 − b2 = n ⋅ k2
with numbers k1 , k2 , then
a1 + b1 = a1 + b1 = a2 + b2 + n ⋅ (k1 + k2 ) = a2 + b2 = a2 + b2 .
Example 3.2.9 For n = 3 we get Z/3 = {0, 1, 2} with
0 = {..., −3, 0, 3, 6, ...} = 3Z
1 = {..., −2, 1, 4, 7, ...} = 1 + 3Z
2 = {..., −1, 2, 5, 8, ...} = 2 + 3Z
3. GROUPS 55

Figure 3.5: residue classes modulo 3

see also Figure 3.5.

The operation can be described in terms of the group table
+ 0 1 2
0 0 1 2
1 1 2 0
2 2 0 1

for example, we see that 2 + 2 = 2 + 2 = 4 = 1.

Example 3.2.10 For every divisor a of n, and d = n
a the subset

{0, a, 2a, ..., (d − 1)a} ⊂ Z/n

is a subgroup (exercise).
For example, for n = 6 and a = 2 we ge the subgroup
{0, 2, 4} ⊂ Z/6.
If we compare the group table
+ 0 2 4
0 0 2 4
2 2 4 0
4 4 0 2
of this group with that of Z/3, we observe, that the elements of
the two groups have different names, but obey the same calcu-
lation rules.
The identification of the subgroup {0, 2, 4} ⊂ Z/6 with Z/3
is an example of a group isomorphism, that is, a bijective map,
which is compatible with the group structures. The group iso-
morphism
ϕ∶ Z/3 Ð→ {0, 2, 4}
0 + 3Z z→ 0 + 6Z
1 + 3Z z→ 2 + 6Z
2 + 3Z z→ 4 + 6Z
3. GROUPS 56

satisfies, for example,

ϕ (1 + 1) = ϕ (2) = 4 = 2 + 2 = ϕ (1) + ϕ (1) .

We write then
Z/3 ≅ {0, 2, 4}
and more generally we have

Z/d ≅ {0, a, 2a, ..., (d − 1)a} .

Definition 3.2.11 A group homomorphism ϕ between two

groups G1 and G2 is a map

ϕ ∶ G1 Ð→ G2

which satisfies

ϕ (a ○ b) = ϕ (a) ○ ϕ (b) ∀a, b ∈ G1

that is, which is compatible with the group structures.

Note that ○ on the left side is the operation in G1 , and on

the right side is the operation in G2 .

Remark 3.2.12 If ϕ ∶ G1 Ð→ G2 is a group homomomorphism,

then
ϕ (e1 ) = e2
where ei ∈ Gi denotes the respective neutral element.
The kernel of ϕ

Ker ϕ = {a ∈ G1 ∣ ϕ (a) = e2 }

and the image of ϕ

Im ϕ = ϕ (G1 )
are subgroups of G1 and G2 , respectively.

For the proof see Exercise 3.4.

For example, for the above group homomorphismus ϕ ∶ Z/3 →
Z/6 given by 1 ↦ 2, we have

Im ϕ = {0, 2, 4}
Ker ϕ = {0} .
3. GROUPS 57

Lemma 3.2.13 A group homomorphism ϕ ∶ G1 Ð→ G2 is injec-

tive if and only if
Ker ϕ = {e1 } ,
that is, the kernel contains only the neutral element e1 of G1 .

Proof. We first remark, that for b ∈ G1

(ϕ (b))−1 = ϕ (b−1 )

since
ϕ (b) ○ ϕ (b−1 ) = ϕ(b ○ b−1 ) = ϕ(e1 ) = e2 ,
and the inverse is unique. For a, b ∈ G1 we hence have

ϕ (a) = ϕ (b) ⇐⇒ ϕ (a ○ b−1 ) = e2 ⇐⇒ a ○ b−1 ∈ Ker ϕ.

using that

ϕ (a) ○ (ϕ (b))−1 = ϕ (a) ○ ϕ (b−1 ) = ϕ(a ○ b−1 )

Hence if Ker ϕ = {e1 }, then ϕ (a) = ϕ (b) implies, that a = b.

Is on the other hand ϕ injective, then it follows from

ϕ (a) = e2 = ϕ (e1 )

that a = e1 .

Definition 3.2.14 Injective group homomorphisms are called

(group-) monomorphisms,surjective (group-)homomorphisms
(group-) epimorphisms.
A (group-) isomorphism

ϕ ∶ G1 Ð→ G2

ia a bijective homomorphism. The inverse map

ϕ−1 ∶ G2 Ð→ G1

is then also a isomorphism. We write G1 ≅ G2 .

See also Exercise 3.4.

Example 3.2.15 1) The inclusion of a subgroup H ↪ G is a

monomorphism.
3. GROUPS 58

2) The map
Z Ð→ nZ
k z→ n⋅k
is for n ≥ 1 an isomorphism.

3) The exponential function

(R, +) Ð→ (R>0 , ⋅)
x z→ exp (x) = ex

in Figure 3.6 is a homomorphism, since by the functional

equation of the exponential function ex1 +x2 = ex1 ⋅ ex2 for all
xi ∈ R. Since the exponential function is continuous and
strictly monotone with limx↦∞ ex = ∞ and limx↦−∞ ex = 0,
it defines even an Isomorphismus.
4

–2 –1 0 1 2

Figure 3.6: Exponential function

4) In contrast, with C∗ = C/{0} the map

(C, +) Ð→ (C∗ , ⋅)
z z→ exp (z) = ez

is an epimorphism, but not an isomorphism. The kernel is

Ker (exp ∶ C Ð→ C∗ ) = 2πiZ ∶= {2πin ∣ n ∈ Z} .

3. GROUPS 59

5) Let n ≥ 2. The signatur or the signum

sign ∶ Sn Ð→ ({±1} , ⋅)
n
σ(i)−σ(j)
σ z→ sign (σ) = ∏ i−j
i,j=1
i<j

is an epimorphism and

Ker (sign) = An

is called alternating group.

The definition of sign translates into the following program
(in the Syntax of Maple):
sgn:=proc(sigma)
local s,j,i;
s:=1;
for j from 1 to nops(sigma) do
for i from 1 to j-1 do
s:=s*(sigma[i]-sigma[j])/(i-j);
od;
od;
return(s);
end proc:

where we represent the permutation σ by the list (σ(1), ... , σ(n)).

As an example, we consider the permutation from Figure
3.2. For the rotation
1 2 3 4
σ=( )
1 3 4 2

using the above formula we obtain

1−3 1−4 1−2 3−4 3−2 4−2
sign(σ) = ⋅ ⋅ ⋅ ⋅ ⋅
1−2 1−3 1−4 2−3 2−4 3−4
3−2 4−2
= ⋅ = (−1)2 = 1
2−3 2−4
and for the two reflections
1 2 3 4 1 2 3 4
τ1 = ( ) τ2 = ( )
1 3 2 4 1 2 4 3
3. GROUPS 60

that sign(τi ) = −1. Indeed we have for all transpositions τ ,

that sign(τ ) = −1. We will prove this below.
Since sign is a group homomorphism, it follows from σ =
τ1 ⋅ τ2 directly that

sign(σ) = sign(τ1 ) ⋅ sign(τ2 ) = 1.

As we will see , one can easily compute the signum via the
homomorphismu-property, by writing a permutation as a
product of permutations with known signum.
See also Exercise 3.5.

6) If a, b ∈ N and gcd (a, b) = 1. Then

Z/ab ≅ Z/a × Z/b

This is a reformulation of the Chinese remainder theorem.

In practice, groups are often specified in terms of generators:

Definition 3.2.16 Let E be a subset of a group G. Then ⟨E⟩

is the smallest subgroup of G, which contains all elements of E.
Equivalently ⟨E⟩ is the intersection of all subgroups U with E ⊂
U ⊂ G (since the intersection of subgroups is again a subgroup).
We call ⟨E⟩ the subgroup generated by E.
A group G is called cyclic, if there is a g ∈ G with

G = ⟨g⟩ .

For g ∈ G we obviously have

⟨g⟩ = {g r ∣ r ∈ Z}

with
g r = g ○ ... ○ g
´¹¹ ¹ ¹ ¹ ¹ ¹ ¸¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
r

and = gr for r < 0. If the operation is written additively

(g −1 )−r
as + we write r ⋅ g instead of g r .

Example 3.2.17 1) The residue class group Z/n is cyclic

generated by 1.
3. GROUPS 61

2) The group (Z, +) cyclic generated by 1.

3) The subgroup nZ ⊂ (Z, +) is cyclic generated by n, so nZ =

⟨n⟩. According to Example 3.2.15 we have nZ ≅ Z.

We will prove later, that all cyclic groups are up to isomor-

phism of the form Z or Z/n (see Example 3.3.13).

Definition 3.2.18 Let g ∈ G be an element of a group. Then

ord (g) = ∣⟨g⟩∣

is called the order of g.

See also Exercise 3.9.

Example 3.2.19 For the rotation of the tetrahedron by 120○

1 2 3 4
σ=( )
1 3 4 2

we have
⟨σ⟩ = {id = σ 0 , σ 1 , σ 2 } ≅ Z/3
and hence ord (σ) = 3.

3.2.2 Group actions

Groups are considered in mathematics mainly since they can be
used to describe symmetries. In order to interpret groups as
groups of symmetries, the effect of the elements of a group on
the elements of a set are specified through the notion of a group
action:

Definition 3.2.20 Let (G, ○) be a group and M a set. An ac-

tion of G on M (from the left) is a map

⋅∶ G×M Ð→ M
(g, m) z→ g⋅m

which satisfies the following conditions:

3. GROUPS 62

1)
e⋅m=m
for all m ∈ M .
2)
(a ○ b) ⋅ m = a ⋅ (b ⋅ m)
for all a, b ∈ G and m ∈ M .
Remark 3.2.21 In an analogous way, one can specify operatons
from the right
⋅ ∶ M × G Ð→ M
(m, g) z→ m ⋅ g
with m ⋅ e = m and (m ⋅ a) ⋅ b = m ⋅ (a ○ b).
It seems to be superfluous to have both notations, however
there are settings, where there are two distinct canonical defini-
tions for an action from the left and from the right, and where
both definitions are used at the same time. An example is the
action of a subgroup H ⊂ G on G by H × G → G, (h, g) ↦ h ○ g
from the left, and by G × H → G, (g, h) ↦ g ○ h from the right,
which we will later discuss in detail.
Remark 3.2.22 To put it differently, an action of G on M is
a group homomorphism
ϕ∶ G Ð→ S (M )
M Ð→ M
g ↦ ϕ (g) ∶= ( )
m ↦ g⋅m
of G into the group of self-mappings of M .
Proof. We check whether ϕ (g) for alle g ∈ G is bijective and
whether ϕ is a homomorphism: Let g ⋅m1 = g ⋅m2 for m1 , m2 ∈ M .
Then
m1 = e ⋅ m1 = (g −1 ○ g) ⋅ m1 = g −1 ⋅ (g ⋅ m1 )
= g −1 ⋅ (g ⋅ m2 ) = (g −1 ○ g) ⋅ m2 = e ⋅ m2 = m2 .
Each m ∈ M is in the image of ϕ (g), since m = e⋅m = g ⋅(g −1 ⋅m).
Moreover,
ϕ (g ○ h) = (m ↦ (g ○ h) ⋅ m) = (m ↦ g ⋅ (h ⋅ m))
= (m ↦ g ⋅ m) ○ (m ↦ h ⋅ m) = ϕ (g) ○ ϕ (h) .
3. GROUPS 63

Example 3.2.23 Sn acts on {1, ..., n} by

Sn × {1, ..., n} Ð→ {1, ..., n}
(σ, j) z→ σ(j)
Another key example is the action of the group of motions of
Rn :
Definition 3.2.24 A Euclidean motion f ∶ Rn → Rn is a map,
which respects the Euclidean distance
√
∥x∥ ∶= ∑ni=1 x2i
that is, with
∥x − y∥ = ∥f (x) − f (y)∥
for all x, y ∈ Rn . Figure 3.7 shows a motion, which is the com-
position of a translation and a roto-reflection. The set E (n) of

Figure 3.7: Example of a motion of R2 .

Euclidean motions of Rn is with the composition a group, the

group of motions.
Let M ⊂ Rn be a subset. The group
Sym (M ) = {A ∈ E (n) ∣ A (M ) = M }
is called the symmetry group of M .
Example 3.2.25 (Symmetry group) We describe the symme-
try group Sym (D) of the equilateral triangle D.
 3. GROUPS 64

Every symmetry is a rotation or reflection

id

Every symmetry is uniquely determined by its action on the ver-

tices of the triangle. By numbering the vertices, we can consider
every symmetry as a bijective map {1, 2, 3} → {1, 2, 3}. Indeed,
we have a group isomorphism ϕ

Sym (D) = {id }

ϕ↓ ↧ ↧ ↧ ↧ ↧ ↧

S3 = { id (1 ↔ 2) (1 ↔ 3) (2 ↔ 3)}

which is induced by the action of Sym (D) on the vertices of the

triangle
Sym (D) × {1, 2, 3} Ð→ {1, 2, 3} .
Is, for example,
r120 =

the rotation by 120○ , then the operation gives a map

(r120 , 1) ↦ 2, (r120 , 2) ↦ 3, (r120 , 3) ↦ 1

hence
1 2 3
ϕ(r120 ) = ( )
2 3 1
3. GROUPS 65

Example 3.2.26 (Orbit and Stabilizer) Given a point of the

equilateral triangle D, we want to investigate on which other
points under the operation

Sym (D) × D Ð→ D

this point can map. This set is called the orbit, the number of
elements the lenght of the orbit. Examples of orbits are

The operation on D induce an operation

Sym (D) × 2D Ð→ 2D

of the set of all subsets of D. In the orbit of the black subset is

also the white subset:

Otherwise one can consider the set of all elements of Sym (D),
that keep a given point (or a subset) fixed. The corner 1 is fixed
by {id, (2 ↔ 3)}, the middlepoint m by Sym (D) and the point
p1 only by the identity. The black subset is fixed by

{id, , }

We observe that these sets are subgroups of Sym (D), and that
the product of the group orders with length the respective orbits,
3. GROUPS 66

in each case ∣Sym (D)∣ = 6, give.

Orbit fixed by
1 {1, 2, 3} {id, (2 ↔ 3)} 3 ⋅ 2 = 6
m {m} Sym (D) 1⋅6=6
p1 {p1 , ..., p6 } {id} 6⋅1=6

This is true in general, and is called the orbit counting theorem

(see Section ??).

We first formalize the ideas of orbit and stabilizer:

Definition 3.2.27 Let G × M → M be an action. For m ∈ M

we call
Gm = {g ⋅ m ∣ g ∈ G} ⊂ M
the orbit of m. If N ⊂ M is any subset, then

Stab (N ) = {g ∈ G ∣ gN = N } ,

where gN = {g ⋅ n ∣ n ∈ N }, is called the stabilizer of N .

The most important case is that of the stabilizer of a 1-
element set: For an element m ∈ M let

Stab (m) = {g ∈ G ∣ g ⋅ m = m} = Stab ({m}) .

Remark 3.2.28 Two orbits Gm1 and Gm2 are either equal or
disjoint. To be in the same orbit is, hence, an equivalence rela-
tion.

Proof. If there exists an

m3 ∈ Gm1 ∩ Gm2

then there are g1 , g2 ∈ G with

m3 = g1 ⋅ m1 = g2 ⋅ m2

hence
m2 = g2−1 ⋅ (g1 ⋅ m1 ).
For every g ∈ G we hence have

g ⋅ m2 = g ⋅ (g2−1 ⋅ (g1 ⋅ m1 )) = (g ○ g2−1 ○ g1 ) ⋅ m1 ∈ Gm1

3. GROUPS 67

that is
Gm2 ⊂ Gm1 .
Similarily we have the other inclusion, that is Gm2 = Gm1 .
The second claim one easily checks using the definition of an
equivalence relation.

Definition 3.2.29 The set of orbits we denote by M /G (quo-

tient of M by G). Every element m ∈ Gm1 we call a represen-
tative of the orbit, since Gm = Gm1 . Morover,

π∶ M Ð→ M /G
m z→ Gm

is called the quotient map.

With the above remark we see:

Definition and Theorem 3.2.30 Let G × M → M be an ac-

tion. A complete set of representatives of the orbits is a
subset R ⊂ M , such that every orbit Gm contains exactly one
element of R.
Then M is the disjoint union
⋅
M = ⋃G ⋅ r
r∈R

The representation of permutations in mapping notation is

not efficient: For the permutation

1 2 3 4 5 6 7 8 9
σ=( )
2 3 1 4 5 6 7 8 9

we do not have to remember the images of 4, ... , 7. The images

of 1, 2, 3 we can encode in a diagram

This is the idea of a so-called cycle:

3. GROUPS 68

Definition 3.2.31 If σ ∈ Sn , then the action of ⟨σ⟩ on the set

{1, ..., n} decomposes the set into orbits

⟨σ⟩ x = {x, σ (x) , σ 2 (x) , ..., σ t−1 (x)}

and t minimal with σ t (x) = x. If there is only one orbit of length

t > 1 (that is, all others have length 1), then σ is called a cycle
of order t, and we write

σ = (x, σ (x) , σ 2 (x) , ..., σ t−1 (x)) ,

that is we code, in addition to the orbit, also the order in which

we go through the orbit by iteratively applyíng σ. Transposi-
tions are cycles of length 2. For the neutral element we write
().

One could also use for a cycle a circle notation as above, but
that would use a bit too much space, and would not be easy to
enter on a computer console.

Remark 3.2.32 The cycle

σ = (a1 , ... , at ) ∈ Sn

is thus the map

{1, ... , n} Ð→ {1, ... , n}

a1 z→ a2
a2 z→ a3
⋮
at−1 z→ at
at z→ a1
a z→ a otherwise,

and ord(σ) = t.

Example 3.2.33 For the rotation

1 2 3 4
σ=( )
1 3 4 2
3. GROUPS 69

of the tetrahedron by 120○ (see also Example 3.2.19) we obtian

the decomposition into orbits

{1, 2, 3, 4} = {1} ∪˙ {2, 3, 4} .

Hence σ is a cycle and by remembering the order in which we go

through the orbit by iteratively applying σ, we get

σ = (2, 3, 4),

that is, 2 ↦ 3, 3 ↦ 4, 4 ↦ 2. The notation of a cycle one can

open up at any point, hence

σ = (2, 3, 4) = (3, 4, 2) = (4, 2, 3).

The rotation
1 2 3 4
σ2 = ( )
1 4 2 3
of the tetrahedron by 240○ gives the same decomposition into or-
bits {1, 2, 3, 4} = {1} ∪˙ {2, 3, 4}, but

σ 2 = (2, 4, 3).

Note that it is easy to compute square of cycles σ: We just have

to go two steps in the cycle to get the image of an element (and
similarly for higher powers).

Not every permutation is a cycle: Under the action of

1 2 3 4 5 6 7 8
σ=( )
2 3 4 1 6 7 8 5

there are two orbits of length 4 and

σ = (1, 2, 3, 4) ○ (5, 6, 7, 8)

is the product of two cycles. Since orbits are disjoint, we can

always decompose a permutation into disjoint cycles:

Theorem 3.2.34 We have:

1) Every element of Sn is a product of disjoint cycles.

3. GROUPS 70

2) Ever element of Sn is a product of transpositions.

Proof. Let σ ∈ Sn .

1) Let {x1 , ..., xr } be a complete set of representatives of the

orbits of the operation of ⟨σ⟩ on {1, ..., n}. If we restrict σ
as a map on the orbit ⟨σ⟩ xi , then we obtain a cycle σi and

σ = σ1 ○ ... ○ σr

2) Using 1) we may assume, that σ is a cycle (y0 , ..., yt−1 ).

Then, as one easily checks,

(y0 , ..., yt−1 ) = (y0 , y1 ) ○ ... ○ (yt−2 , yt−1 ) .

In the cycle notation of permuations, one usually leaves the

symbol ○ away.

Example 3.2.35 Let

1 2 3 4 5 6 7 8 9
σ=( )
4 1 2 3 9 8 7 6 5

The operation of ⟨σ⟩ decomposes

{1, ..., 9} = {1, 2, 3, 4} ∪˙ {5, 9} ∪˙ {6, 8} ∪˙ {7}

in disjoint orbits

σ = (1, 4, 3, 2) (5, 9) (6, 8)

= (1, 4) (4, 3) (3, 2) (5, 9) (6, 8) .

See also Exercise 3.7.

Remark 3.2.36 If σ = τ1 ○ ... ○ τr with transpositions τi , then we

can compute the signum of σ directly as

sign(σ) = (−1)r ,

since sign is a group homomorphism and sign τ = −1 for all trans-

positions τ .
3. GROUPS 71

Proof. For a transposition τ , we compute

τ (i) − τ (j)
sign (τ ) = ∏ .
i<j i−j

Assume τ = (k, l) and k < l.Then

⎧
⎪
⎪ −1 for i = k and j = l
⎪
τ (i) − τ (j) ⎪⎪
⎪ 1 for i, j ∉ {k, l}
= ⎨ l−j
i−j ⎪
⎪ k−j for i = k and j ≠ l
⎪
⎪
⎪
⎪ i−k
⎩ i−l for i ≠ k and j = l.

If we further split up the last two cases into the j with k = i < j < l
or l < j, and into the i with k < i < j = l or i < k, respectively,
then we get
l−j l−j i−k i−k
sign(τ ) = − ∏ ⋅ ∏ ⋅ ∏ ⋅ ∏ .
j with k − j j with k − j i with i − l i with i − l
l<j k<j<l k<i<l i<k
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸¹
>0
¹¹¹¹¹¹¹¹¹¶
>0 1

The second and the third product cancel, and all numerators an
denominators of the first and last product are negative. Hence
sign τ < 0 so sign τ = −1.

Example 3.2.37 For

σ = (1, 4, 3, 2) (5, 9) (6, 8)

= (1, 4) (4, 3) (3, 2) (5, 9) (6, 8)

we obtain
sign(σ) = (−1)5 = −1.

Remark 3.2.38 For a representation of a permutation σ = c1 ○

.... ○ cr as a product of disjoint cycles ci of length mi one can
compute the order of σ as

ord(σ) = lcm(m1 , ..., mr ).

For the proof see Exercise 3.9.

3. GROUPS 72

Example 3.2.39 For σ = (1, 4, 3, 2) (5, 9) (6, 8) we get

ord(σ) = lcm(4, 2, 2) = 4.

We can check this also by a direct calculation

σ 2 = (1, 4, 3, 2)2 (5, 9)2 (6, 8)2 = (1, 3)(2, 4)

σ 3 = (1, 2, 3, 4)(5, 9)(6, 8)
σ 4 = id .

For subgroups of Sn the computer algebra system GAP [13]

algorithms for computing essentially all object introduced in this
chapter.

Example 3.2.40 We determine ord(σ) for

σ = (1, 4, 3, 2) (5, 9) (6, 8)

using GAP:
sigma:=(1,4,3,2)(5,9)(6,8);
(1, 4, 3, 2)(5, 9)(6, 8)
sigma^2;
(1, 3)(2, 4)
sigma^3;
(1, 2, 3, 4)(5, 9)(6, 8)
sigma^4;
()
Hence ord(σ) = 4. To apply Remark 3.2.38, instead, we use:
Order(sigma);
4
Note that, in contrast to the usual convention, to compute
σ ○ τ for σ, τ ∈ Sn we have to enter τ ∗ σ in GAP (that is, maps
take their argument on the left hand side). We check in GAP,
that with τ = (2, 5) we get

σ ○ τ = (1, 4, 3, 2) (5, 9) (6, 8) ○ (2, 5)

= (1, 4, 3, 2, 9, 5)(6, 8).

tau:=(2,5);;
tau*sigma;
(1, 4, 3, 2, 9, 5)(6, 8)
3. GROUPS 73

Remark 3.2.41 The symmetric group S3 is generated by (1, 2)

and (2, 3)
S3 = ⟨(1, 2) , (2, 3)⟩
since (1, 2) (2, 3) = (1, 2, 3) and (1, 2) (2, 3) (1, 2) = (1, 3). In
general, we have

Sn = ⟨(1, 2) , (2, 3) , ..., (n − 1, n)⟩ ,

see also die Exercises 3.21 and 3.22.

Example 3.2.42 In GAP we can define a group via a generat-

ing system as follows:
G:=Group((1,2),(2,3));
Group([(1, 2), (2, 3)])
Elements(G);
[(), (2, 3), (1, 2), (1, 2, 3), (1, 3, 2), (1, 3)]

3.2.3 Action by translation

So far, we have considered the action of Sym(M ) on a set M
and of Sn on {1, ..., n}. A further very important example is the
action of a group (G, ○) on itself

G × G Ð→ G
(g, h) ↦ g ○ h

given by the group operation (this gives an action both from the
left and from the right). The action plays in important role in
the proof of the following theorem, which is of key importance
for practical calculations with groups: It allows to consider any
finite group as a subgroup of Sn . In this representation, the
group can then be handled by the computer.

Theorem 3.2.43 (Cayley) Every group G is isomorphic to a

subgroup of the group of self-mappings S (G).
In particular for n ∶= ∣G∣ < ∞, we can consider G as a sub-
group of Sn ≅ S (G).
3. GROUPS 74

Proof. The map

ϕ∶ G → S (G)
G → G
g ↦ ( )
h ↦ g○h
is a group homomorphism and
Ker ϕ = {g ∈ G ∣ g ○ h = h ∀h ∈ G} = {e}
(using the uniqueness of the neutral element), hence ϕ is injec-
tive. Hence,
G ≅ Im(ϕ) ⊂ S (G) .

For finite groups one can specify the action

G × G Ð→ G
(g, h) ↦ g ○ h
via a table, the group table.
Example 3.2.44 The group
G = Z/4 = {0, 1, 2, 3}
has the group table
+ 0 1 2 3
0 0 1 2 3
1 1 2 3 0
2 2 3 0 1
3 3 0 1 2
In ever row or column, every element of the group occurs exactly
once.
The rows of the table specify ϕ(g), in the example we have
0 1 2 3
ϕ(0) = ( )
0 1 2 3
0 1 2 3
ϕ(1) = ( )
1 2 3 0
0 1 2 3
ϕ(2) = ( )
2 3 0 1
0 1 2 3
ϕ(3) = ( )
3 0 1 2
3. GROUPS 75

where we use the map notation introduced for elements of Sn also

for elements of S(G).
Finally, we can number the elements of G to do the identifi-
cation S(G) ≅ S4 .
A group is abelian if and only if its group table is symmetric
with respect to the diagonal. The associaticity law one cannot
see in any obvious way from the table.
Analogous to the action of a group on itself, one can also
consider the action of a subgroup on the group:
Example 3.2.45 As shown in Example 3.2.7, the subgroups of
(Z, +) are of the form
nZ = {n ⋅ k ∣ k ∈ Z} .
A group action of nZ on Z (from the right) is given by
Z×nZ→ Z
(a, n ⋅ k) ↦ a + n ⋅ k
The orbits are exactly the residue classes modulo n
a = a + nZ = {a + n ⋅ k ∣ k ∈ Z} .
Analogusly, we can also operate by addition from the left, and
obtain the same orbits, since + is commutative. Due to the usual
notation a + nZ for residue classes, one prefers the action from
the right.
For n = 4, we obtain by the action of 4Z on Z the orbits
4Z 1 + 4Z 2 + 4Z 3 + 4Z
⋮ ⋮ ⋮ ⋮
−4 −3 −2 −1
0 1 2 3
4 5 6 7
⋮ ⋮ ⋮ ⋮
As seen in Example 3.2.8, the set of orbits with the operation
a + b = (a + nZ) + (b + nZ) = (a + b) + nZ = a + b
is again a group, called Z/n.
Later, we will investigate, when a set of orbits of a subgroup
is again a group.
3. GROUPS 76

We first introduce the concept of orbits of a subgroup in

general:

Definition 3.2.46 (Cosets) Let H ⊂ G be a subgroup. Then

the operation in G defines an action of H on G

H × G Ð→ G, (h, g) z→ h ○ g

from the left, and similarly from the right

G × H Ð→ G, (g, h) z→ g ○ h.

For g ∈ G the orbits of this action

Hg ∶= H ○ g ∶= {h ○ g ∣ h ∈ H}

and
gH ∶= g ○ H ∶= {g ○ h ∣ h ∈ H}
are called the right and left cosets of g, respectively.

Theorem 3.2.47 Let H ⊂ G be a subgroup. Every two cosets of

H have the same number of elements.

Proof. Let a, b ∈ G. Then aH and bH are in bijection by

multiplication with ba−1 from the left
g z→ b ○ a−1 ○ g
1∶1
G Ð→ G
∪ ∪
aH Ð→ bH
a○h z→ b ○ a−1 ○ a ○ h = b ○ h
(what is the inverse?). The right and left cosets aH and Ha are
in bijection via conjugation with a
g z→ a−1 ○ g ○ a
1∶1
G Ð→ G
∪ ∪
aH Ð→ Ha
a○h z→ a−1 ○ a ○ h ○ a = h ○ a
(what is the inverse?). The operation by conjugation will also
be discussed in Exercise 3.12.
3. GROUPS 77

Corollary 3.2.48 (Index formula) Let H ⊂ G be a subgroup.

Then
∣G∣ = ∣G/H∣ ⋅ ∣H∣
in particular in a finite group G, the order of the subgroup ∣H∣
divides ∣G∣.

Definition 3.2.49 If H ⊂ G is a subgroup, then

[G ∶ H] ∶= ∣G/H∣

is called the index of H in G.

We remark, that

H → aH
h↦a○h

is a bijection (see the proof of Theorem 3.2.47), hence

∣aH∣ = ∣H∣ .

We prove now the index formula:

Proof. By Definition and Theorem 3.2.30 the group G is the
disjoint union of all aH with a from a complete set of represen-
atives R, so if ∣G∣ < ∞ then

∣G∣ = ∑ ∣aH∣ = ∣R∣ ⋅ ∣H∣

a∈R

(using Theorem 3.2.47). If ∣G∣ = ∞, then also ∣G/H∣ = ∞ or

∣H∣ = ∞.

Example 3.2.50 The group G = Z/6 has the order 6 and the
subgroups

{0, ..., 5}
Ò Ó
{0, 2, 4} {0, 3}
Ó Ò
{0}

with the orders 1, 2, 3 and 6.

3. GROUPS 78

Remark 3.2.51 Note that in a group, there is not necessarily

for every divisor a subgroup with the respective order, for exam-
ple,
A4 = {σ ∈ S4 ∣ sign(σ) = 1}
does not have a subgroup of the order 6. The following GAP
code computes all possible order of subgroups of A4 :
G:=AlternatingGroup(4);;
Order(G);
12
L:=ConjugacyClassesSubgroups(G);;
List(List(L,Representative),Size);
[ 1, 2, 3, 4, 12 ]
In the context of the so-called Sylow theorems, one can prove,
that there is a subgroup of the respective order for every prime
power divisor of ∣G∣.

From the index formula (Theorem 3.2.48) we get taking H =

⟨g⟩:

Corollary 3.2.52 In every finite group G the order of an ele-

ment g ∈ G is a divisor of the group order ∣G∣, that is, ord (g) ∣
∣G∣.

Example 3.2.53 In G = Z/6, the elements 1 and 5 = −1 have

order 6, the elements 2 and 4 have order 3, and the element 3
has order 2. The neutral element 0 has order 1.

Corollary 3.2.54 Every group G with ∣G∣ prime is cyclic.

Proof. Using the index formula, we get, that G only has the
subgroups {e} and G. Hence, for every e ≠ g ∈ G we have

{e} ≠ ⟨g⟩ = G
3. GROUPS 79

3.3 Normal subgroups

3.3.1 Normal subgroups and the quotient group
Let H be a subgroup of (G, ○) and
G/H = {gH ∣ g ∈ G}
the set of left cosets
gH = {g ○ h ∣ h ∈ H}
of H, that is, the set of orbits of the translation action G×H → G,
(g, h) ↦ g ○ h of H on G.
Example 3.3.1 For H = nZ ⊂ Z = G we have already seen that
the set
G/H = Z/n = {0, ... , n − 1}
of cosets
a = a + nZ
with the operation induced by the addition in Z
a + b ∶= a + b
has the structure of a group.
More generally, is it true that G/H with the operation
aH ⋅ bH ∶= (a ○ b)H
induced by that of G becomes a group? As in the case of Z/n
the problem arises from the question whether the operation is
well-defined, that is, whether it is independent from the choice
of representatives a, b of the cosets aH and bH.
Example 3.3.2 Let us go back to the respective calculation for
Z/n: Let a1 = a2 and b1 = b2 , that is, a1 = a2 +k ⋅n and b1 = b2 +l⋅n
for some k and l. Then
a1 + b1 = a2 + k ⋅ n + b2 + l ⋅ n = a2 + b2 + (k + l) ⋅ n,
hence a1 + b1 = a2 + b2 . In this calculation, we have commuted
the summands k ⋅ n and b2 . This was possible, since G = Z is
abelian.
3. GROUPS 80

Now to the general setting. If G/H with aH ⋅ bH = (a ○ b)H

is a group, then the quotient map
π ∶ G Ð→ G/H, g z→ gH
is a group homomorphism, since for all a, b ∈ G we have
π(a) ⋅ π(b) = aH ⋅ bH = (a ○ b)H = π(a ○ b).
Moreover
π (e) = eH = H ∈ G/H
is the neutral element, hence H = Ker (π). Indeed, for the kernel
of a group homomorphism, we observe in general:

Remark 3.3.3 Let

ϕ ∶ G Ð→ F
be a group homomorphism and
H = Ker (ϕ) ⊂ G.
Then for g ∈ G and
gHg −1 ∶= {g ○ h ○ g −1 ∣ h ∈ H} ,
we have
gHg −1 = H.

Proof. If h ∈ Ker ϕ, then for all g ∈ G

−1 −1
ϕ (g ○ h ○ g −1 ) = ϕ (g) ○ ϕ (h) ○ ϕ (g) = ϕ (g) ○ ϕ (g) = e,
so g ○ h ○ g −1 ∈ H and hence
gHg −1 ⊂ H.
If we replace g by g −1 , we get the other inclusion.
Subgroups with this property of the kernel, are called normal
subgroups:

Definition 3.3.4 A subgroup H ⊂ G is called a normal sub-

group of G (written H ⊲ G), if
gHg −1 = H for all g ∈ G
(equivalently gH = Hg for all g ∈ G or gHg −1 ⊂ H for all g ∈ G).
3. GROUPS 81

More generally than the above example we have:

Remark 3.3.5 If ϕ ∶ G Ð→ F is a group homomorphism and

M ⊂ F is a normal subgroup, then ϕ−1 (M ) ⊂ G is a normal
subgroup. If ϕ is surjective and N ⊂ G is a normal subgroup,
then ϕ (N ) ⊂ F is a normal subgroup.

We prove this in Exercise 3.17.

Theorem 3.3.6 Let H ⊂ G be a subgroup. The set G/H is with

the induced operation

aH ⋅ bH = (a ○ b)H

a group, if and only if H is a normal subgroup. We then call

G/H the quotient group.

Proof. We have already seen that H normal is necessary for

G/H becoming a group. The condition is also sufficient: Let
H ⊂ G be a normal subgroup. We prove that

aH ⋅ bH = (a ○ b)H

specifies a well-defined operation, that is, given other represen-

tatives
a2 ∈ a1 H b 2 ∈ b1 H
we have to show that

(a2 ○ b2 )H = (a1 ○ b1 )H.

Write
a2 = a1 ○ h b2 = b1 ○ h′
with h, h′ ∈ H. Since H is a normal subgroup, we have

Hb1 = b1 H,

so there exists an h′′ ∈ H with

h ○ b1 = b1 ○ h′′

and hence

(a2 ○ b2 )H = (a1 ○ h ○ b1 ○ h′ )H = (a1 ○ b1 ○ h′′ ○ h′ )H = (a1 ○ b1 )H.

3. GROUPS 82

Note that hH = H for all h ∈ H, since multiplication by h is a

bijektive map H → H.
We check that this operation on G/H indeed defines a group
structure: Associativity

(aH ⋅ bH) ⋅ cH = aH ⋅ (bH ⋅ cH)

follows from (a ○ b) ○ c = a ○ (b ○ c). Moreover,

eH = H

is the neutral element, and

−1
(aH) = a−1 H

the inverse of aH.

Example 3.3.7 Every subgroup of an abelian group is a normal

subgroup. For example, the subgroups nZ ⊂ (Z, +) are normal.
We test this using the definition: For all g ∈ Z we have

g + nZ = {g + nk ∣ k ∈ Z}
= {nk + g ∣ k ∈ Z} = nZ + g.

The quotient group is the residue class group

Z/nZ = Z/n = {0, ..., n − 1} .

The neutral element is 0 = 0 + nZ = nZ and the inverse is −a =

−a = n − a.

Example 3.3.8 An = ker(sign) ⊂ Sn is by Remark 3.3.3 a nor-

mal subgroup. We test this using the definition: For all τ ∈ Sn
and σ ∈ An we have

sign(τ ○ σ ○ τ −1 ) = sign(τ ) sign(σ) sign(τ )−1 = sign(σ) = 1.

Remark 3.3.9 Every subgroup U ⊂ G of index [G ∶ U ] = 2 is a

normal subgroup of G.

The short proof is Exercise 3.16. See also Exercise 3.13.

3. GROUPS 83

3.3.2 Homomorphism theorem

If ϕ ∶ G Ð→ F is a monomorphism, then we can consider G ≅
Im (ϕ) ⊂ F as a subgroup of F . Otherwise, we can make ϕ
injective using the quotient group construction:

Theorem 3.3.10 (Homomorphism theorem) Let ϕ ∶ G Ð→

F be a group homomorphism. Then

G/ Ker ϕ ≅ Im (ϕ) .

Proof. We define

ϕ
̃ ∶ G/ Ker ϕ Ð→ Im ϕ
ϕ
̃ (a Ker ϕ) ∶= ϕ (a)

This is well-defined, since for

a′ = a ○ h ∈ a Ker ϕ with h ∈ Ker ϕ

we have
ϕ (a′ ) = ϕ (a) ⋅ ϕ (h) = ϕ (a) ⋅ e = ϕ (a) .
Since ϕ is a homomorphism, also ϕ ̃ is a homomorphism, it is
surjective on the image of ϕ, and injective, since

ϕ
̃ (a Ker ϕ) = e
⇒ ϕ (a) = e ⇒ a ∈ Ker ϕ
⇒ a Ker ϕ = Ker ϕ = eG/ Ker ϕ .

Hence ϕ ∶ G Ð→ F factorizes in
ϕ
G Ð→ F
projection ↓ ↑ inclusion
G/ Ker ϕ ≅ Im ϕ

Example 3.3.11 Let n ≥ 2. Applied to sign ∶ Sn → ({−1, 1}, ⋅)

with kernel An and im(sign) = {−1, 1} ≅ Z/2 we obtain from
Theorem 3.3.10, that

Sn /An ≅ Z/2.
3. GROUPS 84

Example 3.3.12 The Klein four-group

V4 = {() , (1, 2) (3, 4) , (1, 3) (2, 4) , (1, 4) (2, 3)}

is a normal subgroup of S4 and for the quotient group we have

S4 /V4 ≅ S3 .

We prove this in Exercise 3.18, where we interpret the isomor-

phism geometrically by considering S4 as the symmetry group of
the tetrahedron.
We can also prove S4 /V4 ≅ S3 using GAP:
S4:=SymmetricGroup(4);;
NormalSubgroups(S4);
[ Group(()),
Group([ (1,4)(2,3), (1,3)(2,4) ]),
Group([ (2,4,3), (1,4)(2,3), (1,3)(2,4) ]),
Sym( [ 1 .. 4 ] ) ]
V4:=Group((1,2)(3,4),(1,3)(2,4));;
Elements(V4);
[ (), (1,2)(3,4), (1,3)(2,4), (1,4)(2,3) ]
Q:=S4/V4;;
Order(Q);
6
IsomorphismGroups(Q,CyclicGroup(6));
fail
IsomorphismGroups(Q,SymmetricGroup(3));
[ f1, f2 ] -> [ (2,3), (1,2,3) ]

Example 3.3.13 (Classification of cyclic groups) A cyclic

group G is a group which can be generated by a single element
g ∈ G, that is, G = ⟨g⟩. Then

ϕ ∶ (Z, +) Ð→ ⟨g⟩ = G
k z→ gk

is an epimorphism. The order ord (g) = ∣G∣ can be finite or

infinite. In case ord (g) is infinite, ϕ is an isomorphism, since
only g 0 = e , and every element of G is of the form g k . If ord (g)
 3. GROUPS 85

is finite, then Ker ϕ = ⟨n⟩ = nZ (since every subgroup of Z is of

the form nZ), and the homomorphism theorem gives

Z/nZ ≅ ⟨g⟩
k ↦ gk

So we have shown: Every cyclic group G of finite order is iso-

morphic to Z/nZ with n = ∣G∣, every cyclic group of infinite order
is isomorphic to Z.

3.4 Exercises
Exercise 3.1 Create paper models of the Platonic solids: tetra-
hedron, cube, octahedron, dodecahedron and icosahedron (see Fig-
ure 3.1).

Exercise 3.2 Let G be a set with an operation

○ ∶ G × G Ð→ G
(a, b) ↦ a○b

which satisfies the following axioms:

(G1) Associativity

a ○ (b ○ c) = (a ○ b) ○ c ∀a, b, c ∈ G.

(G2’) There is a left-neutral element, that is, there is an

e∈G

with
e ○ a = a ∀a ∈ G.

(G3’) Existence of the left-inverse, that is, ∀a ∈ G ∃a−1 ∈ G with

a−1 ○ a = e.

Show:

1) For a, b ∈ G we have: If a ○ b = e, then also b ○ a = e.

3. GROUPS 86

2) We have a ○ e = a for all g ∈ G.

3) The neutral element of G is unique.

4) The inverses of the elements of G are unique.

5) For a, b ∈ G we have (a ○ b) = b−1 ○ a−1 .

−1

6) For a ∈ G we have (a−1 )

−1
= a.

Exercise 3.3 Check whether the following definitions give a semi-

group, a monoid, or a group:

1) R ∪ {−∞} with the operation

a♡b = max{a, b},

2) 3 + 6Z = {3 + 6 ⋅ k ∣ k ∈ Z}with addition,

3) Rn with the operation

(a1 , ..., an ) + (b1 , ..., bn ) = (a1 + b1 , ..., an + bn ).

Exercise 3.4 Let ϕ ∶ G1 Ð→ G2 be a group homomorphism.

Show that:

1) Ker (ϕ) ⊂ G1 and Bild (ϕ) ⊂ G2 are subgroups.

2) If ϕ is a group isomorphism, then the inverse map

ϕ−1 ∶ G2 Ð→ G1

is a group isomorphism.

Exercise 3.5 Prove that for n ≥ 2

sign ∶ Sn Ð→ ({1, −1} , ⋅)

n
σ(i)−σ(j)
σ z→ sign (σ) = ∏ i−j
i,j=1
i<j

is a group epimorphism.
3. GROUPS 87

Exercise 3.6 In the following game, we can shift neighboring

squares vertically or horizonally to the empty square. Using such
moves, is it possible to transform the configuration

into the standard configuration:

Exercise 3.7 Write

1 2 3 4 5 6 7 1 2 3 4 5 6 7
σ=( ), τ =( ),
2 1 5 6 4 3 7 1 3 4 2 5 7 6

σ ○ τ and τ ○ σ both as a product of disjoint cycles, as well as,

a product of transpositions. Determine the respective orders and
signatures.

Exercise 3.8 1) Show that, if a, b ∈ Z with a, b ≥ 1 and gcd (a, b) =

1, then
Z/ (a ⋅ b) ≅ Z/a × Z/b.

2) Determine the pre-image of (8 + 10Z, −11 + 21Z) under the

group isomorphism

Z/210 ≅ Z/10 × Z/21.

Exercise 3.9 1) Let G be a group and let x, y ∈ G with x ⋅ y =

y ⋅ x and ⟨x⟩ ∩ ⟨y⟩ = {e}. Show that:

ord (x ⋅ y) = kgV (ord (x) , ord (y)) .

3. GROUPS 88

2) Let
σ = c1 ⋅ ... ⋅ cr ∈ Sn
be the product of disjoint cycles ci with lengths mi . Deter-
mine ord (σ).

Exercise 3.10 Which orders occur among the elements of S6 ?

Exercise 3.11 1) Write every element of S4 as the product

of disjoint cycles.

2) Assign a partition of 4 to every σ ∈ S4 (that is a sum 4 =

p1 + ... + pr with pi ≥ 1). This partition is called the cycle
type of σ.

3) Interpret each cycle type geometrically by interpreting S4

as the symmetry group of the tetrahedron (Figure 3.8).

Figure 3.8: Tetraedron

Exercise 3.12 1) Show that the map

G × G Ð→ G
(a, b) z→ a ○ b ○ a−1

defines a group action, the conjugation, of G on G from

the left.
3. GROUPS 89

Figure 3.9: Regular 5-gon

2) The orbit of b ∈ G
bG ∶= {a ○ b ○ a−1 ∣ a ∈ G}
is called the conjugacy classes of b. Determine all the
conjugacy classes of S3 .
Exercise 3.13 Let G be the symmetry group of the regular pen-
tagon (Figure 3.9). Determine
1) the order of G (prove your claim),
2) all elements of G as permutations of the vertices,
3) all subgroups of G and which thereof are normaldivisors.

Exercise 3.14 Let G = Sym (O) be the symmetry group of the

octahedron O. By numbering the vertices

of O, we can consider G as a subgroup of S6 .

3. GROUPS 90

1) Find the group order of G with the help of the orbit counting
formula.
2) Determine generators of G, and prove your claim using
GAP.
3) Find all conjugacy classes of G using GAP.
4) Interpret the elements of G geometrically.
Hint: Use the GAP commands Group, Order and ConjugacyClasses.
Exercise 3.15 Show that there are exactly 11 isomorphism classes
of (undirected) graphs with 4 vertices.
Exercise 3.16 Let H be a subgroup of G. Show that if [G ∶ H] =
2, then H is a normal subgroup of G.
Exercise 3.17 Let ϕ ∶ G Ð→ F be a group homomorphism.
Prove:
1) If M ⊂ F is a normal subgroup, then ϕ−1 (M ) ⊂ G is a
normal subgroup.
2) If ϕ is surjective and N ⊂ G a normal subgroup, then
ϕ (N ) ⊂ F is a normal subgroup.
3) Give an example of a group homomorphism whose image
is not a normal subgroup.
Exercise 3.18 Prove that the Kleinian four-group
V4 = {() , (1, 2) (3, 4) , (1, 3) (2, 4) , (1, 4) (2, 3)}
is a normal subgroup of S4 , and for the quotient group we have
S4 /V4 ≅ S3 .
Give a geometric interpretation by considering S4 as the symme-
try group of the tetraehedron.
Hint: Every symmetry of the tetrahedron T ⊂ R3 with vertices
e1 = (1, −1, −1) e2 = (−1, 1, −1) e3 = (−1, −1, 1) e4 = (1, 1, 1)
permutes the coordinate axes of R3 , see Figure 3.10. This induces
a group homomorphism
ϕ ∶ S4 → S3 .
3. GROUPS 91

Figure 3.10: Tetrahedron with diagonals connecting mid-points

of the edges

Exercise 3.19 Prove that the Kleinian four-group

V4 = {() , (1, 2) (3, 4) , (1, 3) (2, 4) , (1, 4) (2, 3)}
is a normal subgroup of S4 and for the quotient group, we have
S4 /V4 ≅ S3
Give a geometric interpretation by considering S4 as the symme-
try group of the tetraedron.
Exercise 3.20 Lrt G be the symmetry group of the Ikosaeders.
1) Determine the grou order of G.
2) Find generatofs of the symmetry group G of the ikosaederon
(Figure 3.11) as a subgroup of S12 . Prove you claim using
GAP.
Exercise 3.21 Prove:
1) I
1 ⋯ n−1 n
σ=( ) ∈ Sn ,
σ (1) σ (n − 1) k
then
(n − 1, n) ⋅ ... ⋅ (k, k + 1) ⋅ σ ∈ Sn−1 .
3. GROUPS 92

Figure 3.11: Icosahedron with numbering of the vertices

2) The symmetric group Sn is generated by the transpositions

(1, 2) , (2, 3) , ..., (n − 1, n), that is

Sn = ⟨(1, 2) , (2, 3) , ..., (n − 1, n)⟩ .

Exercise 3.22 Let G ⊂ Sn be a subgroup with (1, 2) ∈ G and

(1, 2, ..., n) ∈ G. Prove that

G = Sn .
4

Rings and fields

4.1 Basics
Definition 4.1.1 A ring (R, +, ⋅) is a set R together with two
operations
+ ∶ R × R Ð→ R, (a, b) z→ a + b
⋅ ∶ R × R Ð→ R, (a, b) z→ a ⋅ b
for which the following axioms are true
(R1) (R, +) is an abelian group,
(R2) multiplication ⋅ is associative,
(R3) the operations are distributive, that is,
a ⋅ (b + c) = a ⋅ b + a ⋅ c
(a + b) ⋅ c = a ⋅ c + b ⋅ c
for all a, b, c ∈ R.
Furthermore if there exists an identity, that is,
(R4) there is an element 1 ∈ R with
a⋅1=1⋅a=a
for all a ∈ R, we say that R is a ring with 1 (as neutral
element of the monoid (R, ⋅) the 1 is unique),
and if

93
4. RINGS AND FIELDS 94

(R5) the multiplication ⋅ is commutative, that is,

a⋅b=b⋅a
for all a, b ∈ R, then R is called a commutative ring.
If ∅ ≠ U ⊂ R with + and ⋅ is a ring, then we call U a subring of
R. If R is a ring with 1, we also require (as a rule) 1 ∈ U .
We write the zero- and identity element as 0R and 1R , in case
they appear in the context of several rings.
Example 4.1.2 1) R = {0} is a ring with 0 = 1, the so-called
zero-ring.
2) Z , Q , R , C are commutative rings with 1.
3) The even numbers 2Z ⊂ Z form a commutative ring with-
out 1.
4) If R1 , R2 are rings, then the cartesian product R1 × R2 is a
ring with componentwise addition
(a1 , a2 ) + (b1 , b2 ) = (a1 + b1 , a2 + b2 )
and multiplication
(a1 , a2 ) ⋅ (b1 , b2 ) = (a1 ⋅ b1 , a2 ⋅ b2 ) .
Definition 4.1.3 Let R and S be rings. a ringhomomor-
phism
ϕ ∶ R Ð→ S
is a map, that satisfies
ϕ (a + b) = ϕ (a) + ϕ (b)
and
ϕ (a ⋅ b) = ϕ (a) ⋅ ϕ (b)
for all a, b ∈ R (in particular, ϕ ∶ (R, +) Ð→ (S, +) is a group
homomorphism). If R and S are rings with 1, we require (as a
rule) that
ϕ (1R ) = 1S .
The concepts mono-, epi-, and isomorphism are defined in a sim-
ilar way as for groups.
4. RINGS AND FIELDS 95

Remark 4.1.4 The image of ϕ (R) ⊂ S, as well as the kernel

Ker ϕ = {r ∈ R ∣ ϕ (r) = 0} ⊂ R,

are subrings.

Remark 4.1.5 For a ring R with 1, only in the special case of

the zero map, Ker ϕ is also a ring with 1, since

1R ∈ Ker ϕ ⇔ ϕ (r) = ϕ (r ⋅ 1R ) = ϕ (r)⋅ϕ (1R ) = 0 ∀r ∈ R ⇔ Ker ϕ = R.

This is because Ker ϕ is an ideal. In the next section we will

come back to ideals.

Definition 4.1.6 Let R be a commutative ring with 1. The

polynomial ring R [x] over R in the variable x is a set of
expressions
f = a0 x0 + a1 x1 + ... + an xn
with n ∈ N0 , ai ∈ R, an ≠ 0.
We call call deg (f ) ∶= n the degree of f and put deg (0) =
−∞.
For i > deg (f ) we put ai = 0.
Addition and multiplication of polynomials are defined as follows.

(a0 x0 + a1 x1 + ... + an xn ) + (b0 x0 + b1 x1 + ... + bm xm )

= (a0 + b0 ) x0 + (a1 + b1 ) x1 + ... + (amax(n,m) + bmax(n,m) ) xmax(n,m)

and

(a0 x0 + a1 x1 + ... + an xn ) ⋅ (b0 x0 + b1 x1 + ... + bm xm )

= c0 x0 + c1 x1 + ... + cn+m xn+m

such that
k
ck = ∑ aj bk−j .
j=0

Polynomial rings in more than one variable are defined induc-

tively
R [x1 , ..., xr ] = R [x1 , ..., xr−1 ] [xr ] .
4. RINGS AND FIELDS 96

4.2 The group of units of Z/n

Definition 4.2.1 Let R be a commutative ring with 1. An ele-
ment u ∈ R is called a unit of R, if there is a w ∈ R with

w ⋅ u = 1.

The set of units is called R× . With u also w is a unit and (R× , ⋅)

is a group, the group of units of R.
If 1 ≠ 0 and
R× = R/{0},
then R is called a field.

Remark: The inverse w = u−1 in R× is unique.

An element a ∈ Z/n is invertible if and only if there is a b ∈ Z
with a ⋅ b = 1, that is, if there are b, k ∈ Z with

a⋅b+n⋅k =1

Such b and k we can obtain using the extended Euclidean algo-

rithm, provided
gcd (a, n) = 1.
On the other hand, if we have such a representation of 1, then a
and n have to be corpime (since otherwise every common divisor
of them also divides 1). We can hence describe the group of units
directly:

Theorem 4.2.2 For n ∈ N we have

×
(Z/n) = {a ∈ Z/n ∣ gcd (a, n) = 1}

The elements are called prime residue classes. The group

(Z/n) we also call the prime residue class group modulo n.
×

As a direct corollary, we obtain:

Corollary 4.2.3 The ring Z/n is a field if and only if n is a

prime number.
4. RINGS AND FIELDS 97

Example 4.2.4 The residue class 8 ∈ Z/15 has an inverse, that

is, 8 ∈ (Z/15) , since
×

gcd (8, 3 ⋅ 5) = 1

With the extended Euclidean algorithm, we obtain a representa-

tion of the greatest common divisor

1 = (2) ⋅ 8 + (−1) ⋅ 15

hence
−1
8 =2

Definition 4.2.5 The Euler ϕ-function ϕ ∶ N → Z defined by

×
ϕ (n) ∶= ∣(Z/n) ∣ = ∣{r ∈ Z ∣ 1 ≤ r ≤ n, gcd (r, n) = 1}∣

gives for n the order of the group of units (Z/n) .

×

Theorem 4.2.6 (Theorem of Fermat-Euler) For all a, n ∈

Z, n ≥ 1 with gcd (a, n) = 1 we have

aϕ(n) ≡ 1 mod n.

Proof. By Corollary 3.2.52, the order of every element g of a

group G divides the group order, hence

g ∣G∣ = e.

Applied to a ∈ (Z/n) we obtain

×

aϕ(n) = 1.

For prime numbers p, we have

ϕ (p) = p − 1,

hence
ap−1 ≡ 1 mod p if p ∤ a
and thus (since for p ∣ a we have ap ≡ 0 ≡ a mod p):
4. RINGS AND FIELDS 98

Corollary 4.2.7 (Fermat’s little theorem) If p is a prime

number and a ∈ Z, then

ap ≡ a mod p.

To compute the Euler ϕ-function, one uses that it is multi-

plicative over coprime products. For this, we first note that the
Chinese remainder theorem group isomorphism is indeed a ring
isomorphism (see Exercise 3.8):

Theorem 4.2.8 Let m1 , m2 ∈ N be copime. Then

Z/m1 m2 ≅ Z/m1 × Z/m2 .

Proof. As shown in Exercise 3.8, by

π∶ Z/m1 m2 Ð→ Z/m1 × Z/m2

a + m1 m2 Z z→ (a + m1 Z, a + m2 Z)

we get a well-defined isomorphism of abelian groups with respect

to +. Moreover

π(ab + m1 m2 Z) = (ab + m1 Z, ab + m2 Z)
= ((a + m1 Z) ⋅ (b + m1 Z), (a + m2 Z) ⋅ (b + m2 Z))
= (a + m1 Z, a + m2 Z) ⋅ (b + m1 Z, b + m2 Z)
= π(a + m1 m2 Z) ⋅ π(b + m1 m2 Z),

hence π is a ring isomorphism.

Example 4.2.9 Using Theorem 2.4.1 we obtain by determining

the solution set of the simultaneous congruences

x ≡ 1 mod 3
x ≡ 7 mod 15 ⇐⇒ {
x ≡ 2 mod 5

According to Theorem 4.2.8 we can rephrase this equivalence as

Z/15 ≅ Z/3 × Z/5 and

7 ↦ (1 , 2)
4. RINGS AND FIELDS 99

Computing the preimage of (1, 2) under this isomorphism is solv-

ing the simultaneous congruence

x ≡ 1 mod 3
x ≡ 2 mod 5

Vice versa, computing the image of 7 is just reduction modulo 3

and 5.
With regard to the multiplication, we have, for example,

π(7) ⋅ π(4) = (1, 2) ⋅ (1, 4) = (1, 8) = (1, 3)

= π(13) = π(7 ⋅ 4).

By Theorem 4.2.8 the ϕ-function is multiplicative:

Corollary 4.2.10 If m1 , m2 ∈ N are corpime, then

ϕ (m1 m2 ) = ϕ (m1 ) ϕ (m2 ) .

Proof. We have a + m1 m2 Z ∈ (Z/m1 m2 ) if and only if

×

(a + m1 Z, a + m2 Z) ∈ (Z/m1 × Z/m2 ) ,
×

equivalently, if a + mi Z ∈ (Z/mi ) for alle i, since multiplication

×

is defined component-wise. Hence

× × ×
(Z/m1 m2 ) = (Z/m1 ) × (Z/m2 )

In particular, we obtain:

Remark 4.2.11 If n = p ⋅ q is the product of two primes, then

ϕ (n) = (p − 1)(q − 1).

4.3 Ideals and quotient rings

In this section, we aim at generalizing the construction of the ring
Z/n. To do so, we investigate, in which sense we can give the
quotient group the structure of a ring. Let R be a commutative
ring with 1. Every subgroup I ⊂ (R, +) is a normal subgroup,
4. RINGS AND FIELDS 100

we hence can construct the quotient group R/I, the surjective

group homomorphism
π∶ (R, +) Ð→ (R/I, +)
r z→ r =r+I

has Ker π = I, and the neutral element of R/I with respect to +

is 0 + I = I.
If we want also a multiplication on R/I, such that π is a ring
homomorphism, then the multiplication has to be induced by
the multiplication in R, since

r1 ⋅ r2 = π (r1 ) ⋅ π (r2 ) = π (r1 r2 ) = r1 r2 .

However, multiplication by multiplication of the representative

may not be well-defined. If r2′ = r2 + b with b ∈ I is a different
representative of r2 + I, then

r1 ⋅ r2′ = r1 ⋅ r2 + r1 ⋅ b,

so we need r1 ⋅ b ∈ I for all r1 ∈ R and b ∈ I. Subgroups of (R, +)

with this property are called ideals:

Definition 4.3.1 Let R be a commutative ring with 1. An ideal

is a non-empty subset I ⊂ R with

a+b∈I
ra ∈ I

for all a, b ∈ I and r ∈ R.

We remark, that with a ∈ I also the additive inverse −a is in I.

All together, we have proven (as an easy exercise the dis-
tributive law in R/I follows directly from that in R):

Theorem 4.3.2 Let I ⊂ R be an ideal. Then the quotient group

R/I carries the structure of a commutative ring with 1 by mul-
tiplication of representatives

(r1 + I) ⋅ (r2 + I) ∶= r1 r2 + I.

The neutral element of R/I with respect to ⋅ is 1+I. We call R/I

the quotient ring of R by I.
4. RINGS AND FIELDS 101

Ideals play an important role in the theory of rings.

Example 4.3.3 1) If I1 , I2 ⊂ R are ideals, then also their
intersection I1 ∩ I2 .
2) Let a1 , ..., an ∈ R. Then
(a1 , ..., an ) ∶= {∑ni=1 ri ai ∣ ri ∈ R}
is an ideal, the ideal generated by the generating system
a1 , ..., an .
3) The ideals of Z are exactly the subgroups
nZ = {na ∣ a ∈ Z} = (n)
with n ∈ Z.
For the ideal I = nZ ⊂ Z, we obtain by Theorem 4.3.2,
that Z/nZ is a commutative ring with 1. The elements are
exactly the residue classes 0̄, 1̄, ..., n − 1, that is,
Z/nZ = Z/n.

4) Let ϕ ∶ R Ð→ S be a ring homomorphism. The kernel

Ker ϕ = {r ∈ R ∣ ϕ (r) = 0} ⊂ R
is an ideal: If r′ ∈ R and ϕ (r) = 0, then also
ϕ (r′ ⋅ r) = ϕ (r′ ) ⋅ ϕ (r) = 0
As in the case of groups we have:
Theorem 4.3.4 (Homomorphism theorem) Let ϕ ∶ R Ð→
S be a ring homomorphism. Then
R/ Ker ϕ ≅ Im ϕ
Proof. From Theorem 3.3.10, we obtain an isomorphism
ϕ
̃ ∶ R/ Ker ϕ Ð→ Im ϕ
r = r + Ker ϕ z→ ϕ (r)
of the additive abelian groups. Moreover, ϕ
̃ is a ring homomor-
phism, since
ϕ
̃ (r1 ⋅ r2 ) = ϕ
̃ (r1 ⋅ r2 ) = ϕ (r1 ⋅ r2 )
= ϕ (r1 ) ⋅ ϕ (r2 ) = ϕ ̃ (r2 ) .
̃ (r1 ) ⋅ ϕ
4. RINGS AND FIELDS 102

4.4 Integral domains and fields

Definition 4.4.1 Let R be a commutative ring with 1.

1) An element a ∈ R is called a zero divisor of R, if there

exists an x ∈ R/{0} with

x ⋅ a = 0.

2) If R does not have any zero divisors except 0, then R is

called an integral domain.

We already have seen, that an element of R cannot be both

a unit and a zero divisor.

Example 4.4.2 1) Z is an integral domain. The units are

+1 and −1, hence

Z× = {+1, −1} .

2) Every field K is an integral domain, for example, Q, R, C.

The units are K × = K/{0}.

3) Z/6 = {0, 1, ...5} is not an integral domain, 2, 3, 4 (and 0)

are zero divisors, 1 and 5 are units.
See also Exercise 4.2.

4) If R is an integral domain, then also R [x] and

×
R [x] = R×

are, since if f ⋅ g = 1, then

0 = deg (1) = deg (f ⋅ g) = deg (f ) + deg (g)

hence deg (f ) = deg (g) = 0.

5) The ring of Gaussian integers

Z [i] = {a + i ⋅ b ∣ a, b ∈ Z} ⊂ C
4. RINGS AND FIELDS 103

is an integral domain and

Z [i] = {1, −1, i, −i} .

×

Clearly these are units:

1⋅1=1
(−1) ⋅ (−1) = 1
i ⋅ (−i) = 1

Prove as an exercise that there are no further units.

Remark 4.4.3 For integral domains R we can use, similarly to

the construction of Q from Z, the calculation with fractions
a
Q(R) = { ∣ a, b ∈ R, b ≠ 0}
b
to form a field, the quotient field, see Exercise 4.6.
For example for a field K we obtain in this way the field of
rational functions.

K(x) = Q(K[x]).

For finite integral domains, there is no need for the quotient field
construction:

Theorem 4.4.4 Every finite integral domain is a field.

We prove this in Exercise 4.5. This observation gives another

proof of Corollary 4.2.3:

Corollary 4.4.5 If p is a prime number, then

Fp = Z/pZ

is a field.

Proof. Follows directly from Theorem 4.4.4: If a ⋅ b = 0 for

a, b ∈ Fp then p ∣ ab, hence p ∣ a or p ∣ b, that is, a = 0 or b = 0.
4. RINGS AND FIELDS 104

Remark 4.4.6 Let K be a field and

χ ∶ Z Ð→ K
n z→ n ⋅ 1K
the characteristic map. The kernel is an ideal
Ker χ = pZ
with p ≥ 0. We call
char (K) = p ≥ 0
the characteristic of K. There are two possible cases:
1) p = 0, that is, χ is injective. In this case, Z and hence also
Q is a subring of K.
2) p > 0. Then
Z/pZ → K
is by the homomorphism theorem 4.3.4 injective, hence
Fp = Z/pZ is a subring of K and thus an integral domain.
Hence p must be a prime number, because if p = a ⋅ b with
a, b > 1, then ā ⋅ b̄ = 0̄, hence ā, b̄ ≠ 0̄ are zero divisors.
So every field contains either Q or Fp .
Remark 4.4.7 One can show, that up to isomorphism, there
is for any prime power pr exactly one field K with ∣K∣ = pr
elements. This field is called Fpr and has char (Fpr ) = p. See
also Exercise 4.4, where we construct a field with 4 elements.
Caveat:
F4 ≠ Z/4
since 2̄ ⋅ 2̄ = 0̄ ∈ Z/4, that is Z/4 is not an integral domain. In
general, Fpr is constructed as an algebaic extension of Fp .

4.5 Exercises
Exercise 4.1 Let R ein ring. Show, that for all x, y ∈ R
0x = x0 = 0
(−x) y = x (−y) = −xy
(−x) (−y) = xy
4. RINGS AND FIELDS 105

Exercise 4.2 Find the multiplication and addition tables of the

ring Z/10Z. Which elements of Z/10Z are units, which are zero
divisors? Find also the multiplication table of the group of units
×
(Z/10Z)

Exercise 4.3 Let K be a field.

1) Show that the set of polynomials K[x] with coefficients

in K and the addition and multiplication from Definition
4.1.6 is an integral domain.
2) Implement the addition and multiplication in K[x].

Exercise 4.4 Show that there is a field with exactly 4 elements

by specifying the addition and multiplication tables.
Hint: Denote the elements of K by 0, 1, a, a + 1.

Exercise 4.5 Show:

1) Every integral domain with finitely many elements is a

field.
2) In a finite ring, every element is either a unit or a zero
divisor.

Exercise 4.6 Let R be an integral domain and S = R/{0}. We

construct the ring von fractions
r
Q (R) = { ∣ r ∈ R, s ∈ S}
s
as Q (R) = (R × S) / ∼ with the equivalence relation
(r, s) ∼ (r′ , s′ ) ⇔ rs′ − sr′ = 0
and write r
s ∶= [(r, s)]. Addition and multiplication are given
r1 r2 r1 s2 + r2 s1
+ =
s1 s2 s1 s2
r1 r2 r1 r2
⋅ =
s1 s2 s1 s2
1) Show: Addition and multiplication are well-defined and
Q(R) is a field.
4. RINGS AND FIELDS 106

2) Implement the arithmetic in Q = Q(Z), that is, addition,

multiplication, inverse, a function to decide equality, and
a function, which finds for every element a representative
which is reduced to lowest terms.

3) Can you modify your implementation such that it also works

for the field of rational functions Q(X) = Q(Q[X])?

Exercise 4.7 Let R be an integral domain. Show:

1) For a, b, c ∈ R, c ≠ 0 it follows from ac = bc, that a = b.

2) For all a ∈ R we have a ∣ 0 and a ∣ a and 1 ∣ a.

3) Let a, b, c ∈ R. If c ∣ b and b ∣ a, then c ∣ a.

4) If a ∈ R and u ∈ R× and a ∣ u, then a ∈ R× .

5) Let a, b, d ∈ R with d ∣ a and d ∣ b. Then d ∣ (xa + yb) for all

x, y ∈ R.

6) Let a, b ∈ R. Then (a) ⊂ (b) ⇐⇒ b ∣ a.

7) If a, b ∈ R, then

a ∣ b and b ∣ a ⇐⇒ ∃u ∈ R× with a = ub ⇐⇒ (a) = (b)

We then say that, a and b are associated.

This is an equivalence relation.

Exercise 4.8 Let F2 be the field with the two elements 0 and 1.
Find all elements of

K = F2 [x] /(x2 + x + 1)

and the addition and multiplication table of K. Show that K is

a field.
5

Vector spaces

5.1 Overview
Linear Algebra is concerned with the description of vector spaces,
the most frequently occuring structure of mathematics. The
reason for this lies in the fact that they arise in the description of
the solution spaces of linear systems of equations. We illustrate
this at an example: If we want to find the set V of all polynomials

f = x1 t3 + x2 t2 + x3 t + x4 ∈ R[t]

of degree ≤ 3 with zeros in t = −1 and t = 2 and an inflection

point in t = 0, we have to find all f with

f (−1) = 0
f ′′ (0) = 0
f (2) = 0.

The coefficients of f hence must satisfy the system of equations

−x1 + x2 − x3 + x4 = 0
2x2 = 0
8x1 + 4x2 + 2x3 + x4 = 0

All these equations are linear (that is of degree 1) in the variables

xi . Such a system, we call a linear system of equations. Since
none of the equations has a constant term, it is a homogeneous
linear system of equations.

107
5. VECTOR SPACES 108

Definition 5.1.1 A polynomial f ∈ K[x1 , ... , xn ] is called ho-

mogeneous, if all terms of f have the same degree.
Example 5.1.2 The polynomials x1 + x2 and x21 x2 + x1 x22 are
homogeneous, x1 + 1 and x21 x2 + x1 x2 are not.
In contrast to general systems of polynomial equations, linear
systems can be solved easily.

5.2 Gauß algorithm

Let K be a field. We can manipulate a linear system as follows:
Remark 5.2.1 If l1 , l2 ∈ K[x1 , ..., xn ] are polynomials and 0 ≠
c ∈ K, then for all x ∈ K n we have
l1 (x) = 0 l1 (x) = 0
} ⇐⇒ {
l2 (x) = 0 l2 (x) + c ⋅ l1 (x) = 0
and
l1 (x) = 0 ⇐⇒ c ⋅ l1 (x) = 0
and
l1 (x) = 0 l2 (x) = 0
} ⇐⇒ {
l2 (x) = 0 l1 (x) = 0
Fixing a total ordering on the set of variables, for example,
x1 > x2 > ... > xn , we can use this observation to solve linear
systems.
Definition 5.2.2 If f = cs xs + ... + cn xn with cs ≠ 0, then
L(f ) = xs
is called the lead variable (or the lead monomial of f ),
LC(f ) = cs
the lead coefficient of f ,
LT(f ) = cs xs
the lead term of f , and
tail(f ) = f − LT(f ) = cs+1 xs+1 + ... + cn xn
the tail of f .
5. VECTOR SPACES 109

Example 5.2.3 For

f = 2x2 + 5x3 + x4

we have

L(f ) = x2
LC(f ) = 2
LT(f ) = 2x2
tail(f ) = 5x3 + x4 .

Theorem 5.2.4 For a homogeneous linear system of equations

given by l1 , ..., lr ∈ K[x1 , ..., xn ], all li ≠ 0, Algorithm 5.1 finds an
equivalent system, such that all equations have pairwise different
lead monomials.

Algorithm 5.1 Gauß algorithm

1
1: for all i do li = LC(l ⋅ li
i)
2: while exist i ≠ j with L(li ) = L(lj ) do
3: lj = lj − li
4: if lj = 0 then
5: delete lj
6: else
1
7: lj = LC(l j)
⋅ lj

Example 5.2.5 In the above example

l1 = −x1 + x2 − x3 + x4 = 0
l2 = 2x2 = 0
l3 = 8x1 + 4x2 + 2x3 + x4 = 0

the algorithm proceeds as follows:

• l1 ∶= −l1 , l2 ∶= 12 l2 and l3 ∶= 18 l3

x1 − x2 + x3 − x4 = 0
x2 = 0
1 1 1
x1 + 2 x2 + 4 x3 + 8 x4 = 0
5. VECTOR SPACES 110

• l3 ∶= l3 − l1

x1 − x2 + x3 − x4 = 0
x2 = 0
3 3 9
2 x2 − 4 x3 + 8 x4 = 0

• l3 ∶= 23 l3

x1 − x2 + x3 − x4 = 0
x2 = 0
1 3
x2 − 2 x3 + 4 x4 = 0

• l3 ∶= l3 − l2

x1 − x2 + x3 − x4 = 0
x2 = 0
− 12 x3 + 3
4 x4 = 0

• l3 ∶= −2l3

x1 − x2 + x3 − x4 = 0
x2 = 0
3
x3 − 2 x4 = 0

Remark 5.2.6 If we sort the li according to increasing lead

monomal L(li ) we obtain the row echelon form of the system.

Example 5.2.7 In Example 5.2.5 the system is already in row

echelon form, while for example

l1 = x1 = 0
l2 = x3 = 0
l3 = x2 = 0

ist not.

Remark 5.2.8 By applying Algorithm 5.2 we can always achieve,

that the variable L(li ) occurs exactly once in li . We then say that
the system is in a reduced row echelon form.
5. VECTOR SPACES 111

Algorithm 5.2 Reduction

1: while exist i ≠ j with L(lj ) in tail(li ) with coeff c do
2: li = li − c ⋅ lj

Example 5.2.9 In Example 5.2.5 we obtain by l1 ∶= l1 + l2

x1 + x3 − x4 = 0
x2 = 0
3
x3 − 2 x4 = 0

and l1 ∶= l1 − l3 the reduced row echelon form

1
x1 + 2 x4 = 0
x2 = 0
3
x3 − 2 x4 = 0

Remark 5.2.10 (Solution set) From the reduced row echelon

form, we can read off the solution set of the linear system directly:
We solve for the lead variables, while the remaining variables can
assume arbitrary values: We first write

V = {x ∈ K n ∣ L(li ) = − tail(li ) for all i} .

From this implicit (that is given by equations) representation

of the solutoin set, we obtain the parametric representation of
the solution set, by replacing in the vector x the variable L(li )
by − tail(li ).

Example 5.2.11 The solution set of Example 5.2.5 is

1 3
V = {x ∈ R4 ∣ x1 = − x4 , x2 = 0, x3 = x4 } .
2 2
The variable x4 can assume arbitrary values, while x1 , x2 , x3
are then determined. The parametric form is then

⎛− 2 x4 ⎞
⎧ 1 ⎫
⎪
⎪
⎪ ⎪
⎪
⎪
⎪
⎪⎜ 0 ⎟ ⎪
⎪
V = ⎨⎜ 3 ⎟ ∣ x4 ∈ R⎬
2 x4
⎪
⎪ ⎜ ⎟ ⎪
⎪
⎪
⎪ ⎝ ⎠ ⎪
⎪
⎪
⎩ x 4 ⎪
⎭
5. VECTOR SPACES 112

Example 5.2.12 Hence, the set of polynomials

f = x1 t3 + x2 t2 + x3 t + x4 ∈ R[t]

in Section 5.1 of degree ≤ 3 with zeros in t = −1 and t = 2 and

inflection point in t = 0 is
1 3
V = {− x4 ⋅ t3 + x4 ⋅ t + x4 ∣ x4 ∈ R} .
2 2
Figure 5.1 shows the function graphs of some f ∈ V (considered
as maps R → R, x ↦ f (x)). In particular, we observe that each
such function indeed automatically has in t = −1 a double zero.

–3 –2 –1 0 1 2 3

–2

–4

Figure 5.1: Cubic polynomials with zeroes −1 and 2 and inflec-

tion point at 0.
 5. VECTOR SPACES 113

We observe that in Example 5.2.11, the elements of V depend

only linearly on x4 , hence, we can write

V = {x4 ⋅ f ∣ x4 ∈ R}

with
1 3
f = − t3 + t + 1.
2 2
To put it differently, with f also all its R-multiples are in V .
This is a more general property of solution sets of linear systems
of equations, and motivates the definition of a vector space.

5.3 Vector spaces and bases

Definition 5.3.1 Let K be a field. A K-vector space is a set
V together with two operations

V ×V Ð→ V (addition)
(v, w) z→ v+w

K ×V Ð→ V (scalar multiplication)
(λ, v) z→ λ⋅v

which obey the following axioms:

(V1) (V, +) is an abelian group,

(V2) Associativity
λ ⋅ (µ ⋅ v) = (λ ⋅ µ) ⋅ v
for all λ, µ ∈ K and v ∈ V ,

(V3) 1 ⋅ v = v for all v ∈ V,

(V4) Distributive laws

(λ + µ) ⋅ v = λ ⋅ v + µ ⋅ v
λ ⋅ (v + w) = λ ⋅ v + λ ⋅ w

for all λ, µ ∈ K and v, w ∈ V .

The elements of a vector space we call vectors.

5. VECTOR SPACES 114

Note that here, we use + both for the addition in K and in

V , and ⋅ both for the multiplication in K and the scalar multi-
plication. Which of the two is meant is clear from the type of
elements connected by the operation.

Example 5.3.2 Let K be a field. Examples of K-vector spaces

are:

1) K n = {(a1 , ..., an ) ∣ ai ∈ K} mit

(a1 , ..., an ) + (b1 , ..., bn ) ∶= (a1 + b1 , ..., an + bn )

λ (a1 , ..., an ) ∶= (λa1 , ..., λan )

where we also write elements of K n as column vectors, that

is, as
⎛ a1 ⎞
⎜ ⋮ ⎟ ∈ K n,
⎝ an ⎠
The neutral element of K n is the zero vector

⎛ 0 ⎞
0=⎜ ⋮ ⎟
⎝ 0 ⎠

2) the polynomial ring K[x],

3) the set of all polynomials up to degree d, denoted K[x]≤d .

Sums and multiples of homogeneous linear systems of equa-

tions are again solutions. We hence define:

Definition 5.3.3 Let V be a K-vector space. A non-empty sub-

set U ⊂ V is called sub (vector) space, if

u1 , u2 ∈ U Ô⇒ u1 + u2 ∈ U
λ ∈ K, u ∈ U Ô⇒ λ ⋅ u ∈ U .

Remark 5.3.4 1) Together with the addition and scalar mul-

tiplication, U is again a K-vector space .

2) Every sub vector space U contains 0 ∈ V (since there is an

element u ∈ U and 0 = 0 ⋅ u ∈ U ).
5. VECTOR SPACES 115

Theorem 5.3.5 The solution space of a linear system of equa-

tions for x1 , ..., xn over the field K is a sub vector space of K n .
Proof. Consider the system
a1,1 x1 + ... + a1,n xn = 0
⋮
ar,1 x1 + ... + ar,n xn = 0
over the field K. If x, y ∈ K n are solutions, then also x + y and
λ ⋅ x for all λ ∈ K:
If ∑nj=1 ai,j xj = 0 and ∑nj=1 ai,j yj = 0 for all i = 1, ..., r, then
n n n
∑j=1 ai,j (xj + yj ) = ∑j=1 ai,j xj + ∑j=1 ai,j yj = 0
and
n n
∑j=1 ai,j (λ ⋅ xj ) = λ ⋅ ∑j=1 ai,j xj = 0.

Example 5.3.6 1) Sub vector spaces of R3 are {0}, the lines

through 0, the planes through 0 (exercise) and R3 . We will
see later that these are the only sub vector spaces of R3 .
2) K [x]≤d = {f ∈ K [x] ∣ deg f ≤ d} ⊂ K [x] is a sub vector
space.
3) The sets
U1 = {(x, y) ∈ R2 ∣ y ≥ a}
with a ∈ R (see Figure 5.2 for a = 0) and
U2 = {(x, y) ∈ R2 ∣ y = x2 }
(see Figure 5.3) are no sub vector spaces of R2 . Why?
Definition and Theorem 5.3.7 Let V be a K-vector sub space
and v1 , ..., vn ∈ V . A vector v ∈ V is a linear combination of
v1 , ..., vn , when there are λi ∈ K with
v = λ1 v1 + ... + λn vn .
The set of all linear combinations
⟨v1 , ..., vn ⟩ ∶= {λ1 v1 + ... + λn vn ∣ λi ∈ K} ⊂ V
is a sub vector space, called the span of v 1 , ..., v n or the sub
vector space spanned by v 1 , ..., v n .
5. VECTOR SPACES 116

-v

Figure 5.2: Half plane

Proof. If v, w ∈ ⟨v1 , ..., vn ⟩, that is, v = ∑ni=1 λi vi and w = ∑ni=1 µi vi

with λi , µi ∈ K, then
n
v + w = ∑ (λi + µi ) vi ∈ ⟨v1 , ..., vn ⟩
i=1

and n
λv = ∑ (λ ⋅ λi ) vi ∈ ⟨v1 , ..., vn ⟩ .
i=1

Example 5.3.8 1) The vectors

⎛ 1 ⎞ ⎛ 0 ⎞
v1 = ⎜ 0 ⎟ , v2 = ⎜ 1 ⎟ ∈ R3
⎝ 0 ⎠ ⎝ 0 ⎠

span the plane E = {z = 0}, since ever vector in the plane

can be written as

⎛ λ1 ⎞
E ∋ ⎜ λ2 ⎟ =λ1 v1 + λ2 v2 .
⎝ 0 ⎠
5. VECTOR SPACES 117

2v

Figure 5.3: Parabola

The vectors
⎛ 1 ⎞ ⎛ 1 ⎞
w1 = ⎜ 1 ⎟ , w2 = ⎜ −1 ⎟ ∈ R3
⎝ 0 ⎠ ⎝ 0 ⎠

also span the plane, so

E = ⟨v1 , v2 ⟩ = ⟨w1 , w2 ⟩

2) The monomials 1, x, ..., xd ∈ K[x] span K[x]≤d .

Definition 5.3.9 Let V be a K-vector space.

1) Vectors v1 , ..., vn ∈ V are called a generating system of

V , if
V = ⟨v1 , ..., vn ⟩ .

2) Vectors v1 , ..., vn ∈ V are called linearly independent, if

λ1 v1 + ... + λn vn = 0

implies that
λ1 = ... = λn = 0,
otherwise linearly dependent.
5. VECTOR SPACES 118

3) A generating system v1 , ..., vn of V consisting of linearly

independent vectors is called a basis of V .

Algorithm 5.3.10 Vectors v1 , ..., vn ∈ K m are linearly indepen-

dent if and only if the homogeneous linear system of equations

x1 v1 + ... + xn vn = 0

has only the solution

⎛ x1 ⎞ ⎛ 0 ⎞
⎜ ⋮ ⎟=⎜ ⋮ ⎟
⎝ xn ⎠ ⎝ 0 ⎠

This can be decided via the Gauß algorithm.

Algorithm 5.3.11 A basis of the solution space of a homoge-

neous linear system of equations is obtained from the parametric
representation given in Remark 5.2.10 by substituting a unit basis
for the free variables.

Example 5.3.12 The system

l1 = x1 + 2x2 − 2x5 = 0
l2 = x3 + x5 = 0
l3 = x4 + 2x5 = 0

in Q[x] has already reduced row echelon form, so the solution set
in its parametric representation is
⎧
⎪
⎪
⎪ ⎛ −2x2 + 2x5 ⎞ ⎫
⎪
⎪
⎪
⎪
⎪ x2 ⎪
⎪
⎪⎜
⎪ ⎜
⎟
⎟ ⎪
⎪
V = ⎨⎜ −x5 ⎟ ∣ x2 , x 5 ∈ Q ⎬
⎪ ⎜ ⎟ ⎪
⎪
⎪
⎪ ⎜ −2x5 ⎟ ⎪
⎪
⎪
⎪
⎪ ⎝ ⎠ ⎪
⎪
⎪
⎩ x5 ⎪
⎭
and we hence obtain a basis by considering (x2 , x5 ) = (1, 0) and
(x2 , x5 ) = (0, 1):
⎛ −2 ⎞ ⎛ −2 ⎞
⎜ 1 ⎟ ⎜ 0 ⎟
⎜ ⎟ ⎜ −1
⎟
V = ⟨⎜ 0 ⎟,⎜ ⎟⟩
⎜ ⎟ ⎜ ⎟
⎜ 0 ⎟ ⎜ −2 ⎟
⎝ 0 ⎠ ⎝ 1 ⎠
5. VECTOR SPACES 119

5.4 Dimension
In this section we will see that any two (finite) bases of a vec-
tor space have the same dimension. This number is called the
dimension of the vector space, which is the key classifying in-
variant: Every n-dimensional K-vector space is isomorphic to
K n.

Theorem 5.4.1 Let V be a K-vector space and Ω = (v1 , ..., vn )

a list of vectors in V . Then it is equivalent:

1) Ω is a basis of V .

2) Ω is generating system of V , which cannot be shortened.

3) Ω is system of linearly independent vectors in V , which

cannot be extended.

4) Every vector in V can be written in a unique way as a

linear combination of the elements of Ω.

Remark 5.4.2 Let Ω be a finite basis of V . The linear com-

bination map

lcΩ ∶ Kn Ð→ V
⎛ a1 ⎞
⎜ ⋮ ⎟ z→ a1 v1 + .... + an vn
⎝ an ⎠

is bijective. Its inverse

n
coΩ = lc−1
Ω ∶ V Ð→ K

is called the coordinate representation with resepect to Ω.

Example 5.4.3 If we choose for the vector space V = K[x]≤2

the basis Ω = (1, x, x2 ), we get the bijection

lcΩ ∶ K3 Ð→ K[x]≤2
⎛ a0 ⎞
⎜ a1 ⎟ z→ a0 + a1 x + a2 x 2
⎝ a2 ⎠
5. VECTOR SPACES 120

The coordinate representation of 3x2 + x is

⎛ 0 ⎞
2
coΩ (3x + x) = ⎜ 1 ⎟
⎝ 3 ⎠

Definition 5.4.4 A vector space is called finite dimenional,

if it has a finite generating system.

Corollary 5.4.5 By Theorem 5.4.1, from any finite generating

system of a vector space, we can choose a basis. Hence any finite
dimenisional vector space has a basis.

We mention without proof:

Definition and Theorem 5.4.6 (Fundamental theorem) Let

V a finite dimensional K-vector space. Then any two basis of V
have the same number of elements.
This number we call the dimension dimK V of V over K.
If V is not finite-dimensional, then we set dimK V = ∞.

Example 5.4.7 Using Theorem 5.4.6 and the specified basis, we

have

1) The unit vectors e1 , ... , en are a basis of K n , hence dim K n =

n,

2) The monomials 1, ... , xd are a basis of K[x]≤d , hence dim K[x]≤d =

d + 1,

3) dim K[x] = ∞, since any finite set of polynomials only in-

volves polynomials of finite degree.

4) R is a Q-vector space of infinite dimension (if R would have

finite dimension n over Q, then there would be a bijective
map Qn → R. So with Q also R would be countable , a
contradiction). Hence dimQ R = ∞ (but dimR R = 1 with
the basis e1 = 1).
5. VECTOR SPACES 121

5.5 Vector space homomorphisms

The maps lcΩ and coΩ respect the vector space structures on
K n and V , that is, it does not play a role whether we first do a
calculation with elements of K n and then apply lcΩ or we first
apply lcΩ and then do the respective calculation. Indeed we have

⎛⎛ a1 ⎞ ⎛ b1 ⎞⎞ ⎛ a1 + b1 ⎞ n
lcΩ ⎜⎜ ⋮ ⎟ + ⎜ ⋮ ⎟⎟ = lcΩ ⎜ ⋮ ⎟ = ∑ (ai + bi ) vi
⎝⎝ an ⎠ ⎝ bn ⎠⎠ ⎝ an + bn ⎠ i=1
n n ⎛ a1 ⎞ ⎛ b1 ⎞
= ∑ ai vi + ∑ bi vi = lcΩ ⎜ ⋮ ⎟ + lcΩ ⎜ ⋮ ⎟
i=1 i=1 ⎝ an ⎠ ⎝ bn ⎠
and
⎛ ⎛ a1 ⎞⎞ ⎛ λa1 ⎞ n
lcΩ ⎜λ ⋅ ⎜ ⋮ ⎟⎟ = lcΩ ⎜ ⋮ ⎟ = ∑ (λai ) vi
⎝ ⎝ an ⎠⎠ ⎝ λan ⎠ i=1
n ⎛ a1 ⎞
= λ ∑ ai vi = λ ⋅ lcΩ ⎜ ⋮ ⎟.
i=1 ⎝ an ⎠
This means that lcΩ is a Homomorphism of vector spaces:
Definition 5.5.1 A K-vector space homomorphism is a K-
linear map F ∶ V → W of K-vector spaces, that is
F (v1 + v2 ) = F (v1 ) + F (v2 )
for all vi ∈ V and
F (λv) = λF (v)
for all v ∈ V and λ ∈ K.
The terms Mono-, Epi- and Isomorphism are used analogsly
to the cases of groups and rings.
Example 5.5.2
lc(1,x,...,xd ) ∶ K d+1 Ð→ K[x]≤d
⎛ a0 ⎞
⎜ ⋮ ⎟ z→ a0 + a1 x + ... + ad xd
⎝ ad ⎠

is a K-vector space isomorphism.

5. VECTOR SPACES 122

For the classification of finite dimensional vector spaces al-

ready the dimension is sufficient:

Theorem 5.5.3 (Classification theorem for vector spaces)

Let V be a K-vector space of dimension n < ∞. Then V is iso-
morphic to K n . We write

V ≅ K n.

Proof. By Remark 5.4.5 and Definition and Theorem 5.4.6 we

observe that V has a basis Ω = (v1 , ..., vn ) with n elements, and
by what we have said above

lcΩ ∶ K n → V

is an isomorphism.

Definition and Theorem 5.5.4 A n × m-matrix A over K is

a table
⎛ a1,1 ⋯ a1,m ⎞
A=⎜ ⋮ ⋮ ⎟ = (ai,j ) i=1,...,n
⎝ an,1 ⋯ an,m ⎠ j=1,...,m

The set of n × m-matrices with entry-wise addition and scalar

multiplication is a K-vector space, which we denote by K n×m .
By matrix multiplication

⎛ a11 ⋯ a1,m ⎞ ⎛ x1 ⎞ ⎛ (∑j=1 a1,j xj ) ⎞

m

⎜ ⋮ ⋮ ⎟ ⋅ ⎜ ⋮ ⎟ ∶= ⎜ ⋮ ⎟
⎝ an,1 ⋯ an,m ⎠ ⎝ xm ⎠ ⎝ (∑mj=1 a n,j x j ) ⎠

a vector space homomorphism

K m → K n, x ↦ A ⋅ x

is given, which we denote again by A. The image of x is just the

xj -linear combination of the columns Ai ∈ K n of A = (A1 ∣ ... ∣
Am ), that is,

⎛ x1 ⎞ m
(A1 ∣ ... ∣ Am ) ⋅ ⎜ ⋮ ⎟ = ∑ xj ⋅ Aj .
⎝ xm ⎠ j=1
5. VECTOR SPACES 123

Proof. The map A = lc(A1 ,...,Am ) , which is a homomorphism, as

remarked above.
Example 5.5.5 We have

1 2 3 ⎛ 1 ⎞ 1⋅1+2⋅2+3⋅3 14
( )⋅⎜ 2 ⎟=( )=( )
4 5 6 ⎝ 3 ⎠ 4⋅1+5⋅2+6⋅3 32

using the formula for matrix multiplication. On the other hand,

using the interpretation of the map as a linear combination map
we get:

1 2 3 ⎛ 1 ⎞ 1 2 3 14
( )⋅⎜ 2 ⎟ = 1⋅( )+2⋅( )+3⋅( )=( )
4 5 6 ⎝ 3 ⎠ 4 5 6 32

Example 5.5.6 The derivative

d
dx ∶ R[x] Ð→ R[x]
is a R-vector space homomorphism, since
d
(∑d ai xi ) = ∑di=1 iai xi−1
dx i=0
and hence the image depends linearly on the coefficients ai of the
input polynomial (check!). It is not a monomorphism since
d d
0= 1,
dx dx
however it is an epimorhpism, since
d ai i+1
(∑di=0 x ) = ∑di=0 ai xi ,
dx i+1
that is, every polynomial has a antiderivative.
Definition 5.5.7 Let F ∶ V Ð→ W be a K-vector space homo-
morphism. For bases Ω = (v1 , ..., vm ) of V and ∆ = (w1 , ..., wn )
of W we define a K-vector space homomorphism
M∆Ω (F ) ∶ K m Ð→ K n
by
M∆Ω (F ) ∶= co∆ ○F ○ lcΩ .
5. VECTOR SPACES 124

So we have a diagram
F
V Ð→ W
lcΩ ↑ ↑ lc∆
Km Ð→
Ω
Kn
M∆ (F )

that is
F = lc∆ ○M∆Ω (F ) ○ coΩ
We already know how to use the isomorphisms lcΩ and coΩ .
The key observation is that M∆Ω (F ) can be realized by matrix
multiplication (which, for example, can be easily implemented
in a computer):

Theorem 5.5.8 Let F ∶ K m → K n be a homomorphism and

A = (ai,j ) ∈ K n×m with
n
F (ej ) = ∑ai,j ei
i=1

that is in the columns of

A = (F (e1 ) ∣ ... ∣ F (em ))

are the images of the unit basis vectors. Then

F (c) = A ⋅ c.

Proof. For
⎛ c1 ⎞
c = ⎜ ⋮ ⎟ ∈ Km
⎝ cm ⎠
we have
m m
F (c) = F (∑cj ej ) = ∑cj F (ej )
j=1 j=1

= A ⋅ c.

Every homomorphism F ∶ K m → K n is hence given by multi-

plication by an n × m-matrix A.
5. VECTOR SPACES 125

Definition 5.5.9 Using the above notation, we call M∆Ω (F ) ∈

K n×m the represenating matrix of F with regard to the bases
Ω of V and ∆ of W .

The representing matrix can now easily be computed:

Remark 5.5.10 The i-th column of M∆Ω (F ) contains the coef-

ficients of the representation of F (vi ) with respect to the basis
∆.

Proof. The i-the column of M∆Ω (F ) is

M∆Ω (F )(ei ) = (co∆ ○F ○ lcΩ )(ei ) = (co∆ ○F )(vi ) = co∆ (F (vi )),

so
M∆Ω (F ) = (co∆ (F (v1 )) ∣ ... ∣ co∆ (F (vm ))) ∈ K n×m

Example 5.5.11 We consider the derivative

d
dx ∶ R[x]≤3 Ð→ R[x]≤2

and and the bases Ω = (1, x, x2 , x3 ) and ∆ = (1, x, x2 ). Then

d
(xs ) = s ⋅ xs−1
dx
hence
d ⎛ 0 1 0 0 ⎞
M∆Ω ( ) = ⎜ 0 0 2 0 ⎟.
dx ⎝ 0 0 0 3 ⎠
Using this, we can now compute, for example,
d d
(x3 − 5x2 + 7x − 11) = lc∆ (M∆Ω ( ) ⋅ coΩ (x3 − 5x2 + 7x − 11))
dx dx
⎛ 0 1 0 0 ⎛ −11 ⎞⎞
⎜ ⎛ ⎞ ⎜ 7 ⎟⎟
= lc∆ ⎜⎜ 0 0 2 0 ⎟ ⋅ ⎜ ⎟⎟
⎜⎝ ⎜ −5 ⎟⎟
⎝ 0 0 0 3
⎠
⎝ 1 ⎠⎠
⎛ 7 ⎞
= lc∆ ⎜ −10 ⎟ = 3x2 − 10x + 7
⎝ 3 ⎠
5. VECTOR SPACES 126

The practical implementation of a K-vector space homomor-

phism F ∶ V → W in a computer can we hence do as follows:
After using co∆ the convert the input vector in V to a vector
in K m , the actual computation is realized as matrix multipli-
cation with the representing matrix M∆Ω (F ), and the output is
then interpreted as a vector in W using lcΩ . Here M∆Ω (F ) can
be precomputed and can then be reused for any input vector.

5.6 Exercises
Exercise 5.1 Let V be a K -vector space and U ⊂ V a vector
sub space. Show:
1) (−1) ⋅ v = −v for all v ∈ V .
2) U is with the addition and skalar multiplication induced by
those of V a K-vector space.

Exercise 5.2 Determine the solutions space V ⊂ Q5 and a basis

thereof for each of the following linear systems of equations:
1)
x1 + 2x2 + 2x3 − 2x4 − x5 = 0
−2x1 − 3x2 − x3 + 8x4 + x5 = 0
x1 + 4x2 + 8x3 + 8x4 − 4x5 = 0
2x1 + 5x2 + 7x3 + 2x4 − 4x5 = 0

2)
x1 + x2 + x3 + x4 − x5 = 0
x1 + 2x2 + 3x3 + 4x4 − 5x5 = 0
x1 + 4x2 + 9x3 + 16x4 − 25x5 = 0
x1 + 8x2 + 27x3 + 64x4 − 125x5 = 0

Exercise 5.3 Determine for all t ∈ Q a basis of the solution

space Vt ⊂ Q3 of the homogeneous linear system of equations
−x1 + x2 − 2x3 = 0
x1 + (t − 1) ⋅ x2 + 2x3 = 0
2x1 + (t − 2) ⋅ x2 2
+ (t − t + 4) ⋅ x3 = 0
5. VECTOR SPACES 127

Exercise 5.4 Let

l1 = a1,1 x1 + ... + a1,n xn = 0

⋮
lr = ar,1 x1 + ... + ar,n xn = 0

with ai,j ∈ Q a homogeneous linear system of equations. Write a

function which

1) transforms the system in row echelon form.

2) transforms the system in reduced row echelon form.

3) finds a basis of the solution space.

Exercise 5.5 Let d ≥ 2 and

R [x]≤d = {f ∈ R [x] ∣ deg f ≤ d}

the vector space of polynomial of degree ≤ d.

1) Determine, whether the following sets are sub vector spaces

of R [x]≤d :

U1 = {f ∈ R [x]≤d ∣ f (0) = 0}
U2 = {f ∈ R [x]≤d ∣ f (0) = 1}
U3 = {f ∈ R [x]≤d ∣ f (1) = 0}
U4 = {f ∈ R [x]≤d ∣ f ′ (0) + f ′′ (0) = 0}
U5 = {f ∈ R [x]≤d ∣ f ′ (0) ⋅ f ′′ (0) = 0}

2) For each Ui which is a sub vector space, find a basis.

Exercise 5.6 Do the following vectors

⎛ 1 ⎞ ⎛ 0 ⎞ ⎛ 0 ⎞ ⎛ 0 ⎞
⎜ 0 ⎟ ⎜ 1 ⎟ ⎜ 0 ⎟ ⎜ 0 ⎟
⎜ ⎟,⎜ ⎟,⎜ ⎟,⎜ ⎟ ∈ R4
⎜ −2 ⎟ ⎜ 0 ⎟ ⎜ 1 ⎟ ⎜ 1 ⎟
⎝ 0 ⎠ ⎝ −3 ⎠ ⎝ −1 ⎠ ⎝ 1 ⎠

form a basis of R4 ? Proof you claim.

5. VECTOR SPACES 128

Exercise 5.7 Show: For every b ∈ R the following d + 1 polyno-

mials
2 d
1, (x − b) , (x − b) , ..., (x − b) ∈ R [x]≤d
form a basis of R [x]≤d .

Exercise 5.8 Let p be a prime number and Fp = Z/p the finite

field with p element.

1) Show: Ever d-dimensional Fp -vector space V has exactly

pd elements.
3
2) Let V = (F2 ) and

⎛ 1 ⎞ ⎛ 1 ⎞
v1 = ⎜ 1 ⎟ v2 = ⎜ 1 ⎟
⎝ 1 ⎠ ⎝ 0 ⎠

Find all elements of the sub vector space ⟨v1 , v2 ⟩ ⊂ V and

all vectors v3 ∈ V , such that v1 , v2 , v3 form a basis of V .
d
3) How many different bases of (Fp ) are there? Give a for-
mula.

Exercise 5.9 Determine which subsets of

{x3 + x, x2 , x3 , x2 + 1, x, 1}
form a basis of R [x]≤3 .

Exercise 5.10 Let K be a field, and let U, V ⊂ K n be sub vector

spaces given by bases u1 , ..., us of U and v1 , ..., vt of V .

1) Show that U ∩ V ⊂ K n is a sub vector space.

2) Describe an algorithm to find a basis of U ∩ V .
3) Apply your method to the sub vector spaces

⎛ 4 ⎞ ⎛ 2 ⎞ ⎛ 2 ⎞ ⎛ 4 ⎞ ⎛ 2 ⎞
⎜ 0 ⎟ ⎜ 1 ⎟ ⎜ 0 ⎟ ⎜ 3 ⎟ ⎜ 2 ⎟
U = ⟨⎜ ⎟, ⎜ ⎟⟩ V = ⟨⎜ ⎟, ⎜ ⎟, ⎜ ⎟⟩
⎜ 2 ⎟ ⎜ 1 ⎟ ⎜ 3 ⎟ ⎜ 2 ⎟ ⎜ 3 ⎟
⎝ 4 ⎠ ⎝ 1 ⎠ ⎝ 1 ⎠ ⎝ 3 ⎠ ⎝ 1 ⎠

of Q4 .
Index

abelian, 49 cycle, 68
action, 61 cyclic, 60
alternating group, 59
anti-symmetric, 18 darstellende Matrix, 125
antiderivative, 123 degree, 95
associated, 106 dimension, 120
associative, 17 divides, 31
associativity, 49 division with remainder, 30

base case, 8 empty set, 4

basis, 118 epimorphism, 57
bijective, 12 equivalence relation, 20
Euclidean motions, 63
canonical map, 20 Euklid’s first theorem, 34
Cantor, Georg, 4 Euklid’s second theorem, 34
Cardano, Geronimo, 3 Euler Phi-function, 97
cardinality, 7 even numbers, 94
cartesian product of groups, 52 exponenetial function, 58
cartesian product of sets, 7 extended Euclidean algorithm,
characteristic, 104 36
commutative, 49, 94
commutative ring, 94 Fermat, Pierre de, 1
complement, 5 Fermats letzter Satz, 1
complete set of representatives, Ferrari, Lodovico, 3
67 field, 96
congruent, 32 fundamental theorem on vector
conjugacy classes, 89 spaces, 120
conjugation, 88 Galois, Evariste, 3
coordinate representation, 119 Gauß algorithm, 109
coprime, 32 generating system, 101, 117
coset, 76 generator, 60
countable, 120 graph of the map, 12

129
INDEX 130

greatest common divisor, 35 linearly dependent, 117

group, 49 linearly independent, 117
group homomorphism, 56
group of motions, 63 map, 11
group of residue classes, 54 Matrix, 122
group of self mappings, 73 Matrixmultiplikation, 122
group of self-mappings, 51, 62 monoid, 49
group of units, 96 monomorphism, 57
group table, 55, 74 motion, 63

homogeneous, 108 natural numbers, 5

neutral element, 49
ideal, 100 normal subgroup, 80
identity, 93 number of elements, 7
identiy map, 17
if and only if, 5 operation, 49
image, 12, 56 orbit, 66
implicit form, 111 order, 49
index, 77 order of a group element, 61
index formel, 77 parametric, 111
induction hypothesis, 9 partial ordering, 18
induction step, 8 partitions, 20
injective, 12 Peano axioms, 27
integers, 5 permutation, 51
integral domain, 102 polynomial ring, 95
inverse map, 15 power set, 8
inverses, 49 preimage, 12
isomorphism, 57 prime factors, 33
kernel, 56 prime number, 32
Klein four-group, 84 prime number theorem, 34
prime residue class group, 96
lead coefficient, 108 prime residue classes, 96
lead monomial, 108
lead term, 108 quotient, 67
lead variable, 108 quotient group, 81
least common multiple, 35 quotient map, 67
linear combination, 115 quotient ring, 100
linear combination map, 119 rational functions, 103
linear system of equations, 107 rational numbers, 5
INDEX 131

reduced row echelon form, 110 unit, 96

reflexive, 18
relation, 11 vector, 113
representative, 20 vector space, 113
representative of an orbit, 67 Vektorraumhomomorphismus, 121
residue class, 32 Wiles, Andrew, 2
residue class group, 54 without loss of generality, 9
ring, 93 word, 51
ring with 1, 93
ringhomomorphism, 94 zero divisor, 102
row echelon form, 110 zero-ring, 94
semigroup, 49
set, 4
sieve of Eratosthenes, 42
signatur, 59
signum, 59
simultaneous congruences, 38
source, 12
span, 115
stabilizer, 66
sub vector space, 114
subgroup, 52
subring, 94
subset, 5
surjective, 12
symmetric, 19
symmetric group, 51
symmetry group, 63

tail, 108
target, 12
Tartaglia, Nicolo, 3
theorem of Fermat-Euler, 97
total ordering, 18
transitive, 18
transposition, 51, 68
trial division, 41

union, 5
Bibliography

[1] The Axiom Group: Axiom, http://www.axiom-

developer.org/ (2012).

[2] M. Artin: Algebra, Birkhäuser (2003).

[3] J. Böhm: Grundlagen der Algebra und Zahlentheorie,

Springer (2016).

[4] J. Böhm: Mathematik für Informatiker: Algebraische

Strukturen, Lecture Notes (2018).

[5] J. Böhm: Mathematik für Informatiker: Kombinatorik und

Analysis, Lecture Notes (2018).

[6] J. Böhm: Mathematik für Informatiker: Analysis, Lecture

Notes (2019).

[7] J. Böhm: Mathematik für Informatiker: Kombinatorik,

Stochastik und Statistik, Lecture Notes (2019).

[8] S. Bosch: Algebra, Springer (1993).

[9] P. Bundschuh: Einführung in die Zahlentheorie, Springer

(1998).

[10] Decker, W.; Greuel, G.-M.; Pfister, G.; Schönemann, H.:

Singular 4-1-1 — A computer algebra system for polyno-
mial computations. http://www.singular.uni-kl.de (2018).

[11] G. Fischer: Lineare Algebra, Vieweg (2010).

[12] G. Fischer, R. Sacher: Einführung in die Algebra, Teubner

(1983).

132
BIBLIOGRAPHY 133

[13] The GAP Group, GAP – Groups, Algorithms, and Program-

ming, Version 4.8.10 ; http://www.gap-system.org, (2018).

[14] Grayson, D. R.; Stillman, M. E.: Macaulay2, a soft-

ware system for research in algebraic geometry, available
at http://www.math.uiuc.edu/Macaulay2/ (2019).

[15] G.H. Hardy, E.M. Wright: An introduction to the theory of

numbers, Oxford (1956).

[16] J. C. Jantzen, J. Schwermer: Algebra, Springer (2006).

[17] C. Karpfinger, K. Meyberg: Algebra, Spektrum Akademis-

cher Verlag (2008).

[18] E. Kunz: Algebra, Vieweg (1994).

[19] B. Kreußler, G. Pfister: Mathematik für Informatiker: Al-

gebra, Analysis, Diskrete Strukturen, Springer (2009).

[20] Bosma, W.; Cannon J.; Playoust C.: The Magma alge-
bra system. I. The user language, J. Symbolic Comput., 24
(1997), 235–265.

[21] Maple (Waterloo Maple Inc.): Maple 2018,

http://www.maplesoft.com/ (2018).

[22] Maxima: Maxima, a Computer Algebra System. Version

5.25.1, available at http://maxima.sourceforge.net/ (2011).

[23] Wolfram Research, Inc.: Mathematica Edition: Version 11

(2018).

[24] MATLAB. Natick, Massachusetts: The MathWorks Inc.,

http://www.mathworks.de/products/matlab/ (2018).

[25] Hearn, A. C.: REDUCE 3.8, available at http://reduce-

algebra.com/ (2009).

[26] R. Remmert, P. Ullrich: Elementare Zahlentheorie,

Birkhäuser (1987).

[27] P. Ribenboim: Die Welt der Primzahlen, Springer (2006).

BIBLIOGRAPHY 134

[28] R. Schulze-Pillot: Einführung in die Algebra und Zahlen-

theorie, Springer (2008).

[29] V. Shoup: A Computational Introduction to Number The-

ory and Algebra, Cambridge University Press (2005).

[30] W. Willems: Codierungstheorie und Kryptographie,

Birkhäuser (2008).

[31] J. Wolfart: Einführung in die Algebra und Zahlentheorie,

Vieweg (1996).

[32] G. Wüstholz: Algebra, Vieweg (2004).