Alireza Ghaffari-Hadigheh

Farzali Izadi

A First course in Mathematical

English for math students,
First Edition
July 21, 2020

Azarbaijan Shahid Madani University

Contents

1 Selected Topics in Calculus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Derivative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Definite and Indefinite Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2 Selected Topics in Linear Algebra . . . . . . . . . . . . . . . . . . . . . . . . 9

2.1 Vector Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.1 Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.2 Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2 Eigenvalues and Eigenvectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3 Selected Topics in Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.1 Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2 Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2.1 Mean . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.2 Arithmetic Mean . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.3 Geometric Mean . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.4 Harmonic Mean . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.5 Median . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.6 Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.7 Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.8 Standard Deviation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.9 Coefficient of Variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.10 Variance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4 Selected Topics in Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

4.1 Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2 Problems of Probability Theory . . . . . . . . . . . . . . . . . . . . . . . . . . 30
VI Contents

4.2.1 Independence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2.2 Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.3 Conditional Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3 The Poisson Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3.1 The Poisson Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

5 Selected Topics in Algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

5.1 Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.2 Rings and Homomorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.3 Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

6 Selected Topics in Numerical Analysis . . . . . . . . . . . . . . . . . . . 45

6.1 rounding Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
6.2 Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
6.3 Solving Non-Linear Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
6.4 Numerical Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
6.5 Numerical Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
6.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

7 Selected Topics in Differential Equation . . . . . . . . . . . . . . . . . . 55

7.1 Differential Equation in a Nutshell . . . . . . . . . . . . . . . . . . . . . . . . . 55
7.2 The General Solution of a Differential Equation . . . . . . . . . . . . . 56
7.3 Isogonal and orthogonal Trajectories . . . . . . . . . . . . . . . . . . . . . . 57
7.4 Initial Value Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
7.5 Boundary Value Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
7.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

8 Selected Topics in Topology and Geometry . . . . . . . . . . . . . . . . 65

8.1 What is Topology? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
8.2 Euler Characteristic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
8.3 Homology Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
8.3.1 Simplexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
8.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

9 Selected Topics in Discrete Mathematics . . . . . . . . . . . . . . . . . 75

9.1 The Language of Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
9.2 Combinatorics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
9.3 Graphs and Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
9.4 Recursion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
9.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
 Contents VII

10 Selected Topics in Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

10.1 The Origins of Operations Research . . . . . . . . . . . . . . . . . . . . . . . . 85
10.2 Linear Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
10.3 The Transportation and Assignment Problems . . . . . . . . . . . . . . 88
10.3.1 Transportation Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
10.3.2 Assignment Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
10.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

11 Selected Topics in Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

11.1 Compact space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
11.2 Hilbert space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
11.3 Orthogonal Sets of Vectors and Basis . . . . . . . . . . . . . . . . . . . . . . 97
11.4 Isomorphic Hilbert Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
11.5 Operators on Hilbert Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
11.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

12 Selected Topics in Number Theory . . . . . . . . . . . . . . . . . . . . . . . . 105

12.1 What is number theory? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
12.2 Main subdivisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
12.2.1 Elementary tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
12.2.2 Analytic number theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
12.2.3 Algebraic number theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
12.2.4 Diophantine geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
12.2.5 Probabilistic number theory . . . . . . . . . . . . . . . . . . . . . . . . 107
12.2.6 Computational number theory . . . . . . . . . . . . . . . . . . . . . . 107
12.3 Modular arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
12.3.1 Definition of congruence relation . . . . . . . . . . . . . . . . . . . . 108
12.3.2 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
12.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

13 Famous problems in Math history . . . . . . . . . . . . . . . . . . . . . . . . . 113

13.1 Fermat’s Last Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
13.2 The Four Color Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
1
Selected Topics in Calculus

1.1 Limits
1 4y
Central to calculus is the value of the slope of a line, 4x , but
when both the numerator and denominator become almost zero. To
evaluate the slope, that ratio, under those vanishing conditions,
requires the idea of a limit. And central to the idea of a limit is
the idea of a sequence of rational numbers.
We encounter such a sequence in geometry when we determine
a value for π, which is the ratio of the circumference of a circle to
the diameter. To do that, we inscribe in the circle a regular polygon.
The ratio of the perimeter of the polygon to the diameter, which
we can actually calculate, will be an approximation to π. And as
we increase the number of sides – that is, if we consider a sequence
of polygons: 60 sides, 61 sides, 62, 63, 64, and so on – then the
sequence of those ratios gets closer and closer to π. Now, the circle
is never equal to any polygon. But by considering a sufficiently
large number of sides, the difference between the circle and that
polygon, the error, will be less than any small number we name.
Less even than
0.00000000000000000000000000000001!
That is the idea of a sequence approaching a limit, or a boundary.
By that process, we can approximate the value of π as closely as
we possibly can. For instance, the reader surely can recognize the
1
This section has been quoted from:
“http://www.salonhogar.net/themathpage/aCalc/limits.htm”.
2 1 Selected Topics in Calculus

number that is the limit of this sequence of rational numbers.

3, 3.1, 3.14, 3.141, 3.1415, 3.14159, 3.141592, . . .
In fact, when we approximate any irrational number, it will be the
limit of the sequence of its rational approximations. And again, the
limit, the irrational number, will not be a member of the sequence.
We speak of a sequence as being infinite, which is a brief way
of saying that, no matter how many terms we have named
already, we could always name one more. And if the sequence
has a limit, then each term we add will be closer.

The limit of a variable

Consider this sequence of values of a variable x:

1.9, 1.99, 1.999, 1.9999, 1.99999, . . .
These values are getting closer and closer to 2 – they are ap-
proaching 2 as their limit. 2 is the smallest number such that no
matter which term of that sequence we consider, it will be less
than 2.
We can define “closer and closer” mathematically, by saying
that the differences between the terms of that sequence and 2 be-
come and remain less than any small number we might name. That
is, we can define a limit by considering the sequence of differences,
x − 2:

1.9 − 2, 1.99 − 2, 1.999 − 2, 1.9999 − 2, 1.99999 − 2, . . .

Now name any positive number, however small, for example,
“The width of a hydrogen atom”. Then if we consider a suffi-
cient number of those differences, the absolute value of a difference
(which is negative) will be less than that small number. The error,
in other words, between the terms of a sequence and their limit,
can be made as small as we please.
When a variable x approaches a limit l, we symbolize that as
x → l. Read: “The values of x approach l as a limit,” or simply,“x
approaches l”. We also say that a sequence converges to a limit.
The sequence above converges to 2.
 1.2 Derivative 3

1.2 Derivative
2
The problem of finding the tangent line to a curve and the prob-
lem of finding the velocity of an object both involve finding the
same type of limit. This special type of limit is called a derivative
and we will see that it can be interpreted as a rate of change in
any of the sciences or engineering.
In general, suppose an object moves along a straight line accord-
ing to an equation of motion s = f (t), where s is the displacement
(directed distance) of the object from the origin at time t. The
function f that describes the motion is called the position func-
tion of the object. In the time interval from t = a to t = a + h the
change in position is f (a + h) − f (a). The average velocity over
this time interval is
displacement f (a + h) − f (a)
Average velocity = = ,
time h
which is the same as the slope of the secant line.
Now suppose we compute the average velocities over shorter and
shorter time intervals [a, a + h]. In other words, we let h approach
0. We define the velocity (or instantaneous velocity) at time t = a
to be the limit of these average velocities:
f (a + h) − f (a)
v(a) = lim
h→0 h
This means that the velocity at time t = a is equal to the slope of
the tangent line at this point.
We have seen that the same type of limit arises in finding the
slope of a tangent line or the velocity of an object. In fact, limits
of the form
f (a + h) − f (a)
lim
h→0 h
arise whenever we calculate a rate of change in any of the sciences
or engineering, such as a rate of reaction in chemistry or a marginal
cost in economics. Since this type of limit occurs so widely, it is
given a special name and notation, namely derivative.
2
This section is quoted from:
Stewart, James, Daniel K. Clegg, and Saleem Watson. Calculus: early transcen-
dentals. Cengage Learning, 2020.
4 1 Selected Topics in Calculus

1.3 Definite and Indefinite Integrals

3
A physicist who knows the velocity of a particle might wish to
know its position at a given time. An engineer who can measure
the variable rate at which water is leaking from a tank wants to
know the amount leaked over a certain time period. A biologist
who knows the rate at which a bacteria population is increasing
might want to deduce what the size of the population will be at
some future time. In each case, the problem is to find a function
F whose derivative is a known function f . If such a function F
exists, it is called an antiderivative of f .

Indefinite Integral

The Fundamental Theorem of Calculus is appropriately named

because it establishes a connection between the two branches of
calculus: differential calculus and integral calculus. Differential cal-
culus arose from the tangent problem, whereas integral calculus
arose from a seemingly unrelated problem, the area problem.
Newtons mentor at Cambridge, Isaac Barrow (1630 - 1677),
discovered that these two problems are actually closely related.
In fact, he realized that differentiation and integration are inverse
processes. The Fundamental Theorem of Calculus gives the precise
inverse relationship between the derivative and the integral. It
was Newton and Leibnitz who exploited this relationship and used
it to develop calculus into a systematic mathematical method.
In particular, they saw that the Fundamental Theorem enabled
them to compute areas and integrals very easily without having
to compute them as limits of sums.
We end this section by bringing together the two parts of the
Fundamental Theorem.
Theorem 1.1 Suppose f is continuous on [a, b].
Z x
1. If g(x) = f (t) dt, then g 0 (x) = f (x).
a

3
This section is quoted from:
Stewart, James, Daniel K. Clegg, and Saleem Watson. Calculus: early transcen-
dentals. Cengage Learning, 2020.
 1.3 Definite and Indefinite Integrals 5
Z b
2. f (t) dt = F (b) − F (a),where F is an antiderivative of f ,
a
that is F 0 = f .
Both parts of the Fundamental Theorem establish connections
between antiderivativesZ and definite integrals. Part 1 says that if
x
f is continuous, then f (t) dt is an antiderivative of f . Part
Z b a

2 says that f (t) dt can be found by evaluating F (b) − F (a),

a
where F is an antiderivative of f . We need a convenient notation
for antiderivatives that makes them easy to work with. Because
of the relation given by the Fundamental Z Theorem between an-
tiderivatives and integrals, the notation f (x) dx is traditionally
used for an antiderivative of and is called an indefinite integral.

Definite Integral

A limit of the form

n
X
lim f (x∗i )4x = lim [f (x∗1 )4x + f (x∗2 )4x + · · · + f (x∗n )4x]
n→∞ n→∞
i=0

arises when we compute an area. We also saw that it arises when

we try to find the distance traveled by an object. It turns out that
the same type of limit occurs in a wide variety of situations even
when f is not necessarily a positive function. Later, we will see that
limits of this form also arise in finding lengths of curves, volumes
of solids, centers of mass, force due to water pressure, and work,
as well as other quantities. We therefore give this type of limit a
special name and notation.
Definition 1.1 If a function f is defined for a ≤ x ≤ b,
we divide the interval [a, b] into n subintervals of equal width
4x = (b − a)/n. We let x0 (= a), x1 , x2 , . . . , xn (= b) be the end-
points of these subintervals and we let x∗0 , x∗1 , x∗2 , . . . , x∗n be any
sample points in these subintervals, so x∗i lies in the ith subinter-
val [xi , xi+1 ]. Then the definite integral of f from a to b is
6 1 Selected Topics in Calculus
Z b n
X
f (x) dx = lim f (x∗i )4x
a n→∞
i=0

provided that this limit exists. If it does exist, we say that f is

integrable on [a, b].

1.4 Exercises

1. Translate the following sentences.

1. A limit is the value that a function (or sequence) approaches as
the input (or index) approaches some value. Limits are essential
to calculus (and mathematical analysis in general) and are used
to define continuity, derivatives, and integrals.
2. The derivative of a function of a real variable measures the
sensitivity to change of the function value (output value) with
respect to a change in its argument (input value).
3. The slope or gradient of a line is a number that describes both
the direction and the steepness of the line.
4. The absolute value or modulus of a real number x, denoted by
|x|, is the non-negative value of x without regard to its sign.
5. A function is a relation between sets that associates to every
element of a first set exactly one element of the second set.
6. A differentiable function of one real variable is a function whose
derivative exists at each point in its domain.
7. A variable is a symbol used to represent an arbitrary element
of a set. In addition to numbers, variables are commonly used
to represent vectors, matrices and functions.
8. A sequence is an enumerated collection of objects in which
repetitions are allowed and order does matter.
9. An integral assigns numbers to functions in a way that can
describe displacement, area, volume, and other concepts that
arise by combining infinitesimal data.
10. There are many ways of formally defining an integral, not all
of which are equivalent. The most commonly used definitions
of integral are Riemann integrals and Lebesgue integrals.
 1.4 Exercises 7

2. Write down other forms of the following words.

Noun Verb adjective adverb

Continuous
Derivative
Converge
limit
Computational
Calculation
Originally
Approximately
Increasing

3. Use the correct form of the word.

1. A rigorous definition of (Continuous)........................ of real
functions is usually given in a first course in calculus in terms
of the idea of a limit.
2. Pointwise (converge) ........................ is one of various senses
in which a sequence of functions can converge to a particular
function.
3. The principles of (integral) ........................ were formulated in-
dependently by Isaac Newton and Gottfried Wilhelm Leibniz
in the late 17th century, who thought of the integral as an
infinite sum of rectangles of infinitesimal width.
4. The process of finding a derivative is called (derivative) ............
5. In analytic geometry, an asymptote of a curve is a line such
that the distance between the curve and the line (approach)
......................... zero as one or both of the x or y coordinates
tends to infinity.
6. Familiar examples of (inscribe) ......................... figures include
circles inscribed in triangles or regular polygons, and triangles
or (regulate)................ polygons inscribed in circles.
8 1 Selected Topics in Calculus

4. Fill gaps with one of the following words

1. curvature 2. subspace 3. straight 4. denominator
5. variable 6. numerator 7. integral 8. real
9. function 10. set 11. diameter 12. ratio
1. The notion of line or .......... line was introduced by ancient
mathematicians to represent straight objects (i.e., having no
curvature) with negligible width and depth.
2. Intuitively, the .......... is the amount by which a curve deviates
from being a straight line, or a surface deviates from being a
plane.
3. ............... is a subset of a space which is a space in its own
right.
4. In a formula, a dependent variable is a .......... that is implicitly
a function of another (or several other) variables.
5. ................ is the part of a fraction that is above the line and
signifies the number to be divided by the ..........
6. .......... is a mathematical correspondence that assigns exactly
one element of one set to each element of the same or another
...........
7. Indefinite .......... is any function whose derivative is a given
function.
8. A ........ number that is not rational is called irrational.
9. The number π is a mathematical constant. It is defined as the
.......... of a circle’s circumference to its .........., and it also has
various equivalent definitions.
2
Selected Topics in Linear Algebra

2.1 Vector Space

1
There are two approaches to linear algebra, each having its virtues.
The first is abstract. A vector space is defined axiomatically as
a collection of objects, called vectors, with a sum and a scalar-
vector product. As the theory develops, matrices emerge, almost
incidentally, as scalar representations of linear transformations. The
advantage of this approach is generality. The disadvantage is that
the hero of our story, the matrix, has to wait in the wings.
The second approach is concrete. Vectors and matrices are de-
fined as arrays of scalars - here arrays of real or complex numbers.
Operations between vectors and matrices are defined in terms of
the scalars that compose them. The advantage of this approach
for a treatise on matrix computations is obvious: it puts the ob-
jects we are going to manipulate to the fore. Moreover, it is truer
to the history of the subject. Most decompositions we use today
to solve matrix problems originated as simplifications of quadratic
and bilinear forms that were defined by arrays of numbers.
Although we are going to take the concrete approach, the con-
cepts of abstract linear algebra will not go away. It is impossible
to derive and analyze matrix algorithms without a knowledge of
such things as subspaces, bases, dimension, and linear transforma-
tions. Consequently, after introducing vectors and matrices and
describing how they combine, we will turn to the concepts of lin-
1
This section is quoted from:// GW Stewart, Matrix Algorithms: Volume 1: Basic
Decompositions, SIAM, (1998)
10 2 Selected Topics in Linear Algebra

ear algebra. This inversion of the traditional order of presentation

allows us to use the power of matrix methods to establish the basic
results of linear algebra.
The results of linear algebra apply to vector spaces over an ar-
bitrary field . However, we will be concerned entirely with vectors
and matrices composed of real and complex numbers. What distin-
guishes real and complex numbers from an arbitrary field of scalars
is that they posses a notion of limit. This notion of limit extends
in a straightforward way to finite-dimensional vector spaces over
the real or complex numbers, which inherit this topology by way
of a generalization of the absolute value called the norm. Moreover,
these spaces have a Euclidean geometry - e.g ., we can speak of
the angle between two vectors.

2.1.1 Vectors

Since we are going to define matrices as two-dimensional arrays of

numbers, called scalars, we could regard a vector as a degenerate
matrix with a single column, and a scalar as a matrix with one
element. In fact, we will make such identifications later. However,
the words “scalar ” and “vector ” carry their own bundles of associ-
ations, and it is therefore desirable to introduce and discuss them
independently.

Scalars

Although vectors and matrices are represented on a computer by

floating-point numbers - and we must ultimately account for the
inaccuracies this introduces - it is convenient to regard matrices
as consisting of real or complex numbers. We call these numbers,
scalars.

Real and complex numbers

The set of real numbers will be denoted by R. As usual, |x| will

denote the absolute value of x ∈ R. The set of complex numbers
will be denoted by C. Any complex number z can be written in
the form z = x + i y, where x and y are real and i is the principal
square root of −1. The number x is the real part of z and is written
 2.1 Vector Space 11

Re z. The number y is the imaginary part of z andp

is written Im
z. The absolute value, or modulus, of z is |z| = x2 + y 2 . The
conjugate x − iy of z will be written z̄.

Vectors

In three dimensions, a directed line segment can be specified by

three numbers x, y, and z. The following definition is a natural
generalization of this observation.
Definition 2.1 A vector x of dimension n or n-vector is an
array of n scalars of the form x = (x1 , x2 , . . . , xn ). The scalars
xi are called the components of x. The set of n-vectors with real
components will be written Rn . The set of n-vectors with real or
complex components will be written Cn .

Operations with vectors and scalars

Vectors can be added and multiplied by scalars. These operations

are performed componentwise as specified in the following defini-
tion.
Definition 2.2 Let x and y be n-vectors and α be a scalar.
The sum of x and y is the vector
 
 1x + y1

 
 x2 + y2 
x+y = 
..
.
.

 
 
xn + yn
The scalar-vector product αx is the vector
 
 αx1 
 
 αx2 
αx = 
 ..  .

 . 
 
αxn
The following properties are easily established from the definitions
of the vector sum and scalar-vector product.
12 2 Selected Topics in Linear Algebra

Theorem 2.1 Let x, y, and z be n-vectors and α and β be

scalars. Then
1. x+y =y+x
2. (x + y) + z = x + (y + z)
3. x+0=x
4. x + (−1)x = 0
5. (αβ)x = α(βx)
6. (α + β)x = αx + βx
7. α(x = y) = αx + αy
8. 1.x = x
The properties listed above insure that a sum of products of
the form α1 x1 + α2 x2 + · · · + αn xn is unambiguously defined and
independent of the order of summation. Such a sum of products
is called a linear combination of the vectors x1 , x2 , . . . , xn .
The properties listed in Theorem 2.1 are sufficient to define a
useful mathematical object called a vector space or linear space.
Specifically, a vector space consists of a field F of objects called
scalars and a set of objects x called vectors.

2.1.2 Matrices

Matrices and the matrix-vector product arise naturally in the

study of systems of equations. An m×n system of linear equations

a11 x1 + a12 x2 + · · · +a1n xn = b1

a21 x1 + a22 x2 + · · · +a2n xn = b2
.. .. .. .. .. (2.1)
. . . . .
am1 x1 + am2 x2 + · · · +amn xn = bm
can be written compactly in the form
n
X
aij x= bi , i = 1, 2, . . . , m
j=1

However, matrices provide an even more compact representation.

If we define arrays A, x, and b by
 2.2 Eigenvalues and Eigenvectors 13
     
 a11 a12 . . . a1n  x
 1  b1 
     
 a21 a22 . . . a2n  x  b 
A= , x =  2 , b =  2 ,
 .. .. . . ..   ..   .. 
 . . . .   .   . 
     
am1 am2 . . . amn xn bm

and define the product Ax by the left-hand side of (2.1), then (2.1)
is equivalent to Ax = b. The scalars aij are called the elements of
A. The set m × n matrices with real elements is written Rm×n .
The set of m × n matrices with real or complex components is
written Cm×n . The indices i and j of the elements aij of a matrix
are called respectively the row index and the column index.

2.2 Eigenvalues and Eigenvectors

2
The polynomial p(λ) = det(A − λI) is called the characteristic
polynomial of A. The roots of p(λ) = 0 are the eigenvalues of A.
Since the degree of the characteristic polynomial p(λ) equals n,
the dimension of A, it has n roots, so A has n eigenvalues.
A nonzero vector x satisfying Ax = λx is a (right) eigenvector
for the eigenvalue λ. A nonzero vector y such that y T A = λy T is
a left eigenvector.
Most of algorithms will involve transforming the matrix A into
simpler, or canonical forms, from which it is easy to compute
its eigenvalues and eigenvectors. These transformations are called
similarity transformations. The two most common canonical forms
are called the Jordan form and Schur form. The Jordan form is
useful theoretically but is very hard to compute in a numerically
stable fashion, which is why we will aim to compute the Schur
form instead.
To motivate Jordan and Schur forms, let us ask which matri-
ces have the property that their eigenvalues are easy to compute.
The easiest case would be a diagonal matrix, whose eigenvalues
are simply its diagonal entries. Equally easy would be a triangu-
lar matrix, whose eigenvalues are also its diagonal entries. Below
2
This section is quoted from:
Applied Numerical Linear Algebra - James W. Demmel Cambridge University
Press, 1997
14 2 Selected Topics in Linear Algebra

we will see that a matrix in Jordan or Schur form is triangular.

But recall that a real matrix can have complex eigenvalues, since
the roots of its characteristic polynomial may be real or complex.
Therefore, there is not always a real triangular matrix with the
same eigenvalues as a real general matrix, since a real triangular
matrix can only have real eigenvalues. Therefore, we must either
use complex numbers or look beyond real triangular matrices for
our canonical forms for real matrices.
It will turn out to be sufficient to consider block triangular ma-
trices, i.e. matrices of the form
 
 A11 A12 . . . A1b 
 
 A22 . . . A2b 
 
..
.
 
 
 
Abb

where each Aii is square and all entries below the Aii blocks
are zero. One can easily show that the characteristic polynomial
Y b
det(A − λI) of A is the product det(Aii − λI) of the charac-
i=1
teristic polynomials of the Aii and therefore that the set λ(A) of
[b
eigenvalues of A is the union λ(Aii ) of the sets of eigenvalues of
i=1
the diagonal blocks Aii . The canonical forms that we compute will
be block triangular and will proceed computationally by break-
ing up large diagonal blocks into smaller ones. If we start with
a complex matrix A, the final diagonal blocks will be 1-by-1, so
the ultimate canonical form will be triangular. If we start with a
real matrix A, the ultimate canonical form will have 1-by-1 diago-
nal blocks (corresponding to real eigenvalues) and 2-by-2 diagonal
blocks (corresponding to complex conjugate pairs of eigenvalues);
such a block triangular matrix is called quasi-triangular.
 2.3 Exercises 15

2.3 Exercises

1. Translate the following sentences.

1. The first modern and more precise definition of a vector space
was introduced by Peano in 1888, a theory of linear transfor-
mations of finite-dimensional vector spaces had emerged.
2. An element of a specific vector space may have various nature;
for example, it could be a sequence, a function, a polynomial
or a matrix. Linear algebra is concerned with those properties
of such objects that are common to all vector spaces.
3. A set of vectors is linearly independent if none is in the span
of the others.
4. Matrices allow explicit manipulation of finite-dimensional vec-
tor spaces and linear maps. Their theory is thus an essential
part of linear algebra.
5. Systems of linear equations form a fundamental part of linear
algebra. Historically, linear algebra and matrix theory has been
developed for solving such systems. In the modern presentation
of linear algebra through vector spaces and matrices, many
problems may be interpreted in terms of linear systems.
6. A symmetric matrix is always diagonalizable. There are non-
diagonalizable matrices.
7. A linear form is a linear map from a vector space V over a field
F to the field of scalars F , viewed as a vector space over itself.
8. The inner product is an example of a bilinear form, and it
gives the vector space a geometric structure by allowing for
the definition of length and angles
9. Nearly all scientific computations involve linear algebra. Conse-
quently, linear algebra algorithms have been highly optimized.
10. In multilinear algebra, one considers multivariable linear trans-
formations, that is, mappings that are linear in each of a num-
ber of different variables.
16 2 Selected Topics in Linear Algebra

2. Write down other forms of the following words.

Noun Verb adjective adverb

analysis
axiomatically
combine
dependent
characterize
Calculation
sufficiently
stable
representation

3. Use the correct form of the word.

1. Linear algebra initially emerged as a method for (solve)..........
systems of linear equations.
2. In a coordinate system, there is a dimension and an (associate)
............ number of coordinates.
3. Here, vector spaces are the central objects of study, and linear
(transform) .......... are the mappings between them.
4. By definition of linear independence, 0 may be (express)..........
as a linear combination of elements of S in only one way.
5. In order to (discussion) .......... the “dimension” of a vector
space, it is important to realize that this is not a defining con-
cept of vector space.
6. Matrices are useful in a variety of fields and form the basis for
linear algebra. Their (apply) .......... include solving systems
of linear equations, path-finding in graph theory, and several
applications in group theory (especially representation theory).
7. Matrix multiplication is not commutative. In other words, it is
not (general) .......... true that AB = BA.
 2.3 Exercises 17

4. Fill gaps with one of the following words

1. analysis 2. space 3. definition 4. product
5. normed 6. complete 7. finite 8. linear
9. function 10. integrable 11. metric 12. measures
Vector spaces that are not .......... dimensional often require addi-
tional structure to be tractable. A .......... vector space is a vector
space along with a function called a norm, which measures the
“size” of elements. The norm induces a .........., which .......... the
distance between elements, and induces a topology, which allows
for a .......... of continuous maps. The metric also allows for a def-
inition of limits and completeness - a metric space that is ..........
is known as a Banach space.
A complete metric space along with the additional structure
of an inner .......... (a conjugate symmetric sesquilinear form) is
known as a Hilbert .........., which is in some sense a particularly
well-behaved Banach space. Functional analysis applies the meth-
ods of .......... algebra alongside those of mathematical analysis to
study various .......... spaces; the central objects of study in func-
tional .......... are Lp spaces, which are Banach spaces, and espe-
cially the Lp space of square .......... functions, which is the only
Hilbert space among them. Functional analysis is of particular im-
portance to quantum mechanics, the theory of partial differential
equations, digital signal processing, and electrical engineering. It
also provides the foundation and theoretical framework that un-
derlies the Fourier transform and related methods.
3
Selected Topics in Statistics

3.1 Statistics
1
Statistics is the science of sampling. How one set of measurements
differs from another and what the implications of those differences
might be are its primary concerns. Conceptually, the subject is
rooted in the mathematics of probability, but its applications are
everywhere. Statisticians are as likely to be found in a research lab
or a field station as they are in a government office, an advertising
firm, or a college classroom.
Properly applied, statistical techniques can be enormously effec-
tive in clarifying and quantifying natural phenomena. In general,
statistical techniques are employed either to (1) describe what did
happen or (2) predict what might happen. It is unarguably true
that the interplay between description and prediction.

3.2 Descriptive Statistics

Descriptive statistics are used to describe the main features of a

collection of data in quantitative terms. Descriptive statistics are
distinguished from inferential statistics (or inductive statistics), in
that descriptive statistics aim to quantitatively summarize a data
set, rather than being used to support inferential statements about
1
This section is quoted from:
An Introduction to Mathematical Statistics and Its Applications; Fourth Edition
; Richard J. Larsen, Vanderbilt University and Morris L. Marx University of West
Florida ; New Jersey .(2006) pearson Education’ Inc.
20 3 Selected Topics in Statistics

the population that the data are thought to represent. Even when
a data analysis draws its main conclusions using inductive statis-
tical analysis, descriptive statistics are generally presented along
with more formal analysis. For example in a paper reporting on
a study involving human subjects, there typically appears a table
giving the overall sample size, sample sizes in important subgroups
(e.g. for each treatment or exposure group), and demographic or
clinical characteristics such as the average age, the proportion of
subjects with each gender, and the proportion of subjects with
related comorbidities.

3.2.1 Mean

In statistics, mean has two related meanings:

• the arithmetic mean (and is distinguished from the geometric
mean or harmonic mean).
• the expected value of a random variable, which is also called the
population mean.
For a real-valued random variable X, the mean is the expectation
of X.

3.2.2 Arithmetic Mean

The arithmetic mean (or simply the mean) of a list of numbers is

the sum of all of the list divided by the number of items in the
list. If the list is a statistical population, then the mean of that
population is called a population mean. If the list is a statistical
sample, we call the resulting statistic a sample mean.
The mean is the most commonly-used type of average and is of-
ten referred to simply as the average. The term mean or arithmetic
mean is preferred in mathematics and statistics to distinguish it
from other averages such as the median and the mode.

3.2.3 Geometric Mean

The geometric mean is a type of mean or average, which indicates

the central tendency or typical value of a set of numbers. It is
similar to the arithmetic mean, which is what most people think
of with the word average, except that instead of adding the set of
 3.2 Descriptive Statistics 21

numbers and then dividing the sum by the count of numbers in

the set, n, the numbers are multiplied and then the nth root of
the resulting product is taken. For instance, the geometric mean of
two numbers, say 2 and√8, is just the square root of their product
which equals 4; that is 2 × 8 = 4.
The geometric mean can also be understood in terms of geom-
etry. The geometric mean of two numbers, a and b, is the length
of one side of a square whose area is equal to the area of a rect-
angle with sides of lengths a and b. Similarly, the geometric mean
of three numbers, a, b, and c, is the length of one side of a cube
whose volume is the same as that of a cuboid with sides whose
lengths are equal to the three given numbers.
The geometric mean only applies to positive numbers. It is also
often used for a set of numbers whose values are meant to be
multiplied together or are exponential in nature, such as data on
the growth of the human population or interest rates of a financial
investment. The geometric mean is also one of the three classic
Pythagorean means, together with the aforementioned arithmetic
mean and the harmonic mean.

3.2.4 Harmonic Mean

The harmonic mean (formerly sometimes called the subcontrary

mean) is one of several kinds of average. Typically, it is appropriate
for situations when the average of rates is desired. The harmonic
mean H of the positive real numbers x1 , x2 , . . . , xn is defined to be
n n
H= 1 1 1 = Xn , xi > 0 for all i.
x1
+ x 2
+ · · · + xn
1
i=1
xi

Equivalently, the harmonic mean is the reciprocal of the arith-

metic mean of the reciprocals.

3.2.5 Median

A median is described as the numeric value separating the higher

half of a sample, a population, or a probability distribution, from
the lower half. The median of a finite list of numbers can be found
by arranging all the observations from lowest value to highest value
22 3 Selected Topics in Statistics

and picking the middle one. If there is an even number of obser-

vations, then there is no single middle value, so one often takes
the mean of the two middle values. The median can be used as
a measure of location when a distribution is skewed , when end
values are not known, or when one requires reduced importance
to be attached to outliers, e.g. because they may be measurement
errors. A disadvantage of the median is the difficulty of handling
it theoretically.

3.2.6 Mode

The mode is the value that occurs the most frequently in a data
set or a probability distribution. In some fields, notably education,
sample data are often called scores, and the sample mode is known
as the modal score. Like the statistical mean and the median, the
mode is a way of capturing important information about a random
variable or a population in a single quantity. The mode is in general
different from the mean and median, and may be very different for
strongly skewed distributions. The mode is not necessarily unique,
since the same maximum frequency may be attained at different
values. The most ambiguous case occurs in uniform distributions,
wherein all values are equally likely.

3.2.7 Range

The range is the length of the smallest interval which contains all
the data. It is calculated by subtracting the smallest observation
(sample minimum) from the greatest (sample maximum) and pro-
vides an indication of statistical dispersion. It is measured in the
same units as the data. Since it only depends on two of the obser-
vations, it is a poor and weak measure of dispersion except when
the sample size is large.
The range, in the sense of the difference between the highest and
lowest scores, is also called the crude range. When a new scale for
measurement is developed, then a potential maximum or minimum
will emanate from this scale. This is called the potential (crude)
range. Of course this range should not be chosen too small, in or-
der to avoid a ceiling effect. When the measurement is obtained,
the resulting smallest or greatest observation, will provide the ob-
served (crude) range.
 3.2 Descriptive Statistics 23

3.2.8 Standard Deviation

The standard deviation of a statistical population, a data set, or a

probability distribution is the square root of its variance. Standard
deviation is a widely used measure of the variability or dispersion,
being algebraically more tractable though practically less robust
than the expected deviation or average absolute deviation.
It shows how much variation there is from the “average”
(mean). It may be thought of as the average difference of the
scores from the mean of distribution, how far they are away from
the mean. A low standard deviation indicates that the data points
tend to be very close to the mean, whereas high standard deviation
indicates that the data are spread out over a large range of values.
In addition to expressing the variability of a population, standard
deviation is commonly used to measure confidence in statistical
conclusions.

3.2.9 Coefficient of Variation

The coefficient of variation is a normalized measure of dispersion of

a probability distribution. It is defined as the ratio of the standard
deviation σ to the mean µ : cv = σµ
This is only defined for non-zero mean, and is most useful for
variables that are always positive. It is also known as unitized risk
or the variation coefficient. The coefficient of variation should only
be computed for data measured on a ratio scale. As an example,
if a group of temperatures are analyzed, the standard deviation
does not depend on whether the Kelvin or Celsius scale is used.
However the mean temperature of the data set would differ in
each scale and thus the coefficient of variation would differ. So the
coefficient of variation does not have any meaning for data on an
2
interval scale. The variance-to-mean ratio, σµ , is another similar
ratio, but is not dimensionless, and hence not scale invariant.
A percentile is the value of a variable below which a certain
percent of observations fall. So the 20th percentile is the value (or
score) below which 20 percent of the observations may be found.
The term percentile and the related term percentile rank are often
used in descriptive statistics as well as in the reporting of scores
from norm-referenced tests.
24 3 Selected Topics in Statistics

The 25th percentile is also known as the first quartile(Q1); the

50th percentile as the median or second quartile(Q2); the 75th
percentile as the third quartile (Q3).

3.2.10 Variance

The variance of a random variable or distribution is the expected,

or mean, value of the square of the deviation of that variable from
its expected value or mean. Thus the variance is a measure of
the amount of variation within the values of that variable, taking
account of all possible values and their probabilities or weightings
(not just the extremes which give the range).

3.3 Exercises

1. Translate the following text.

Statisticians gather information through observational studies
and experiments. Observational studies observe and measure spe-
cific characteristics without modifying the subjects under study. In
contrast, a statistical experiment applies a treatment to the sub-
jects to see if a causal relationship exists. (Treatments can also be
called factors, but this can be confusing because latent variables
in factor analysis are also called factors.) Statistical experiments
are designed to compare the outcomes of applying one or more
treatments to experimental units, then comparing the results to a
control group that does not receive a treatment.
Designing a statistical experiment starts with identifying the
question(s) you want to answer. By carefully planning all the de-
tails of the experiment in advance, you decrease the probability
of errors and increase the likelihood that the experiment will pro-
duce good data that leads to sound conclusions after analysis. The
purpose of an experiment is to see if applying a treatment results
in any observable differences in the experimental units. For a dif-
ference to matter, it needs to be measurable.
The independent variable is the one that you plan to change.
The dependent variable is whatever you plan to measure after the
treatment. Most real-life studies also have extraneous variables
 3.3 Exercises 25

that impact the results of the experiment. There are often variables
that you don’t even know about. A confounding variable is an
extraneous variable that varies across the independent variable.

2. Write down other forms of the following words.

Noun Verb adjective adverb

category
predictable
normalize
dependent
indicate
deviation
summarize
inferentially
expectation

3. Use the correct form of the words.

1. Sample data are the result of making a series of (observe)

............s.
2. In a narrow sense, statistics is a branch of probability theory
and its job is to determine stochastic parameters from (exper-
iment) ........... data.
3. (Statistics)........... problems arise if the probability parameters
of an experiment are unknown.
4. A statistical test is always based on certain model (assume)
............s about the population from which our sample comes.
5. Regardless of what we are trying to measure, the qualities that
make a good measure of a scientific concept are high (reliable)
.........., absence of bias, low cost, practicality, objectivity, high
acceptance, and high concept (valid) ..........
26 3 Selected Topics in Statistics

6. Both (category).......... and (quantity).......... variables are often

recorded as numbers, so this is not a reliable guide to the major
distinction between categorical and quantitative variables.

4. Fill gaps with one of the following words

1. distinguished 2. include 3. categories 4. examples

5. Continuous 6. zero 7. data 8. numbers
9. observations 10. both 11. regression 12. presenting

1. Ordinal data are also categorical, but in this case .......... have
an order and can be ranked. Examples include stages of breast
cancer.
2. Numeric data can be discrete or continuous. Discrete data have
fixed values. .......... data can take any value, frequently within
a given range.
3. Data can be broadly .......... as categorical or numeric. Cate-
gorical data may be nominal, ordinal or binary.
4. Binary, or dichotomous, data have only two possible outcomes.
Common .......... are Yes/No or True/False responses, but they
could also include other common epidemiological outcomes,
such as “survive” and“not survive”.
5. Numeric data which may include something such as weight and
length (where the range would be from .......... to, theoretically,
infinity).
6. Nominal Data describes categorical data without an order. Ex-
amples .......... blood groups (O, A, B, AB), eye colour and
marital status.
7. The most obvious first step in assessing a trend is to plot the
............. of interest by year.
8. Time series analysis refers to a particular collection of spe-
cialised .......... methods that illustrate trends in the data.
9. Interval data are numerical data where the differences between
two ........... can be interpreted, but the ratio between two num-
bers is meaningless.
10. Moving averages (or rolling averages) provide a useful way of
.......... time series data.
 3.3 Exercises 27

11. Ratio data are numerical and have a true zero and .......... dif-
ferences and ratios are meaningful.
12. There are four data scales: nominal, ordinal, interval and ratio.
Nominal and ordinal .......... have already been described.
4
Selected Topics in Probability

4.1 Probability
1
Probability theory arose originally in connection with games of
chance and then for a long time it was used primarily to investigate
the credibility of testimony of witnesses in the “ethical” sciences.
Nevertheless, probability has become a very powerful mathemat-
ical tool in understanding those aspects of the world that cannot
be described by deterministic laws. Probability has succeeded in
finding strict determinate relationships where chance seemed to
reign and so terming them “laws of chance” combining such con-
trasting notions in the nomenclature appears to be quite justified.
This introductory chapter discusses such notions as determinism,
chaos and randomness, predictability and unpredictability, some ini-
tial approaches to formalizing randomness and it surveys certain
problems that can be solved by probability theory. This will per-
haps give one an idea to what extent the theory can answer ques-
tions arising in specific random occurrences and the character of
the answers provided by the theory.
Games of chance and the analysis of testimony of witnesses
were originally the basic areas of application of probability theory.
Games of chance involving cards, dice and flipping coins naturally
permitted the creation of appropriate random experiments (this
terminology first appeared in the twentieth century) so that their

1
This section has be quoted from:
Skorokhod, Valeriy. Basic principles and applications of probability theory.
Springer Science & Business Media, 2005.
30 4 Selected Topics in Probability

outcomes had symmetry in relation to the conditions of the exper-

iment. These outcomes were treated as ‘equally likely” and they
were assigned the same probabilities. Thus, if there are s outcomes
in the experiment, each elementary event was assigned a probabil-
ity of 1/s (it is easy to see that an elementary event has that
probability using the additivity of probability and the fact that
the sure event has probability one). If an event is expressed as the
union of r elementary events (r ≤ s), then the probability of A is
r/s by virtue of the additivity. Thus, we arrive at the definition of
probability that has been in use for about two centuries.
The probability of an event A is the quotient of the number of
outcomes favorable to A and the number of all possible outcomes.
The outcomes favorable to A are understood to be those that imply
A.

4.2 Problems of Probability Theory

Initially, probability theory was the study of ways of computing

probabilities of events knowing the probabilities of other given
events. The techniques developed for computing the probabilities
of certain classes of events now form a constituent unit of probabil-
ity but only partly and far from the main part. However, as before,
probability theory only deals with the probabilities of events inde-
pendently of what meaningful sense can be invested in the words
“the probability of event A is p”. This means that probability the-
ory itself does interpret its results meaningfully but in so doing it
does not exclude the term “probability”. There is no statement
like “A always occurs” but rather the statement “A occurs with
probability one”.

4.2.1 Independence

Independence is one of the basic concepts of probability theory. Ac-

cording to Kolmogorov, it is exactly this that distinguishes proba-
bility theory from measure theory. Independence will be discussed
more precisely later on. For the moment, we merely point out
that stochastic independence and physical independence of events
(one event having no effect on another) are identical in content.
 4.2 Problems of Probability Theory 31

Stochastic independence is a precisely-defined mathematical con-

cept to be given below. At this point, we note that independence
was already used in latent form in the definition of random ex-
periment. One of the requirements imposed on an experiment is
the possibility of iterating it indefinitely. To iterate it assumes that
the conditions of the experiment can be reconstructed after which
the one just performed and all of the prior ones have no affect on
the outcome of the next experiment. This means that the events
occurring in different experiments must be independent.

4.2.2 Samples

A sample may be defined in general as follows. There are m finite

sets A1 , A2 , . . . , Am . From each set, we choose an element ai ∈ Ai
one by one. The collection (a1 , . . . , am ) is then the sample. Samples
are distinguished by identification rules (let us say, we are not
interested in the order of the elements in a sample). Each sample
is regarded as an elementary event and the elementary events are
considered to be equally likely.
1. Sampling with replacement. In this instance, the Ai coincide:
Ai = A and the number of samples is nm, where n is the
number of elements in A.
2. Sampling without replacement. A sample is constructed as fol-
lows. A1 = A, A2 = A\{a1 }, . . . , Ak = A\{a1 , . . . , ak−1 }. In
other words, only samples (a1 , . . . , am ), ai ∈ A, are consid-
ered in which all of the elements are distinct. If A has n el-
ements, then the number of samples
  without replacement is
n
n(n − 1) · · · (n − m + 1)/m! = .
m

4.2.3 Conditional Probability

The conditional probability of an event A given event B having

positive probability has occurred is the quantity
P (A ∩ B)
P (A|B) = (4.1)
P (B)
As a function of A, P (A|B) possesses all of the properties of
a probability. The meaning of conditional probability may be ex-
plained as follows. Together with the original experiment, consider
32 4 Selected Topics in Probability

a conditional probability experiment which is performed if event

B has happened in the original experiment. Thus, if the original
experiment has been done n times and B has happened nB times,
then this sequence contains nB conditional experiments. The event
A will have occurred in the conditional experiment if A and B oc-
cur simultaneously, i.e., if A ∩ B occurs. If nA∩B is the number
of experiments in which the event A ∩ B is observed (of the n
carried out), then the relative frequency of occurrence in the nB
conditional experiments is nA∩B /nB . If we replace the relative fre-
quencies by the probabilities, then we have the right-hand side of
(4.1).

4.3 The Poisson Distribution

The Binomial Distribution problems all had  relatively

 small values
n k
for n, so evaluating px (k) = P (X = k) = p (1 − p)n−k was
k
not particularly difficult. But suppose n were 1000 and k, 500.
Evaluating px (500) would be a formidable task for many handheld
calculators, even today. Two hundred years ago, the prospect of
doing cumbersome binomial calculations by hand was a catalyst
for mathematicians to develop some easy-to-use approximations.
One of the first such approximations was the Poisson limit, which
eventually gave rise to the Poisson distribution.
Simeon Denis Poisson (1781-1840) was an eminent French
mathematician and physicist, an academic administrator of some
note, and, according to an 1826 letter from the mathematician
Abel to a friend, Poisson was a man who knew “how to behave
with a great deal of dignity.” One of Poisson’s many interests was
the application of probability to the law, and in 1837 he wrote
Recherches sur la Probabilite de Jugements.
  Included in the latter
n k
is a limit for px (k) = P (X = k) = p (1 − p)n−k that holds
k
when n approaches ∞, p approaches 0, and np remains constant. In
practice, Poisson’s limit is used to approximate hard-to-calculate
binomial probabilities where the values of n and p reflect the con-
ditions of the limit-that is, when n is large and p is small.
 4.4 Exercises 33

4.3.1 The Poisson Limit

2
Deriving an asymptotic expression for the binomial probability
model is a straightforward exercise in calculus, given that np is to
remain fixed as n increases.
Suppose X is a binomial random variable, where
 
n k
px (k) = P (X = k) = p (1 − p)n−k , k = 0, 1, . . . , n
k
If n → ∞ and p → 0 in such a way that A = np remains
constant, then

e−np (np)k
 
n k
lim
n→∞
P (X = k) = lim
n→∞
p (1 − p)n−k =
p→0 p→0 k k!
np=const np=const

4.4 Exercises

1. Translate the following text.

A probability distribution is a mathematical function that pro-
vides the probabilities of occurrence of different possible outcomes
in an experiment. In more technical terms, the probability distri-
bution is a description of a random phenomenon in terms of the
probabilities of events. For instance, if the random variable X is
used to denote the outcome of a coin toss (“the experiment”),
then the probability distribution of X would take the value 0.5
for X = heads, and 0.5 for X = tails (assuming the coin is fair).
Examples of random phenomena can include the results of an ex-
periment or survey.
A probability distribution is specified in terms of an underly-
ing sample space, which is the set of all possible outcomes of the
random phenomenon being observed. The sample space may be
2
This part has been quoted from:
An Introduction to Mathematical Statistics and Its Applications; Fourth Edition
; Richard J. Larsen, Vanderbilt University and Morris L. Marx University of West
Florida ; New Jersey .(2006) pearson Education’ Inc.
34 4 Selected Topics in Probability

the set of real numbers or a set of vectors, or it may be a list of

non-numerical values; for example, the sample space of a coin flip
would be {heads, tails}.
2. Write down other forms of the following words.

Noun Verb adjective adverb

expression
relatively
coincide
constructive
explain
frequency
exactly
discussed
condition

3. Use the correct form of the word.

Probability distributions are generally (devision) .......... into
two classes. A discrete probability (distribute) ............(applicable
to the scenarios where the set of possible outcomes is discrete, such
as a coin toss or a roll of dice) can be encoded by a discrete list
of the probabilities of the outcomes, known as a probability mass
function. On the other hand, a continuous probability distribution
(applicable to the scenarios where the set of possible outcomes can
take on values in a continuous range (e.g. real numbers), such as
the temperature on a given day) is typically (description) ...........
by probability density functions (with the probability of any in-
dividual outcome actually being 0). The normal distribution is a
commonly encountered continuous probability distribution. More
complex experiments, such as those (involve) ........... stochastic
processes defined in continuous time, may demand the use of more
general probability measures.
 4.4 Exercises 35

A probability distribution whose sample space is one-(dimension)

............ (for example real numbers, list of labels, ordered labels
or binary) is called univariate, while a distribution whose sample
space is a vector space of dimension 2 or more is called multivari-
ate. A univariate distribution gives the probabilities of a single
random variable taking on various alternative values; a multivari-
ate distribution (a joint probability distribution) gives the proba-
bilities of a random vector a list of two or more random variables
taking on various (combine) ............. of values. Important and
commonly encountered univariate probability distributions (inclu-
sion) ............ the binomial distribution, the hypergeometric dis-
tribution, and the normal distribution. The multivariate normal
distribution is a commonly encountered multivariate distribution.
4. Fill gaps with one of the following words
1. combination 2. sciences 3. explicit 4. curve
5. theorem 6. distributions 7. random 8. quantities
9. function 10. symmetric 11. properties 12. suitable
Normal distributions are important in statistics and are often
used in the natural and social .......... to represent real-valued ran-
dom variables whose distributions are not known. Their impor-
tance is partly due to the central limit ........... It states that, under
some conditions, the average of many samples (observations) of a
random variable with finite mean and variance is itself a ..........
variable whose distribution converges to a normal distribution as
the number of samples increases. Therefore, physical .......... that
are expected to be the sum of many independent processes (such
as measurement errors) often have distributions that are nearly
normal.
Moreover, Gaussian distributions have some unique .......... that
are valuable in analytic studies. For instance, any linear .......... of
a fixed collection of normal deviates is a normal deviate. Many
results and methods (such as propagation of uncertainty and least
squares parameter fitting) can be derived analytically in ..........
form when the relevant variables are normally distributed.
A normal distribution is sometimes informally called a bell
........... However, many other distributions are bell-shaped (such
as the Cauchy, Student’s t, and logistic ..........).
36 4 Selected Topics in Probability

The normal distribution is a subclass of the elliptical distribu-

tions. The normal distribution is .......... about its mean, and is
non-zero over the entire real line. As such it may not be a ..........
model for variables that are inherently positive or strongly skewed,
such as the weight of a person or the price of a share. Such vari-
ables may be better described by other distributions, such as the
log-normal distribution or the Pareto distribution.
5
Selected Topics in Algebra

5.1 Groups
1
Let S be a set. A mapping

S×S →S
is sometimes called a law of composition (of S into itself). If x, y
are elements of S, the image of the pair (x, y) under this mapping
is also called their product under the law of composition, and will
be denoted by xy. (Sometimes, we also write x · y, and in many
cases it is also convenient to use an additive notation, and thus to
write x + y. In that case, we call this element the sum of x and y.
It is customary to use the notation x + y only when the relation
x + y = y + x holds.)
Let S be a set with a law of composition. If x, y, z are elements
of S, then we may form their product in two ways: (xy)z and
x(yz). If (xy)z = x(yz) for all x, y, z in S then we say that the law
of composition is associative.
An element e of S such that ex = x = xe for all x ∈ S is called
a unit element. When the law of composition is written additively,
the unit element is denoted by 0, and is called a zero element. A
unit element is unique, for if e0 is another unit element, we have
e = ee0 = e0 by assumption. In most cases, the unit element is
written simply 1 (instead of e).

1
This chapter has been quoted from:
Serge Lang, Algebra, 2002 Springer-Verlag New York, Inc.
38 5 Selected Topics in Algebra

A monoid is a set G, with a law of composition which is asso-

ciative, and having a unit element (so that in particular, G is not
empty).
It would be possible to define more general laws of composition,
i.e. maps S1 × S2 → S3 using arbitrary sets. One can then express
associativity and commutativity in any setting for which they make
sense. For instance, for commutativity we need a law of composi-
tion f : S × S → T where the two sets of departure are the same.
Commutativity then means f (x, y) = f (y, x), or xy = yx if we
omit the mapping f from the notation. For associativity, we leave
it to the reader to formulate the most general combination of sets
under which it will work. We shall meet special cases later, for
instance arising from maps

S × S → S and S × T → T
Then a product (xy)z makes sense with x ∈ S, y ∈ S, and z ∈
T . The product x(yz) also makes sense for such elements x, y, z
and thus it makes sense to say that our law of composition is
associative, namely to say that for all x, y, z as above we have
(xy)z = x(yz). If the law of composition of G is commutative, we
also say that G is commutative (or Abelian).
By a submonoid of G, we shall mean a subset H of G containing
the unit element e, and such that, if x, y ∈ H then xy ∈ H (we
say that H is closed under the law of composition). It is then clear
that H is itself a monoid, under the law of composition induced by
that of G.
A group G is a monoid, such that for every element x ∈ G there
exists an element y ∈ G such that xy = yx = e. Such an element y
is called an inverse for x. Such an inverse is unique, because if y 0 is
also an inverse for x, then y 0 = y 0 e = y 0 (xy) = (y 0 x)y = ey = y. We
denote this inverse by x−1 (or by −x when the law of composition
is written additively).
Let G be a group. A subgroup H of G is a subset of G contain-
ing the unit element, and such that H is closed under the law of
composition and inverse (i.e. it is a submonoid, such that if x ∈ H
then x−1 ∈ H). A subgroup is called trivial if it consists of the unit
element alone. The intersection of an arbitrary non-empty family
of subgroups is a subgroup (trivial verification).
 5.2 Rings and Homomorphism 39

Let G, G0 be monoids. A monoid-homomorphism (or simply

homomorphism) of G into G0 is a mapping f : G → G0 such that
f (xy) = f (x)f (y) for all x, y ∈ G, and mapping the unit element
of G into that of G0 . If G, G0 are groups, a group-homomorphism
of G into G0 is simply a monoid-homomorphism.
A homomorphism f : G → G0 is called an isomorphism if there
exists a homomorphism g : G0 → G such that f ◦g and g ◦f are the
identity mappings (in G0 and G respectively). It is trivially verified
that f is an isomorphism if and only if f is bijective. The existence
of an isomorphism between two groups G and G0 is sometimes
denoted by G ≈ G0 . If G = G0 , we say that isomorphism is an
automorphism. A homomorphism of G into itself is also called an
endomorphism.

5.2 Rings and Homomorphism

A ring A is a set, together with two laws of composition called

multiplication and addition respectively, and written as a product
and as a sum respectively, satisfying the following conditions:
RI 1 With respect to addition, A is a commutative group.
RI 2 The multiplication is associative, and has a unit element.
RI 3 For all x, y, Z ∈ A we have (x + y)z = xz + yz (This is called
distributivity.)
As usual, we denote the unit element for addition by 0, and the
unit element for multiplication by 1. We do not assume that 1 6= 0.
We observe that 0x = 0 for all x ∈ A. Because, we have 0x + x =
(0 + 1)x = 1x = x. Hence 0x = 0. In particular, if 1 = 0, then
A consists of 0 alone. For any x, y ∈ A we have (−x)y = −(xy).
Because, we have xy + (−x)y = (x + (−x))y = 0y = 0, so (−x)y
is the additive inverse of xy.
Other standard laws relating addition and multiplication are
easily proved, for instance (−x)(−y) = xy.
Let A be a ring, and let U be the set of elements of A which
have both a right and left inverse. Then U is a multiplicative group.
Indeed, if a has a right inverse b, so that ab = 1, and a left inverse
c, so that ca = 1, then cab = b, whence c = b, and we see that c (or
b) is a two-sided inverse, and that c itself has a two-sided inverse,
40 5 Selected Topics in Algebra

namely a. Therefore U satisfies all the axioms of a multiplicative

group, and is called the group of units of A. It is sometimes denoted
by A∗ , and is also called the group of invertible elements of A. A
ring A such that 1 6= 0, and such that every non-zero element is
invertible is called a division ring.
A left ideal a in a ring A is a subset of A which is a subgroup
of the additive group of A, such that Aa ⊂ a (and hence Aa = a
since A contains 1). To define a right ideal, we require Aa = a, and
a two-sided ideal is a subset which is both a left and a right ideal.
A two-sided ideal is called simply an ideal. Note that (0) and A
itself are ideals.
By a ring-homomorphism one means a mapping f : A → B
where A, B are rings, and such that f is a monoid-homomorphism
for the multiplicative structures on A and B, and also a monoid-
homomorphism for the additive structure. In other words, f must
satisfy:
f (a + a0 ) = f (a) + f (a0 ), f (1) = 1, f (aa0 ) = f (a)f (a0 ), f (0) = 0,
for all a, a0 ∈ A. Its kernel is defined to be the kernel of f viewed
as additive homomorphism.
A ring A is said to be commutative if xy = yx for all x, y ∈ A.
A commutative division ring is called a field. We observe that by
definition, a field contains at least two elements, namely 0 and 1.

5.3 Module

Let A be a ring. A left module over A, or a left A-module M

is an Abelian group, usually written additively, together with an
operation of A on M (viewing A as a multiplicative monoid by RI
2), such that, for all a, b ∈ A and x, y ∈ M we have
(a + b)x = ax + bx and a(x + y) = ax + ay.
We leave it as an exercise to prove that a(−x) = −(ax) and
that 0x = 0. By definition of an operation, we have 1x = x.
In a similar way, one defines a right A-module. We shall deal
only with left A-modules, unless otherwise specified, and hence
call these simply A-modules, or even modules if the reference is
clear.
 5.4 Exercises 41

Let M be an A-module. By a submodule N of M we mean an

additive subgroup such that AN ⊂ N . Then N is a module (with
the operation induced by that of A on M ). A module over a field
is called a vector space.
By a module-homomorphism one means a map f : M → M 0
of one module into another (over the same ring A), which is an
additive group-homomorphism, and such that f (ax) = af (x) for
all a ∈ A and x ∈ M . If we wish to refer to the ring A, we also say
that f is an A-homomorphism, or also that it is an A-linear map.
There are some things in mathematics which satisfy all the
axioms of a ring except for the existence of a unit element. Let A
be a commutative ring. Let E, F be modules. By a bilinear map
g : E × E → F we mean a map such that given x ∈ E, the map
y 7→ g(x, y) is A-linear, and given y ∈ E, the map x 7→ g(x, y) is A-
linear. By an A-algebra we mean a module together with a bilinear
map g : E × E → E. We view such a map as a law of composition
on E. Aside from the examples already mentioned, we note that
the group ring A[G] (or monoid ring when G is a monoid) is an
A-algebra, also called the group (or monoid) algebra.

5.4 Exercises

1. Translate the following sentences.

An integer n is said to be a divisor of an integer i if i is an

integer multiple of n; i.e., i = qn for some integer q. Thus all
integers are trivially divisors of 0. The integers that have integer
inverses, namely ±1, are called the units of Z. If u is a unit and n
is a divisor of i, then un is a divisor of i and n is a divisor of ui.
Thus the factorization of an integer can only be unique up to a unit
u, and ui has the same divisors as i. We therefore consider only
factorizations of positive integers into products of positive integers.
Every nonzero integer i is divisible by 1 and i; these divisors are
called trivial. An integer n is said to be a factor of an integer i if
n is positive and a nontrivial divisor of i. For example, 1 has no
nontrivial divisors and thus no factors. A positive integer greater
than 1 that has no nontrivial divisors is called a prime integer.
42 5 Selected Topics in Algebra

2. Write down other forms of the following words.

Noun Verb adjective adverb

multiplication
uniquely
satisfy
invertible
commutate
specification
divisibility
additive
distribution
3. Use the correct form of the word.
1. A ring is an Abelian group with a second binary (operate)
.......... that is associative, is distributive over the Abelian group
operation, and has an identity element.
2. Although ring addition is (commute) .........., ring multiplica-
tion is not required to be commutative: ab need not necessarily
equal ba.
3. The study of rings (origin) .......... from the theory of polyno-
mial rings and the theory of algebraic integers.
4. The (define) .......... of an ideal in a ring is analogous to that of
normal subgroup in a group.
5. The concept of a module over a ring (general) .......... the con-
cept of a vector space (over a field) by generalizing from mul-
tiplication of vectors with elements of a field (scalar multipli-
cation) to multiplication with elements of a ring.
6. A field is a set on which addition, subtraction, multiplication,
and division are (definition) .......... and behaved as the corre-
sponding operations on rational and real numbers do.
7. Finite fields (also called Galois fields) are fields with (finite)..........
many elements, whose number is also referred to as the order
of the field.
 5.4 Exercises 43

4. Fill gaps with one of the following words

1. divides 2. rings 3. multiple 4. finite
5. degree 6. field 7. factors 8. algebraic
9. remainders 10. element 11. group 12. algebra
1. A polynomial g(x) is said to be a divisor of an polynomial f (x)
if f (x) is a polynomial .......... of g(x); i.e., f (x) = q(x)g(x) for
some polynomial q(x).
2. An important example of a finite abelian group is the set of
.......... Rn = {0, 1, . . . , n − 1} under mod-n addition, where n
is any given positive integer.
3. A finite .......... G of order n is called cyclic if it is isomorphic
to Zn .
4. By degree additivity, the .......... of a polynomial f (x) is equal
to the sum of the degrees of its prime factors, which are unique
by unique factorization.
5. Given a positive integer i, we may factor i into a unique product
of prime .......... by simply factoring out primes no greater than
i until we arrive at the quotient 1.
6. If S is a subgroup of a finite group G, then |S| .......... |G|.
7. What is the order of the smallest non-trivial ring with identity
which is not a ..........?
8. A finite ring is a ring that has a .......... number of elements.
9. A monoid is an algebraic structure intermediate between groups
and semigroups, and is a semigroup having an identity ...........
10. An interesting topic in the ring theory is the classification of
finite ...........
11. An associative algebra is an .......... structure with compatible
operations of addition, multiplication (assumed to be associa-
tive), and a scalar multiplication by elements in some field.
12. Boolean .......... is the branch of algebra in which the values
of the variables are the truth values true and false, usually
denoted 1 and 0, respectively.
6
Selected Topics in Numerical Analysis

6.1 rounding Process

1
The conventional process of rounding or “forcing” a number to n
digits (or figures) consists of replacing that number by an n-digit
approximation with minimum error. When this requirement leads
to two permissible roundings, that one for which the nth digit of
the rounded number is even is generally selected. With this rule,
the associated error is never larger in magnitude than one-half
unit in the place of the nth digit of the rounded number. Thus
4.05149 ' 4.0515, 4.051, 4.05, 4.1, and 4. It may be noted here
that whereas 4.05149 rounds to 4.0515, which in turn rounds to
4.052; nevertheless, 4.05149 rounds directly to 4.051. Thus round-
ing is not necessarily transitive.
The errors introduced in the rounding of a large set of num-
bers, which are to be combined in a certain way, usually (but not
always) tend to be equally often positive and negative, so that
their effects often tend to cancel. The slight favoring of even num-
bers is prompted by the fact that any subsequent operations on
the rounded numbers are then somewhat less likely to necessitate
additional round offs.
Each digit of a number, except a zero which serves only to
fix the position of the decimal point, is called a significant digit
or figure of that number. Thus, the numbers 2.159, 0.04072, and
10.00 each contain four significant figures. Whether or not the last
1
This Part has been quoted from:
Hildebrand, Francis Begnaud. Introduction to numerical analysis, Courier Cor-
poration, 2013.
46 6 Selected Topics in Numerical Analysis

digit of 14620 is significant depends upon the context. If “a number

known to be between 14615 and 14625” is intended, then that zero
is not significant and the number would preferably be written in
the form 1.462 × 104 . Otherwise the form 1.4620 × 104 would be
appropriate.
It may be seen that the concept of significant figures is related
more intimately to the relative error
true value − approximation
Relative error =
true value
than to the error (or the absolute error) itself.

6.2 Interpolation

Anyone who has had occasion to consult tables of mathemati-

cal functions is familiar with the method of linear interpolation
and probably has encountered situations in which this method of
“reading between the lines of the table” has appeared to be unreli-
able. If more reliable interpolates are desired, it is clearly necessary
to make use of more information than that consisting of tabulated
values (ordinates) of a function, corresponding to only two succes-
sive abscissas? Whereas that additional information could consist,
for example, of known values of certain derivatives of the function
at those two points, it is supposed in most of what follows that
the interpolation process is to be based only on tabulated values of
the function itself, with any further available information reserved
for use in estimating the error involved.
There exist a number of interpolation formulas which have this
property, most of which possess certain advantages in certain situ-
ations, but no one of which is preferable to all others in all respects.
Whereas certain of these formulas are expressed explicitly in terms
of all the ordinates on which they depend, others involve only one
or two of the ordinates explicitly and express their dependence
upon other ordinates only in terms of differences of ordinates and
successive differences of differences.
 6.4 Numerical Differentiation 47

6.3 Solving Non-Linear Equations

2

We are here concerned with the problem of solving equations

of the form x = f (x), where f is a given function. The method of
iteration consists in executing the following algorithm:
Algorithm: Choose x0 arbitrarily, and generate the sequence
{xn } recursively from the relation
xn = f (xn−1 ), n = 1, 2, · · · . (6.1)
At the outset, we cannot even be sure that this algorithm is
well defined. It could be that f is undefined at some point f (xn ).
However, let us assume that f is defined on some closed finite
interval I = [a, b], and that the values of f lie in the same interval.
Geometrically this means that the graph of the function y = f (x)
is contained in the square a ≤ x ≤ b, a ≤ y ≤ b.
Under this assumption, if x0 ∈ I, we can say that all elements
of the sequence {xn } are in I. For if some xn ∈ I with n > 0, then
also xn+1 = f (xn ) ∈ I since f has its values in I. Observe that
the above hypotheses are not sufficient to guarantee that equation
(6.1) has a solution. However, if we assume the function f to be
continuous, then the equation has at least one solution.
The above intuitive consideration can be couched in purely
analytical terms, as follows. Consider the function g defined by
g(x) = x − f (x), a ≤ x ≤ b. This function is continuous on
the interval [a, b]; moreover, since f (a) > a, f (b) < b, it satisfies
g(a) ≤ 0, g(b) ≥ 0. By the intermediate value theorem of calculus
it assumes all values between g(a) and g(b) somewhere in the in-
terval [a, b]. Therefore it must assume the value zero, say at x = s.
This implies 0 = s − f (s), or s = f (s). Thus the number s is the
desired solution.

6.4 Numerical Differentiation

3
The mathematical definition of the derivative of f (x) is
2
This Section has been quoted from:
Henrici, Peter. Elements of numerical analysis. (1964).
3
This section has been quoted from:
Hjorth-Jensen, Morten. Computational physics, Lecture notes (2011).
48 6 Selected Topics in Numerical Analysis

df (x) f (x + h) − f (x)
= lim
x h→0 h
where h is the step size. If we use a Taylor expansion for f (x), we
can write
h2 00
0
f (x + h) = f (x) + hf (x) + f (x) + · · ·
2
We can then set the computed derivative fc0 (x) as
f (x + h) − f (x) hf 00 (x)
fc0 (x) ' ' f 0 (x) + + ···
h 2
Assume now that we will employ two points to represent the func-
tion f by way of a straight line between x and x + h. This means
that we could represent
f (x + h) − f (x)
f20 (x) = + O(h).
h
where the suffix 2 refers to the fact that we are using two points
to define the derivative and the dominating error goes like O(h).
This is the forward derivative formula. Alternatively, we could use
the backward derivative formula
f (x) − f (x − h)
f20 (x) = + O(h).
h
If the second derivative is close to zero, this simple two point
formula can be used to approximate the derivative.
Above we have used the interpolating polynomial to approxi-
mate values of a function f at points where f is not known. An-
other use of the interpolating polynomial, of equal or even higher
importance in practice, is the imitation of the fundamental oper-
ations of calculus. In all these applications the basic idea is ex-
tremely simple: Instead of performing the operation on the func-
tion f , which may be difficult or-in cases where f is known at
discrete points only-impossible, the operation is performed on a
suitable interpolating polynomial.
 6.5 Numerical Integration 49

6.5 Numerical Integration

4
We now turn to the problem of numerical evaluation of definite
integrals. The method is the same as numerical differentiation. In-
stead of performing the integration on the function f , which may
be difficult, we perform the integration on a polynomial interpo-
lating f at suitable points. If values of f are available on both
sides of the interval of integration, it seems preferable to perform
the integration on a polynomial that takes into account all these
values.
It is assumed that f is continuous on [a, b] and can be evalu-
ated at arbitrary points of that interval. Several procedures offer
themselves; we shall be able to dismiss two of them very briefly.
1. Newton-Cotes formulas. The most natural idea that offers
itself seems to select a certain number of interpolating points
within [a, b], to interpolate f at these points, and to approxi-
mate the integral off by the integral of the interpolating poly-
nomial. If the interpolating points divide [a, b] into N equal
parts, we arrive in this manner at certain integration formulas
which are called the Newton-Cotes formulas. Unfortunately,
these formulas have, for large values of N , some very unde-
sirable properties. In particular, it turns out that there exist
functions, even analytic ones, for which the sequence of the
integrals of the interpolating polynomials does not converge
towards the integral of the function f . Also, the coefficients in
these formulas are large and alternate in sign, which is unde-
sirable for the propagation of rounding error. For these reasons,
the Newton-Cotes formulas are rarely used for high values of
N . For N = 2, 3, 4 the formulas are identical with certain well-
known integration formulas.
2. Gaussian quadrature. One may try to avoid some of the
short-comings of the Newton-Cotes formulas by relinquishing
the equal spacing of the interpolating points. Gauss discov-
ered that by a proper choice of the interpolating points one
can construct integration formulas which, using N + 1 interpo-
lating points, give the accurate value of the integral if f is a
4
This Section has been quoted from:
Henrici, Peter. Elements of numerical analysis. (1964).
50 6 Selected Topics in Numerical Analysis

polynomial of degree 2N +1 or less. These formulas turn out to

be numerically stable, and they are in successful use at a num-
ber of computation laboratories. The formulas suffer from the
disadvantage, however, that the interpolating points as well as
the corresponding weights are irregular numbers that have to
be stored. This practically (although not theoretically) limits
the applicability of these highly interesting formulas.

6.6 Exercises

1. Translate the following sentences.

Numerical integration is the approximate computation of an in-
tegral using numerical techniques. The numerical computation of
an integral is sometimes called quadrature. Ueberhuber uses the
word “quadrature” to mean numerical computation of a univari-
ate integral, and “cubature” to mean numerical computation of a
multiple integral.
There are a wide range of methods available for numerical inte-
gration. The most straightforward numerical integration technique
uses the Newton-Cotes formulas (also called quadrature formulas),
which approximate a function tabulated at a sequence of regularly
spaced intervals by various degree polynomials. If the endpoints
are tabulated, then the 2- and 3-point formulas are called the
trapezoidal rule and Simpson’s rule, respectively. The 5-point for-
mula is called Boole’s rule. A generalization of the trapezoidal rule
is Romberg integration, which can yield accurate results for many
fewer function evaluations.
If the functions are known analytically instead of being tab-
ulated at equally spaced intervals, the best numerical method of
integration is called Gaussian quadrature. By picking the abscissas
at which to evaluate the function, Gaussian quadrature produces
the most accurate approximations possible. However, given the
speed of modern computers, the additional complication of the
Gaussian quadrature formalism often makes it less desirable than
simply brute-force calculating twice as many points on a regular
grid (which also permits the already computed values of the func-
tion to be re-used).
 6.6 Exercises 51

Modern numerical integrations methods based on information

theory have been developed to simulate information systems such
as computer controlled systems, communication systems, and con-
trol systems since in these cases, the classical methods (which are
based on approximation theory) are not as efficient.
2. Write down other forms of the following words.

Noun Verb adjective adverb

algorithm
equally
interpolate
regular
execute
enumerate
intuitively
instability
classify

3. Use the correct form of the word.

1. spline interpolation is a form of (interpolate) .......... where the

interpolant is a special type of piecewise polynomial called a
spline.
2. Cubic Hermite splines are typically used for interpolation of
numeric data specified at given argument values x1 , x2 , . . . , xn ,
to obtain a smooth (continue) .......... function.
3. The finite element method is a (numeric).......... technique for
finding approximate solutions of partial differential equations
as well as of integral equations.
4. Finite-difference methods are numerical methods for (approxi-
mate) ........... the solutions to differential equations using finite
difference equations to approximate derivatives.
52 6 Selected Topics in Numerical Analysis

5. Spectral methods are techniques used in applied mathematics

and (science) .......... computing to numerically solve certain dif-
ferential equations, often involving the use of the Fast Fourier
Transform.

4. Fill gaps with one of the following words

1. computation 2. approximation 3. secant 4. procedure

5. difference 6. solutions 7. functions 8. successive
9. errors 10. digits 11. finding 12. precision

1. When a number is expressed in scientific notation, the number

of significant .......... (or significant figures) is the number of
digits needed to express the number to within the uncertainty
of calculation.
2. The Bisection Method is a .......... approximation method that
narrows down an interval that contains a root of the function
f (x).
3. The relative error of the quotient or product of a number of
quantities is less than or equal to the sum of their relative
...........
4. The Newton-Raphson method (also known as Newton’s method)
is a way to quickly find a good ..........for the root of a real-
valued function f (x) = 0f (x) = 0.
5. The bisection method is a root-finding method that applies to
any continuous .......... for which one knows two values with
opposite signs.
6. The concepts of accuracy and .......... are both closely related
and often confused.
7. The golden section search algorithm can be used for .......... a
minimum (or maximum) of a single-variable function f (x).
8. The Euler method is a first-order numerical .......... for solving
ordinary differential equations with a given initial value. I
9. If rounding is performed on each of a series of numbers in a
long ......, roundoff error can become important, especially if
division by a small number ever occurs.
 6.6 Exercises 53

10. The secant method is a root-finding algorithm that uses a suc-

cession of roots of .......... lines to better approximate a root of
a function f .
11. There are many different methods that can be used to approx-
imate .......... to a differential equation.
12. Absolute error is the .......... between the measured or inferred
value of a quantity and its actual value.
7
Selected Topics in Differential Equation

7.1 Differential Equation in a Nutshell

1
In the calculus, you studied various methods by which you could
differentiate the elementary functions. Equations which involve
variables and their derivatives are called differential equations.
Let f (x) define a function of x on an interval I : a ≤ x ≤ b. By
an ordinary differential equation we mean an equation involving x,
the function f (x) and one or more of its derivatives. The order of a
differential equation is the order of the highest derivative involved
in the equation.
Let y = f (x) define y as a function of x on an interval I : a ≤
x ≤ b. We say that the function f (x) is an explicit solution or
simply a solution of an ordinary differential equation involving x,
f (x), and its derivatives, if it satisfies the equation for every x in
I, i. e., if we replace y by f (x) , y 0 by f 0 (x), y 00 by f 00 (x), · · · , y (n)
by f (n) (x), the differential equation reduces to an identity in x.
A relation f (x, y) = 0 will be called an implicit solution of the
differential equation
F (x, y, y 0 , . . . , y (n) ) = 0. (7.1)
on an interval I : a ≤ x ≤ b, if

1
This section has been quoted from:
Morris. Tenenbaum, and Harry Pollard. Ordinary differential equations: an el-
ementary textbook for students of mathematics, engineering, and the sciences.
Dover Publications, 1963.
56 7 Selected Topics in Differential Equation

1. It defines y as an implicit function of x on I, i.e., if there exists

a function g(x) defined on I such that f [x, g(x)] = 0 for every
x in I, and if
2. g(x) satisfies (7.1), i.e., if
F [x, g(x), g 0 (x), . . . , g (n) (x)] = 0

for every x in I.

7.2 The General Solution of a Differential Equation

We assume at the outset that you have understood clearly the

material of the previous lesson so that when we say “solve a differ-
ential equation” or “find a solution of a differential equation,” or
“the solution of a differential equation is,” you will know what is
meant. Or if we omit intervals for which a function or a differential
equation is defined, we expect that you will be able to fill in this
omission yourself.
Two conclusions seem to stem from these examples. First, if a
differential equation has a solution, it has infinitely many solutions.
Second, if the differential equation is of the first order, its solution
contains one arbitrary constant; if of the second order, its solution
contains two arbitrary constants; if of the nth order, its solution
contains n arbitrary constants.
It is customary to call a solution which contains n constants
c1 , c2 , . . . , cn an n-parameter family of solutions, and to refer to
the constants c1 to cn as parameters.
We shall now show you how to find the differential equation
when its n-parameter family of solutions is known. You must bear
in mind that although the family will contain the requisite number
of n arbitrary constants, the nth order differential equation whose
solution it is, contains no such constants. In solving problems of
this type, therefore, these constants must be eliminated. Unfortu-
nately a standard method of eliminating these constants is not
always the easiest to use. There are frequently simpler methods
which cannot be standardized and which will depend on your own
ingenuity.
 7.4 Initial Value Problem 57

7.3 Isogonal and orthogonal Trajectories

When two curves intersect in a plane, the angle between them is
defined to be the angle made by their respective tangents drawn at
their point of intersection. Since these lines determine two angles,
it is customary to specify the particular one desired by stating
from which tangent line we are to proceed in a counterclockwise
direction to reach the other.
A curve which cuts every member of a given 1-parameter family
of curves in the same angle is called an isogonal trajectory of the
family.
If two 1-parameter families have the property that every mem-
ber of one family cuts every member of the other family in the same
angle, then each family may be said to be a 1-parameter family of
isogonal trajectories of the other, i. e., the curves of either family
are isogonal trajectories of the other.
An interesting problem is to find a family of isogonal trajecto-
ries that makes a predetermined angle with a given 1-parameter
family of curves. If we call y10 the slope of a curve of a given 1-
parameter family, y 0 the slope of an isogonal trajectory of the fam-
ily, and α their angle of intersection measured from the tangent
line with slope y 0 to the tangent line with slope y10 , then
yl0 − y 0
tan α =
1 + y 0 y10
A curve which cuts every member of a given 1-parameter family
of curves in a 90◦ angle is called an orthogonal trajectory of the
family. If two 1-parameter families have the property that every
member of one family cuts every member of the other family in a
right angle, then each family may be said to be an orthogonal tra-
jectory of the other. Orthogonal trajectory problems are of special
interest since they occur in many physical fields.

7.4 Initial Value Problem

2
In the field of differential equations, an initial value problem is an
ordinary differential equation together with specified value, called
2
This Section has been quoted from:
http://en.wikipedia.org/wiki/Initial value problem
58 7 Selected Topics in Differential Equation

the initial condition, of the unknown function at a given point in

the domain of the solution. In physics or other sciences, modeling
a system frequently amounts to solving an initial value problem;
in this context, the differential equation is an evolution equation
specifying how, given initial conditions, the system will evolve with
time.
An initial value problem is a differential equation

y 0 (t) = f (t, y(t)) with f : R × R → R

together with a point in the domain of f
(t0 , y0 ) ∈ R × R,

called the initial condition.

A solution to an initial value problem is a function y that is a
solution to the differential equation and satisfies
y(t0 ) = y0 .

This statement subsumes problems of higher order, by interpreting

y as a vector. For derivatives of second or higher order, new vari-
ables (elements of the vector y) are introduced. More generally,
the unknown function y can take values on infinite dimensional
spaces, such as Banach spaces or spaces of distributions. For a
large class of initial value problems, the existence and uniqueness
of a solution can be demonstrated.
The Picard-Lindelöf theorem guarantees a unique solution on
some interval containing t0 if f and its partial derivative ∂f
∂y
are con-
tinuous on a region containing t0 and y0 . The proof of this theorem
proceeds by reformulating the problem as an equivalent integral
equation. The integral can be considered an operator which maps
one function into another, such that the solution is a fixed point of
the operator. The Banach fixed-point theorem is then invoked to
show that there exists a unique fixed point, which is the solution
of the initial value problem.
An older proof of the Picard-Lindelöf theorem constructs a se-
quence of functions which converge to the solution of the inte-
gral equation, and thus, the solution of the initial value problem.
Such a construction is sometimes called “Picard’s method” or “the
 7.5 Boundary Value Problem 59

method of successive approximations”. This version is essentially

a special case of the Banach fixed point theorem.
Hiroshi Okamura obtained a necessary and sufficient condition
for the solution of an initial value problem to be unique. This
condition has to do with the existence of a Lyapunov function for
the system.
In some situations, the function f is not of class C 1 , or even
Lipschitz, so the usual result guaranteeing the local existence of
a unique solution does not apply. The Peano existence theorem
however proves that even for f merely continuous, solutions are
guaranteed to exist locally in time; the problem is that there is
no guarantee of uniqueness. An even more general result is the
Carathéodory existence theorem, which proves existence for some
discontinuous functions f .

7.5 Boundary Value Problem

3
A boundary value problem is a differential equation together with
a set of additional restraints, called the boundary conditions. A so-
lution to a boundary value problem is a solution to the differential
equation which also satisfies the boundary conditions.
Boundary value problems arise in several branches of physics
as any physical differential equation will have them. Problems in-
volving the wave equation, such as the determination of normal
modes, are often stated as boundary value problems. A large class
of important boundary value problems are the Sturm-Liouville
problems. The analysis of these problems involves the eigenfunc-
tions of a differential operator.
To be useful in applications, a boundary value problem should
be well posed. This means that given the input to the problem
there exists a unique solution, which depends continuously on the
input. Much theoretical work in the field of partial differential
equations is devoted to proving that boundary value problems
arising from scientific and engineering applications are in fact well-
posed.

3
This section has been quoted from:
http://en.wikipedia.org/wiki/Boundary value problem
60 7 Selected Topics in Differential Equation

Among the earliest boundary value problems to be studied is

the Dirichlet problem, of finding the harmonic functions (solutions
to Laplace’s equation); the solution was given by the Dirichlet’s
principle.
A more mathematical way to picture the difference between an
initial value problem and a boundary value problem is that an ini-
tial value problem has all of the conditions specified at the same
value of the independent variable in the equation (and that value
is at the lower boundary of the domain, thus the term “initial”
value). On the other hand, a boundary value problem has con-
ditions specified at the extremes of the independent variable. For
example, if the independent variable is time over the domain [0, 1],
an initial value problem would specify a value of y(t) and y 0 (t) at
time t = 0, while a boundary value problem would specify values
for y(t) at both t = 0 and t = 1.
If the problem is dependent on both space and time, then in-
stead of specifying the value of the problem at a given point for
all time the data could be given at a given time for all space. For
example, the temperature of an iron bar with one end kept at ab-
solute zero and the other end at the freezing point of water would
be a boundary value problem.
If the boundary gives a value to the normal derivative of the
problem then it is a Neumann boundary condition. For example,
if there is a heater at one end of an iron rod, then energy would
be added at a constant rate but the actual temperature would not
be known. If the boundary gives a value to the problem then it is
a Dirichlet boundary condition. For example, if one end of an iron
rod is held at absolute zero, then the value of the problem would
be known at that point in space. If the boundary has the form of
a curve or surface that gives a value to the normal derivative and
the problem itself then it is a Cauchy boundary condition.
Aside from the boundary condition, boundary value problems
are also classified according to the type of differential operator
involved. For an elliptic operator, one discusses elliptic boundary
value problems. For an hyperbolic operator, one discusses hyper-
bolic boundary value problems. These categories are further sub-
divided into linear and various nonlinear types.
 7.6 Exercises 61

7.6 Exercises

1. Translate the following sentences.

A partial differential equation (or briefly a PDE) is a mathe-
matical equation that involves two or more independent variables,
an unknown function (dependent on those variables), and partial
derivatives of the unknown function with respect to the indepen-
dent variables. The order of a partial differential equation is the
order of the highest derivative involved. A solution (or a partic-
ular solution) to a partial differential equation is a function that
solves the equation or, in other words, turns it into an identity
when substituted into the equation. A solution is called general if
it contains all particular solutions of the equation concerned.
The term exact solution is often used for second- and higher-
order nonlinear PDEs to denote a particular solution Partial dif-
ferential equations are used to mathematically formulate, and thus
aid the solution of, physical and other problems involving functions
of several variables, such as the propagation of heat or sound, fluid
flow, elasticity, electrostatics, electrodynamics, etc.

2. Write down other forms of the following words.

Noun Verb adjective adverb

initiate
general
involve
explicitly
omit
elimination
intersecting
uniqueness
evolution
62 7 Selected Topics in Differential Equation

3. Use the correct form of the word.

1. The main idea behind the Laplace Transformation is that we
can solve an equation (or system of equations) containing dif-
ferential and integral terms by (transform) .......... the equation
in “t-space” to one in “s-space”.
2. The Laplace transform is (invert) .......... on a large class of
functions.
3. The meaning of the integral depends on types of functions of
interest. A necessary condition for existence of the integral is
that f must be locally (integral) ...........on [0, ∞).
4. Two integrable functions have the same Laplace transform only
if they (different) .......... on a set of Lebesgue measure zero.
5. The Laplace transform has a number of properties that make
it useful for (analysis) .......... linear dynamical systems.
4. Fill gaps with one of the following words
1. string 2. gravity 3. horizontal 4. component
5. stretched 6. flexible 7. initial 8. displacement
9. perfectly 10. location 11. tangential 12. measures
Consider a vertical string of length L that has been tightly
stretched between two points at x = 0 and x = L. Because the
string has been tightly .......... we can assume that the slope of
the displaced string at any point is small. So just what does this
do for us? Let’s consider a point x on the string in its equilibrium
position, i.e. the ..........of the point at t = 0. As the string vibrates
this point will be displaced both vertically and horizontally, how-
ever, if we assume that at any point the slope of the string is small
then the .......... displacement will be very small in relation to the
vertical displacement. This means that we can now assume that at
any point x on the string the displacement will be purely vertical.
So, let’s call this ........u(x, t).
We are going to assume, at least initially, that the string is
not uniform and so the mass density of the ..........,ρ(x) may be a
function of x.
Next, we are going to assume that the string is perfectly
...........This means that the string will have no resistance to bend-
ing. This in turn tells us that the force exerted by the string at
 7.6 Exercises 63

any point x on the endpoints will be ..........to the string itself. This
force is called the tension in the string and its magnitude will be
given by T (x, t).
Finally, we will let Q(x, t) represent the vertical ..........per unit
mass of any force acting on the string. Provided we again assume
that the slope of the string is small the vertical displacement of
the string at any point is then given by,
∂u2
 
∂ ∂u
ρ(x) 2 = T (x, t) + ρ(x)Q(x, t)
∂t ∂x ∂x
This is a very difficult partial differential equation to solve so
we need to make some further simplifications.
First, we’re now going to assume that the string is ..........elas-
tic. This means that the magnitude of the tension, T (x, t), will
only depend upon how much the string stretches near x. Again,
recalling that we’re assuming that the slope of the string at any
point is small this means that the tension in the string will then
very nearly be the same as the tension in the string in its equilib-
rium position. We can then assume that the tension is a constant
value, T (x, t) = T0 .
Further, in most cases the only external force that will act upon
the string is ..........and if the string light enough the effects of
gravity on the vertical displacement will be small and so will also
assume that Q(x, t) = 0. This leads to
∂ 2u ∂ 2u
ρ = T0
∂t2 ∂x2
If we now divide by the mass density and define,
T0
c2 =
ρ
we arrive at the 1-D wave equation,
∂ 2u 2
2∂ u
= c
∂t2 ∂x2
The initial conditions will also be a little different here from what
we saw with the heat equation. Here we have a 2nd order time
derivative and so we’ll also need two initial conditions. At any
64 7 Selected Topics in Differential Equation

point we will specify both the ..........displacement of the string as

well as the initial velocity of the string. The initial conditions are
then,
∂u
u(x, 0) = f (x), (x, 0) = g(x)
∂t
8
Selected Topics in Topology and Geometry

8.1 What is Topology?

1
The concept of topological space grew out of the study of the
real line and Euclidean space and the study of continuous function
on these spaces. Here we define what a topological space is, and
we study a number of ways of constructing a topology on a set as
to make it into a topological space. We also consider some of the
elementary concepts associated with topological spaces. Open and
closed sets, limit points, and continuous functions are introduced
as natural generalizations of the corresponding ideas for the real
line and Euclidean space.
The definition of a topological space that is now standard was
a long time in being formulated. Various mathematicians- Fréchet,
Hausdorff, and others- proposed different definitions over a period
of years during the first decades of the twentieth century, but it
took quite a while before mathematicians settled on the one that
seemed most suitable. They wanted, of course, a definition that
was as broad as possible, so that it would include as special cases
all the various examples that were useful in mathematics - Eu-
clidean space, infinite-dimensional Euclidean space, and function
spaces among them - but they also wanted the definition to be
narrow enough that the standard theorems about these familiar
spaces would hold for topological spaces in general. This is always
the problem when one is trying to formulate a new mathemati-
1
This section has been quoted from:
Topology (2nd Edition) James Munkresæ2000 by Prentice Hall Inc. N.J.
66 8 Selected Topics in Topology and Geometry

cal concept, to decide how general its definition should be. The
definition finally settled on may seem a bit abstract, but as you
work through the various ways of constructing topological spaces,
you will get better feeling for what the concept means. A topology
on a set X is a collection τ of subsets of X having the following
properties:
1. ∅ and X are in τ .
2. The union of the elements of any subcollection of τ is in τ .
3. The intersection of the elements of any finite subcollection of τ
is in τ .
A set X for which a topology τ has been specified is called a
topological space.
Properly speaking, a topological space is an ordered pair (X, τ )
consisting of a set X and a topology τ in X, but we often omit
specific mention of τ if no confusion will arise.
If X is a topological space with topology τ , we say that a subset
U of X is an open set of X if U belongs to the collection τ . Using
this terminology, one can say that a topological space is a set X
together with a collection of subsets of X, called open sets, such
that ∅ and X are both open, and such that arbitrary union and
finite intersections of open sets are open.
If X is any set, the collection of all subsets of X is a topology
on X; it is called the discrete topology. The collection consisting
of X and ∅ is also a topology on X; we shall call it the indiscrete
topology, or the trivial topology.
Suppose that τ and τ 0 are two topologies on a given set X. If
τ 0 ⊃ τ , we say that τ 0 is finer than τ ; if τ 0 properly contains τ ,
we say that τ 0 is strictly finer than τ . We also say that τ is coarser
than τ 0 , or strictly coarser, in these two respective situations. We
say τ is comparable topology with τ 0 if either τ ⊃ τ 0 or τ 0 ⊃ τ .
Other terminology is sometimes used for this concept. If τ 0 ⊃ τ ,
some mathematicians would say that τ 0 is larger than τ and τ is
smaller than τ 0 . Many mathematicians use the words “weaker” and
“stronger” in this context. Unfortunately, some of them (particu-
lary analysts) are apt to say that τ 0 is stronger than τ if τ 0 ⊃ τ ,
while others (particulary topologists) are apt to say that τ 0 is
weaker than τ in the same situation.
 8.2 Euler Characteristic 67

If X is a set, a basis for a topology on X is a collection B of

subsets of X (called basis elements) such that
1. For each x ∈ X, there is at least one basis element B containing
x.
2. If x belongs to the intersection of two basis elements B1 and
B2 , then there is a basis element B3 containing x such that
B3 ⊂ B1 ∩ B2 .
if B satisfies these two conditions, then we define the topology
τ generated by B as follows: A subset U of X is said to be open
in X (that is, to be an element of τ ) if for each x ∈ X, there is a
basis element B ∈ B such that x ∈ B and B ⊂ U . Note that each
basis element is itself an element of τ .
Lt X and Y be topological spaces. A function f : X → Y is said
to be continuous if for each open subset V of Y , the set f −1 (V ) is
an open set of X.
Recall that f −1 (V ) is the set of all points x of X for which
f (x) ∈ V , it is empty if V does not intersect the image set f (x)
of f .
Continuity of a function depends not only upon the function
f , but also on the topologies for its domain and range. if we wish
to emphasize this fact, we can say that f is continuous relative to
specific topologies X and Y .

8.2 Euler Characteristic

2
The Euler characteristic is one of the most useful topological in-
variants. Moreover, we find the prototype of the algebraic approach
to topology in it. To avoid unnecessary complication, we restrict
ourselves to points, lines and surfaces in R3 . A polyhedron is a geo-
metrical object surrounded by faces. The boundary of two faces is
an edge and two edges meet at a vertex. We extend the definition
of a polyhedron a bit to include polygons and the boundaries of
polygons, lines or points. We call the faces, edges and vertices of
a polyhedron simplexes (or simplices). Note that the boundary of
two simplexes is either empty or another simplex. (For example,
2
The rest of this chapter has been quoted from: Nakahara, Mikio. Geometry, topol-
ogy and physics. CRC Press, 2003.
68 8 Selected Topics in Topology and Geometry

the boundary of two faces is an edge.) Formal definitions of a sim-

plex and a polyhedron in a general number of dimensions will be
given later. We are now ready to define the Euler characteristic of
a figure in R3 .
Definition 8.1 Let X be a subset of R3 , which is homeomor-
phic to a polyhedron K. Then the Euler characteristic χ(X) of X
is defined by
χ(X) = (number of verticies in K) − (number of edges in K)
+ (number of faces in K). (8.1)

The reader might wonder if χ(X) depends on the polyhedron

K or not. The following theorem due to Poincaré and Alexander
guarantees that it is, in fact, independent of the polyhedron K.
Theorem 8.1 (Poincaré-Alexander) The Euler characteristic
χ(X) is independent of the polyhedron K as long as K is homeo-
morphic to X.
Examples are in order. The Euler characteristic of a point is
χ(·) = 1 by definition. The Euler characteristic of a line is χ(−) =
2−1 = 1, since a line has two vertices and an edge. For a triangular
disc, we find χ(triangle) = 3 − 3 + 1 = 1. An example which is
a bit non-trivial is the Euler characteristic of S1 . The simplest
polyhedron which is homeomorphic to S1 is made of three edges
of a triangle. Then χ(S1 ) = 3 − 3 = 0. Similarly, the sphere S2
is homeomorphic to the surface of a tetrahedron, hence χ(S2 ) =
4 − 6 + 4 = 2. It is easily seen that S2 is also homeomorphic to the
surface of a cube. Using a cube to calculate the Euler characteristic
of S2 , we have χ(S2 ) = 8 − 12 + 6 = 2, in accord with theorem
8.1. Historically this is the conclusion of Eulers theorem: if K is
any polyhedron homeomorphic to S2 , with v vertices, e edges and
f two-dimensional faces, then v − e + f = 2.

8.3 Homology Groups

The mathematical structures underlying homology groups are

finitely generated Abelian groups. Let H be a subgroup of G. We
say x, y ∈ G are equivalent if x − y ∈ H, and write x ∼ y. Clearly
 8.3 Homology Groups 69

∼ is an equivalence relation. The equivalence class to which x be-

longs is denoted by [x]. Let G/H be the quotient space. The group
operation + in G naturally induces the group operation + in G/H
by [x] + [y] = [x + y].
Definition 8.2 If G is finitely generated by r linearly indepen-
dent elements, G is called a free Abelian group of rank r.
Let us recall how the Euler characteristic of a surface is calcu-
lated. We first construct a polyhedron homeomorphic to the given
surface, then count the numbers of vertices, edges and faces. The
Euler characteristic of the polyhedron, and hence of the surface, is
then given by equation (8.1). We abstract this procedure so that
we may represent each part of a figure by some standard object.
We take triangles and their analogues in other dimensions, called
simplexes, as the standard objects. By this standardization, it be-
comes possible to assign to each figure Abelian group structures.

8.3.1 Simplexes

Simplexes are building blocks of a polyhedron. A 0-simplex hp0 i is

a point, or a vertex, and a 1-simplex hp0 p1 i is a line, or an edge.
A 2-simplex hp0 p1 p2 i is defined to be a triangle with its interior
included and a 3-simplex hp0 p1 p2 p3 i is a solid tetrahedron. It is
common to denote a 0-simplex without the bracket; hp0 i may be
also written as p0 . It is easy to continue this construction to any
r-simplex hp0 p1 . . . pr i. Note that for an r-simplex to represent an
r-dimensional object, the vertices pi must be geometrically inde-
pendent, that is, no (r − 1)-dimensional hyperplane contains all the
r + 1 points. Let p0 , . . . , pr be points geometrically independent in
Rm where m ≥ r. The r-simplex hp0 , . . . , pr i is expressed as
( r r
)
X X
σ r = x ∈ Rm |x = ci pi , ci ≥ 0, ci = 1 .
i=0 i=0

(c0 , . . . , cr ) is called the barycentric coordinate of x.

Let K be a set of finite number of simplexes in Rm . If these sim-
plexes are nicely fitted together, K is called a simplicial complex.
By “nicely” we mean:
1. an arbitrary face of a simplex of K belongs to K, that is, if
σ ∈ K and σ 0 ≤ σ then σ 0 ∈ K; and
70 8 Selected Topics in Topology and Geometry

2. if σ and σ 0 are two simplexes of K, the intersection σ ∩ σ 0 is

either empty or a common face of σ and σ 0 , that is, if σ, σ 0 ∈ K
then either σ ∩ σ 0 = ∅ or σ ∩ σ 0 ≤ σ and σ ∩ σ 0 ≤ σ 0 .
Let X be a topological space. If there exists a simplicial complex
K and a homeomorphism f : |K| → X × X is said to be triangu-
lable and the pair (K, f ) is called a triangulation of X. Given a
topological space X, its triangulation is far from unique.

Oriented Simplexes

We may assign orientations to an r -simplex for r ≥ 1. Instead

of h. . .i for an unoriented simplex, we will use (. . .) to denote an
oriented simplex. The symbol σr is used to denote both types of
simplex. An oriented 1-simplex σ1 = (p0 p1 ) is a directed line seg-
ment traversed in the direction p0 → p1 . Now (p0 p1 ) should be
distinguished from (p1 p0 ).
Definition 8.3 The r-chain group Cr (K) of a simplicial com-
plex K is a free Abelian group generated by the oriented r-simplexes
of K. If r > dim K, Cr (K) is defined to be 0. An element of Cr (K)
is called an r-chain.
Before we define the cycle group and the boundary group, we
need to introduce the boundary operator. Let us denote the bound-
ary of an r-simplex σr by ∂r σr . ∂r should be understood as an
operator acting on σr to produce its boundary.
Definition 8.4 If c ∈ Cr (K) satisfies ∂r c = 0, c is called an
r-cycle. The set of r-cycles Zr (K) is a subgroup of Cr (K) and is
called the r-cycle group.
Let K be an n-dimensional simplicial complex. The rth homol-
ogy group Hr (K),0 ≤ r ≤ n, associated with K is defined by
Hr (K) ≡ Zr (K)/Br (K).
If necessary, we define Hr (K) = 0 for r > n or r < 0. If we
want to stress that the group structure is defined with integer
coefficients, we write Hr (K; Z).
 8.4 Exercises 71

8.4 Exercises

1. Translate the following text.

Euclidean geometry is a mathematical well-known system at-
tributed to the Greek mathematician Euclid of Alexandria. Eu-
clid’s text Elements was the first systematic discussion of geom-
etry. It has been one of the most influential books in history, as
much for its method as for its mathematical content. The method
consists of assuming a small set of intuitively appealing axioms,
and then proving many other propositions ( theorems) from those
axioms. Although many of Euclid’s results had been stated by
earlier Greek mathematicians, Euclid was the first to show how
these propositions could be fitted together into a comprehensive
deductive and logical system.
The Elements begin with plane geometry, still often taught in
secondary school as the first axiomatic system and the first exam-
ples of formal proof. The Elements goes on to the solid geometry
of three dimensions, and Euclidean geometry was subsequently ex-
tended to any finite number of dimensions. Much of the Elements
states results of what is now called number theory, proved using
geometrical methods.
For over two thousand years, the adjective “Euclidean” was un-
necessary because no other sort of geometry had been conceived.
Euclid’s axioms seemed so intuitively obvious that any theorem
proved from them was deemed true in an absolute sense. Many
other consistent formal geometries are now known, the first ones
being discovered in the early 19th century. It also is no longer taken
for granted that Euclidean geometry describes physical space. An
implication of Einstein’s theory of general relativity is that Eu-
clidean geometry is only a good approximation to the properties
of physical space if the gravitational field is not too strong.
72 8 Selected Topics in Topology and Geometry

2. Write down other forms of the following words.

Noun Verb adjective adverb

orientation
cyclically
structure
simplicial
simulate
variation
infinitely
triangulable
verification

3. Use the correct form of the word.

To the ancients, the parallel postulate seemed less obvious than
the others; (verify) .......... it physically would require us to inspect
two lines to check that they never intersected, even at some very
(distance) ..........point, and this inspection could potentially take
an infinite amount of time. Euclid himself seems to have considered
it as being (quality) .......... different from the others, as evidenced
by the organization of the Elements: the first 28 propositions he
presents are those that can be proved without it.
Many geometers tried in vain to prove the fifth postulate from
the first four. By 1763 at least 28 (differ) .......... proofs had been
published, but all were found to be incorrect. In fact the parallel
postulate cannot be proved from the other four: this was shown
in the 19th century by the (construct) .......... of alternative ( non-
Euclidean) systems of geometry where the other axioms are still
true but the parallel postulate is replaced by a (conflict) ..........ax-
iom.
One distinguishing aspect of these systems is that the three an-
gles of a triangle do not add to 180: in hyperbolic geometry the
sum of the three angles is always less than 180 and can approach
 8.4 Exercises 73

zero, while in elliptic geometry it is greater than 180. If the paral-

lel postulate is dropped from the list of axioms without (replace)
..........the result is the more general geometry called absolute ge-
ometry.
4. Fill gaps with one of the following words

1. another 2. inevitably 3. following 4. right

5. subtracted 6. centre 7. greater 8. segment
9. sphere 10. straight 11. concepts 12. derived
Following a precedent set in the Elements, Euclidean geometry
has been exposited as an axiomatic system, in which all theorems
(“true statements”) are ........... from a finite number of axioms.
Near the beginning of the first book of the Elements, Euclid gives
five postulates (axioms):
1. Any two points can be joined by a .......... line.
2. Any straight line can be extended indefinitely in a straight line.
3. Given any straight line segment, a circle can be drawn having
the segment as radius and one endpoint as .........
4. All ........... angles are congruent.
5. Parallel postulate. If two lines intersect a third in such a way
that the sum of the inner angles on one side is less than two
right angles, then the two lines ............ must intersect each
other on that side if extended far enough.
These axioms invoke the following ..........: point, straight line
segment and line, side of a line, circle with radius and centre,
right angle, congruence, inner and right angles, sum. The following
verbs appear: join, extend, draw, intersect. The circle described in
postulate 3 is tacitly unique. Postulates 3 and 5 hold only for plane
geometry; in three dimensions, postulate 3 defines a ..........
The Elements also include the .......... five “common notions”:
1. Things that equal the same thing also equal one ...........
2. If equals are added to equals, then the wholes are equal.
3. If equals are .......... from equals, then the remainders are equal.
4. Things that coincide with one another equal one another.
5. The whole is .......... than the part.
9
Selected Topics in Discrete Mathematics

9.1 The Language of Logic

1
Logic is the study of the principles and techniques of reason-
ing. Logic plays a central role in the development of every area of
learning, especially in mathematics and computer science. Com-
puter scientists, for example, employ logic to develop programming
languages and to establish the correctness of programs. Electronics
engineers apply logic in the design of computer chips.
A declarative sentence that is either true or false, but not both,
is a proposition (or a statement), which we will denote by the low-
ercase letter p, q, r, s, or t. The variables p, q, r, s, or t are boolean
variables (or logic variables).
Consider the sentence, This sentence is false. It is certainly a
valid declarative sentence, but is it a proposition? To answer this,
assume the sentence is true. But the sentence says it is false. This
contradicts our assumption. On the other hand, suppose the sen-
tence is false. This implies the sentence is true, which again contra-
dicts our assumption. Thus, if we assume that the sentence is true,
it is false; and if we assume that it is false, it is true. It is a mean-
ingless and self-contradictory sentence, so it is not a proposition,
but a paradox.
A compound proposition is formed by combining two or more
simple propositions called components. Compound propositions
can be formed in several ways. The conjunction of two arbitrary
1
This chapter has been quoted from:
Koshy, Thomas. Discrete mathematics with applications. Elsevier, 2004.
76 9 Selected Topics in Discrete Mathematics

propositions p and q, denoted by p ∧ q, is the proposition p and

q. It is formed by combining the propositions using the word and,
called a connective. If both p and q are true, then p ∧ q is true;
otherwise it is false. A second way of combining two propositions
p and q is by using the connective or. The resulting proposition
p or q is the disjunction of p and q and is denoted by p ∨ q. The
disjunction of two propositions is true if at least one component
is true; it is false only if both components are false.
The negation of a proposition p is It is not the case that p,
denoted by ¬p. You may read ¬p as the negation of p or simply
not p. If a proposition p is true, then ¬p is false; if p is false, then
¬p is true.
Two propositions p and q can be combined to form statements
of the form: If p, then q. Such a statement is an implication, denoted
by p ⇒ q. Since it involves a condition, it is also called a conditional
statement. The component p is the hypothesis (or premise) of the
implication and q the conclusion. To construct the truth table for
an implication If p, then q, we shall think of it as a conditional
promise. If you do p, then I promise to do q. If the promise is kept,
we consider the implication true; if the promise is not kept, we
consider it false. We can use this analogy to construct the truth
table, as shown below. Consider the following implication:
p ⇒ q: If you wax my car, then I will pay you $25.
If you wax my car (p true) and if I pay you $25 (q true), then
the implication is true. If you wax my car (p true) and if I do
not pay you $25 (q false), then the promise is violated; hence the
implication is false. What if you do not wax my car (p false)? Then
I may give you $25 (being generous!) or not. (So q may be true or
false). In either case, my promise has not been tested and hence
has not been violated. Consequently, the implication has not been
proved false. If it is not false, it must be true. In other words, if p
is false, the implication p ⇒ q is true by default. (If p is false, the
implication is said to be vacuously true.)

9.2 Combinatorics
At the beginning of the 18th century, the following problem was
proposed:
 9.3 Graphs and Trees 77

A secretary had written n different letters and addressed n

different envelopes for them. Unfortunately, a wind storm
mixed up the letters and the envelopes. After the storm was
over, each letter was placed in an envelope. In how many
ways can the letters be placed in the envelopes, so that
every letter is in a wrong envelope?
This problem has several variations. A variation involves n
guests checking in their coats at the coat room of a fancy restau-
rant. In how many ways can the attendant return their coats, so
no person gets the right coat?
Before answering these problems, we make the following defini-
tion. A permutation of n distinct items a1 , a2 , · · · , an in which no
item ai appears in its original position i for any i, 1 ≤ i ≤ n, is
called a derangement. We would like to find the number of possible
derangements of n items.
Let Dn denote the number of derangements of n distinct items.
To find an explicit formula for Dn , first we derive a recurrence
relation satisfied by Dn as:
Dn = (n − 1)(Dn−1 + Dn−2 ), n ≥ 2
where D0 = 1 and D1 = 0.
It can be proven that the number of derangements of n distinct
elements is
(−1)n
 
1 1 1 1
Dn = n! 1 − + − + − · · · + ,n ≥ 0
1! 2! 3! 4! n!
It is shown in calculus that
∞
−1
X (−1)n
e =
n=0
n!
so the expression inside the brackets in the formula is the sum of
the first (n + 1) terms in the expansion of e−1 .

9.3 Graphs and Trees

Graph theory, a fascinating branch of mathematics, has numerous
applications to such diverse areas as computer science, engineer-
ing, linguistics, and management science, as well as the natural
78 9 Selected Topics in Discrete Mathematics

and social sciences. Like many important discoveries, graph the-

ory grew out of an interesting physical problem, the celebrated
Königsberg Bridge Puzzle. The outstanding Swiss mathematician
Leonhard Euler solved the puzzle in 1736, thus laying the founda-
tion for graph theory and earning his title as the father of graph
theory.
A graph (or undirected graph) G consists of a nonempty finite
set V of points (called vertices or nodes) and a set E of unordered
pairs of elements in V (called edges). The graph G is the ordered
pair (V, E) : G = (V, E). An edge connecting the vertices u and v
is denoted by {u, v},u − v, or some label. Geometrically, edges are
denoted by arcs or line segments.
An edge emanating from and terminating at the same vertex
is a loop. Parallel edges have the same vertices. A simple graph
contains no loops or parallel edges.
Two vertices v and w in a graph are adjacent, if an edge runs
between them; if a loop occurs at v, v is adjacent to itself. An
isolated vertex is not adjacent to any vertex. Adjacent edges have
a common vertex. An edge is incident with a vertex v if v is an
endpoint of the edge.
The degree of a vertex v in a graph is the number of edges
meeting at v; it is denoted by deg(v). Clearly, a vertex v is isolated
if deg(v) = 0. In addition, a loop at v contributes two to its degree.
The adjacency matrix of a graph with n vertices v1 , v2 , . . . , vn is
an n × n matrix A = (aij ), where aij = number of edges from vi to
vj . Because every edge in a graph is undirected, aij = aji for every
i and j, so the adjacency matrix of every graph is symmetric.
A subgraph of a graph G = (V, E) is a graph G1 = (V1 , E1 )
where V1 ⊆ V and E1 ⊆ E. A simple graph with an edge between
every two distinct vertices is a complete graph. A complete graph
with n vertices is denoted by Kn . The cycle graph Cn of length
n(≥ 3) consists of n vertices v1 , . . . , vn and edges {vi , vi+1 }, where
1 ≤ i ≤ n and vn+1 = v0
If the vertex set V of a simple graph G = (V, E) can be parti-
tioned into two disjoint (nonempty) sets V1 and V2 , so every edge
in G is incident with a vertex in V1 and a vertex V2 , then G is
bipartite. Let G be a bipartite graph with |V1 | = m and |V2 | = n.
 9.3 Graphs and Trees 79

If an edge runs between every vertex in V1 and V2 , G is a complete

bipartite graph, denoted by Km,n .
A simple graph in which each edge e is assigned a positive real
number w is a weighted graph.
A graph is planar if it can be drawn in the plane, so its edges
meet only at the vertices. Such a drawing is a planar representation
of the graph. Let G be a connected planar graph with e edges and
v vertices. Let r be the number of regions formed by a planar
representation of G. Then r = e − u + 2 (Euler’s formula).
Trees are the most important class of graphs and they make
fine modeling tools. A connected, acyclic graph is a tree.
• A connected graph is a tree if and only if there is a unique,
simple path between any two vertices.
• A connected graph with n vertices is a tree if and only if it has
exactly n − 1 edges.
A subgraph H of a connected graph G is a spanning tree of G if
H is a tree containing every vertex of G. Every connected graph
has a spanning tree.
Let G be a connected weighted graph. The weight of a spanning
tree of G is the sum of the weights of its edges. A minimal spanning
tree of G weighs the least. Several algorithms can find a minimal
spanning tree T of a connected weighted graph G. Two of them
are Kruskal’s and Prim’s.
A tree with a root is a rooted tree. Rooted trees are drawn
with the root at the top, especially in computer science; they grow
downward.
Let T be a rooted tree with root v0 . Let v0 − v1 − · · · − vn−1 − vn
be the path from v0 to vn . Then:
• vn−1 is the parent of vn .
• vn is a child of vn−1 .
• Vertices with the same parent are siblings.
• The vertices v0 , v1 , . . . , vn−1 are ancestors of vn .
• The descendants of a vertex v are those vertices for which v is
an ancestor.
• A vertex with no children is a leaf or a terminal vertex.
• A vertex that is not a leaf is an internal vertex.
80 9 Selected Topics in Discrete Mathematics

• The subtree of T rooted at v consists of v, its descendants, and

all edges incident with them.

9.4 Recursion

Recursion is an elegant and powerful problem-solving technique,

used extensively in both discrete mathematics and computer sci-
ence.
Leonardo Fibonacci, the most outstanding Italian mathemati-
cian of the Middle Ages, proposed the following problem around
1202:
Suppose there are two newborn rabbits, one male and the
other female. Find the number of rabbits produced in a year
if:
• Each pair takes one month to become mature.
• Each pair produces a mixed pair every month, from the
second month.
• No rabbits die.
Suppose, for convenience, that the original pair of rabbits was born
on January 1. They take a month to become mature. So there is
still only one pair on February 1. On March 1, they are 2 months
old and produce a new mixed pair, a total of two pairs. Continuing
like this, there will be three pairs on April 1, five pairs on May 1,
and so on.
The numbers 1, 1, 2, 3, 5, 8, · · · are Fibonacci numbers. They
have a fascinating property: Any Fibonacci number, except the
first two, is the sum of the two immediately preceding Fibonacci
numbers. (At the given rate, there will be 144 pairs of rabbits on
December 1.)
This yields the following recursive definition of the nth Fi-
bonacci number Fn :
F1 = F2 = 1 ←− initial conditions
Fn = Fn−1 + Fn−2 ←− recurrence relation
Solving the recurrence relation for a function f means finding an
explicit formula for Fn . The iterative method of solving it involves
two steps:
 9.5 Exercises 81

• Apply the recurrence formula iteratively and look for a pattern

to predict an explicit formula.
• Use induction to prove that the formula does indeed hold for
every possible value of the integer n.
Unfortunately, the iterative method illustrated above can be
applied to only a small and simple class of recurrence relations. The
recurrence relation Fn = Fn−1 + Fn−2 is linear and homogeneous. If
it has a nonzero solution of the form cαn , then cαn = cαn−1 +cαn−2 .
Since cα 6= 0, this yields α2 = α + 1; that is, α2 − α − 1 = 0, so
α must be a solution of the characteristic
√
equation
√
x2 − x − 1 = 0,
and its solutions are a α = 1+2 5 and β = 1−2 5 . You may verify
α + β = 1 and αβ = −1. The general solution is Fn = Aαn + Bβ n .
To find A and B, we have:
F1 = Aα + Bβ = 1
F2 = Aα2 + Bβ 2 = 1
Solving these two equations, we get (Verify):
1 1
A = √ and B = − √ .
5 5

9.5 Exercises
1. Translate the following sentences.
Certain graph problems deal with finding a path between two
vertices such that each edge is traversed exactly once, or finding a
path between two vertices while visiting each vertex exactly once.
These paths are better known as Euler path and Hamiltonian path
respectively. The Euler path problem was first proposed in the
1700s.
• An Euler path is a path that uses every edge of a graph exactly
once. It starts and ends at different vertices.
• An Euler circuit is a circuit that uses every edge of a graph
exactly once. It starts and ends at the same vertex.
There are simple criteria for determining whether a multigraph
has a Euler path or a Euler circuit. For any multigraph to have a
Euler circuit, all the degrees of the vertices must be even.
82 9 Selected Topics in Discrete Mathematics

Theorem: A connected multigraph (and simple graph) with

at least two vertices has a Euler circuit if and only if each of its
vertices has an even degree.
Proof of the above statement is that every time a circuit passes
through a vertex, it adds twice to its degree. Since it is a circuit,
it starts and ends at the same vertex, which makes it contribute
one degree when the circuit starts and one when it ends. In this
way, every vertex has an even degree.
Theorem: A connected multigraph (and simple graph) has an
Euler path but not an Euler circuit if and only if it has exactly
two vertices of odd degree.
The proof is an extension of the proof given above. Since a path
may start and end at different vertices, the vertices where the path
starts and ends are allowed to have odd degrees.

2. Write down other forms of the following words.

Noun Verb adjective adverb

logic
iteratively
contradict
homogeneous
weigh
paradox
correctly
inductive
recurrence

3. Use the correct form of the word.

1. Mathematical proof is an argument we give logically to (valid)

.......a mathematical statement.
 9.5 Exercises 83

2. When we want to (proof) ......... a conditional statement p im-

plies q, we assume that p is true, and follow (imply) ..........to
get to show that q is then true.
3. Symmetric and anti-symmetric (relate).......... are not opposite
because a relation can contain both the properties or may not.
4. A relation is an (equal) .......... relation if it is reflexive, sym-
metric, and transitive.
5. A relation R on a set A is called (refelect) ......... if (a, a) ∈ R
holds for every element a ∈ A.

4. Fill gaps with one of the following words

1. exactly 2. route 3. path 4. diverse

5. electronic 6. octahedron 7. district 8. nodes
9. determine 10. unique 11. complete 12. visiting

A Hamiltonian cycle is a closed loop on a graph where every

node (vertex) is visited exactly once. A loop is just an edge that
joins a node to itself; so a Hamiltonian cycle is a path travel-
ing from a point back to itself, .......... every node en route. If a
graph with more than one node (i.e. a non-singleton graph) has
a Hamiltonian cycle, we call it a Hamiltonian graph. There isnt
any equation or general trick to finding out whether a graph has a
Hamiltonian cycle; the only way to .......... this is to do a complete
and exhaustive search, going through all the options.
Every .......... graph with more than two vertices is a Hamilto-
nian graph. This follows from the definition of a complete graph: an
undirected, simple graph such that every pair of nodes is connected
by a .......... edge. The graph of every platonic solid is a Hamilto-
nian graph. So the graph of a cube, a tetrahedron, an .........., or an
icosahedron are all Hamiltonian graphs with Hamiltonian cycles.
A graph with n vertices (where n > 3) is Hamiltonian if the sum
of the degrees of every pair of non-adjacent vertices is n or greater.
This is known as Ores theorem.
A search for Hamiltonian cycles is not just a fun game for the
afternoon off. It has real applications in such ........ fields as com-
puter graphics, .......... circuit design, mapping genomes, and op-
erations research. For instance, when mapping genomes scientists
84 9 Selected Topics in Discrete Mathematics

must combine many tiny fragments of genetic code (“read”, they

are called), into one single genomic sequence (a “superstring”).
This can be done by finding a Hamiltonian cycle or Hamiltonian
.........., where each of the reads are considered nodes in a graph
and each overlap (place where the end of one read matches the
beginning of another) is considered to be an edge. In a much less
complex application of exactly the same math, school districts use
Hamiltonian cycles to plan the best .......... to pick up students
from across the ........... Here, students may be considered ..........,
the paths between them edges, and the bus wishes to travel a route
that will pass each students house .......... once.
10
Selected Topics in Optimization

10.1 The Origins of Operations Research

1
The roots of operations research (OR) can be traced back many
decades, when early attempts were made to use a scientific ap-
proach in the management of organizations. However, the be-
ginning of the activity called operations research has generally
been attributed to the military services early in World War II.
Because of the war effort, there was an urgent need to allocate
scarce resources to the various military operations and to the ac-
tivities within each operation in an effective manner. Therefore,
the British and then the U.S. military management called upon a
large number of scientists to apply a scientific approach to dealing
with this and other strategic and tactical problems. In effect, they
were asked to do research on (military) operations. These teams of
scientists were the first OR teams. By developing effective methods
of using the new tool of radar, these teams were instrumental in
winning the Air Battle of Britain. Through their research on how
to better manage convoy and antisubmarine operations, they also
played a major role in winning the Battle of the North Atlantic.
Similar efforts assisted the Island Campaign in the Pacific.
When the war ended, the success of OR in the war effort spurred
interest in applying OR outside the military as well. As the indus-
trial boom following the war was running its course, the problems

1
This chapter has been quoted from:
Introduction to Operations Research Seventh Edition, Hillier and Lieberman 2001
The McGraw-Hill Companies.
86 10 Selected Topics in Optimization

caused by the increasing complexity and specialization in organiza-

tions were again coming to the forefront. It was becoming appar-
ent to a growing number of people, including business consultants
who had served on or with the OR teams during the war, that
these were basically the same problems that had been faced by
the military but in a different context. By the early 1950s, these
individuals had introduced the use of OR to a variety of organiza-
tions in business, industry, and government. The rapid spread of
OR soon followed.

10.2 Linear Programming

The development of linear programming has been ranked among

the most important scientific advances of the mid-20th century,
and we must agree with this assessment. Its impact since just
1950 has been extraordinary. Today it is a standard tool that has
saved many thousands or millions of dollars for most companies or
businesses of even moderate size in the various industrialized coun-
tries of the world; and its use in other sectors of society has been
spreading rapidly. A major proportion of all scientific computation
on computers is devoted to the use of linear programming. Dozens
of textbooks have been written about linear programming, and
published articles describing important applications now number
in the hundreds.
Linear programming uses a mathematical model to describe the
problem of concern. The adjective linear means that all the mathe-
matical functions in this model are required to be linear functions.
The word programming does not refer here to computer program-
ming; rather, it is essentially a synonym for planning. Thus, lin-
ear programming involves the planning of activities to obtain an
optimal result, i.e., a result that reaches the specified goal best
(according to the mathematical model) among all feasible alterna-
tives.
Although allocating resources to activities is the most com-
mon type of application, linear programming has numerous other
important applications as well. In fact, any problem whose mathe-
matical model fits the very general format for the linear program-
ming model is a linear programming problem. Furthermore, a re-
 10.2 Linear Programming 87

markably efficient solution procedure, called the simplex method,

is available for solving linear programming problems of even enor-
mous size. These are some of the reasons for the tremendous im-
pact of linear programming in recent decades.
Let us begin with some basic terminology and notation. The
key terms are resources and activities, where m denotes the num-
ber of different kinds of resources that can be used and n denotes
the number of activities being considered. Some typical resources
are money and particular kinds of machines, equipment, vehicles,
and personnel. Examples of activities include investing in partic-
ular projects, advertising in particular media, and shipping goods
from a particular source to a particular destination. In any ap-
plication of linear programming, all the activities may be of one
general kind (such as any one of these three examples), and then
the individual activities would be particular alternatives within
this general category.
The model poses the problem in terms of making decisions
about the levels of the activities, so x1 , x2 , . . . , xn are called the
decision variables. The values of cj , bi , and aij (for i = 1, 2, . . . , m
and j = 1, 2, . . . , n) are the input constants for the model. The cj ,
bi , and aij are also referred to as the parameters of the model.
we can now formulate the mathematical model for this general
problem of allocating resources to activities. In particular, this
model is to select the values for x1 , x2 , . . . , xn so as to
Maximize Z = c1 x1 + c2 x2 + · · · + cn xn ,
subject to the restrictions

a11 x1 +a12 x2 + · · · +a1n xn ≤ b1

a21 x1 +a22 x2 + · · · +a2n xn ≤ b2
.. .. .. .. ..
. . . . .
am1 x1 +am2 x2 + · · · +amn xn ≤ bm

and
x1 ≥ 0, x2 ≥ 0, . . . , xn ≥ 0.
Common terminology for the linear programming model can now
be summarized. The function being maximized, c1 x1 + c2 x2 + · · · +
88 10 Selected Topics in Optimization

cn xn , is called the objective function. The restrictions normally are

referred to as constraints. The first m constraints (those with a
function of all the variables ai1 x1 + ai2 x2 + · · · + ain xn ≤ bi on
the left-hand side) are sometimes called functional constraints (or
structural constraints). Similarly, the xj ≥ 0 restrictions are called
nonnegativity constraints (or nonnegativity conditions).
You may be used to having the term solution mean the final
answer to a problem, but the convention in linear programming
(and its extensions) is quite different. Here, any specification of
values for the decision variables (x1 , x2 , . . . , xn ) is called a solu-
tion, regardless of whether it is a desirable or even an allowable
choice. Different types of solutions are then identified by using an
appropriate adjective.
A feasible solution is a solution for which all the constraints are
satisfied. An infeasible solution is a solution for which at least one
constraint is violated. The feasible region is the collection of all
feasible solutions. An optimal solution is a feasible solution that
has the most favorable value of the objective function. The most
favorable value is the largest value if the objective function is to
be maximized, whereas it is the smallest value if the objective
function is to be minimized.

10.3 The Transportation and Assignment Problems

We continue to broaden our horizons here by discussing two partic-

ularly important (and related) types of linear programming prob-
lems. One type, called the transportation problem, received this
name because many of its applications involve determining how
to optimally transport goods. However, some of its important ap-
plications (e.g., production scheduling) actually have nothing to
do with transportation. The second type, called the assignment
problem, involves such applications as assigning people to tasks.
Although its applications appear to be quite different from those
for the transportation problem, we shall see that the assignment
problem can be viewed as a special type of transportation problem.
 10.3 The Transportation and Assignment Problems 89

10.3.1 Transportation Problem

To describe the general model for the transportation problem, we

need to use terms that are considerably less specific than those
for the components of the prototype example. In particular, the
general transportation problem is concerned (literally or figura-
tively) with distributing any commodity from any group of supply
centers, called sources, to any group of receiving centers, called
destinations, in such a way as to minimize the total distribution
cost.
The requirements assumption: Each source has a fixed
supply of units, where this entire supply must be distributed
to the destinations. (We let si denote the number of units
being supplied by source i, for i = 1, 2, . . . , m.) Similarly,
each destination has a fixed demand for units, where this
entire demand must be received from the sources. (We let
dj denote the number of units being received by destination
j, for j = 1, 2, . . . , n.)
This assumption that there is no leeway in the amounts to be
sent or received means that there needs to be a balance between
the total supply from all sources and the total demand at all des-
tinations.
The feasible solutions property: A transportation prob-
lem will have feasible solutions if and only if
m
X m
X
si = dj
i=1 j=1

In some real problems, the supplies actually represent maximum

amounts (rather than fixed amounts) to be distributed. Similarly,
in other cases, the demands represent maximum amounts (rather
than fixed amounts) to be received. Such problems do not quite fit
the model for a transportation problem because they violate the
requirements assumption. However, it is possible to reformulate the
problem so that they then fit this model by introducing a dummy
destination or a dummy source to take up the slack between the
actual amounts and maximum amounts being distributed. We will
90 10 Selected Topics in Optimization

illustrate how this is done with two examples at the end of this
section.
This reference to a unit cost implies the following basic assump-
tion for any transportation problem.
The cost assumption: The cost of distributing units
from any particular source to any particular destination
is directly proportional to the number of units distributed.
Therefore, this cost is just the unit cost of distribution times
the number of units distributed. (We let cij denote this unit
cost for source i and destination j.)
The only data needed for a transportation problem model are the
supplies, demands, and unit costs. These are the parameters of the
model.
The model: Any problem (whether involving transporta-
tion or not) fits the model for a transportation problem if
it can be described completely in terms of a parameter and
it satisfies both the requirements assumption and the cost
assumption. The objective is to minimize the total cost of
distributing the units.

10.3.2 Assignment Problem

The assignment problem is a special type of linear programming

problem where assignees are being assigned to perform tasks. For
example, the assignees might be employees who need to be given
work assignments. Assigning people to jobs is a common applica-
tion of the assignment problem. However, the assignees need not
be people. They also could be machines, or vehicles, or plants, or
even time slots to be assigned tasks. The first example below in-
volves machines being assigned to locations, so the tasks in this
case simply involve holding a machine. A subsequent example in-
volves plants being assigned products to be produced.
To fit the definition of an assignment problem, these kinds of
applications need to be formulated in a way that satisfies the fol-
lowing assumptions.
1. The number of assignees and the number of tasks are the same.
(This number is denoted by n.)
 10.4 Exercises 91

2. Each assignee is to be assigned to exactly one task.

3. Each task is to be performed by exactly one assignee.
4. There is a cost cij associated with assignee i (i = 1, 2, . . . , n)
performing task j ( j = 1, 2, . . . , n).
5. The objective is to determine how all n assignments should be
made to minimize the total cost.
Any problem satisfying all these assumptions can be solved ex-
tremely efficiently by algorithms designed specifically for assign-
ment problems.
The first three assumptions are fairly restrictive . Many poten-
tial applications do not quite satisfy these assumptions. However,
it often is possible to reformulate the problem to make it fit. For
example, dummy assignees or dummy tasks frequently can be used
for this purpose.

10.4 Exercises

1. Translate the following text.

According to the number of time periods considered in the
model, optimization models can be classified as static (single
time period) or multistage (multiple time periods). Even when
all relationships are linear, if several time periods are incorpo-
rated in the model the resulting linear program could become pro-
hibitively large for solution by standard computational methods.
Fortunately, in most of these cases, the problem exhibits some
form of special structure that can be adequately exploited by the
application of special types of mathematical programming meth-
ods. Dynamic programming is one approach for solving multistage
problems. Further, there is a considerable research effort underway
today, in the field of large-scale linear programming, to develop
special algorithms to deal with multistage problems.
92 10 Selected Topics in Optimization

2. Write down other forms of the following words.

Noun Verb adjective adverb

proportion
specially
assume
restrictive
allocate
complexity
optimally
scientific
solution

3. Use the correct form of the word.

Another important way of (classify).......... optimization mod-
els refers to the behavior of the parameters of the model. If the
parameters are known constants, the optimization model is said
to be deterministic. If the parameters are specified as uncertain
quantities, whose values are (characterization) ..........by probabil-
ity distributions, the optimization model is said to be stochastic.
If some of the parameters are allowed to (variation) ..........system-
atically, and the changes in the optimum solution corresponding
to changes in those parameters are determined, the optimization
model is said to be parametric. (general).........., stochastic and
parametric mathematical programming give rise to much more
difficult problems than deterministic mathematical programming.
Although important theoretical and practical contributions have
been made in the areas of stochastic and parametric program-
ming, there are still no (effect)..........general procedures that cope
with these problems. Deterministic linear programming, however,
can be (efficient)..........applied to very large problems of up to
5000 rows and an almost unlimited number of variables. More-
over, in linear programming, sensitivity analysis and parametric
 10.4 Exercises 93

programming can be conducted effectively after (obtain) ..........

the deterministic optimum solution.
4. Fill gaps with one of the following words
1. amount 2. program- 3. class 4. models
ming
5. efficiently 6. structure 7. variables 8. satisfies
9. progress 10. optimal 11. discrete 12. continuous
A way of classifying optimization .......... deals with the behavior
of the variables in the .......... solution. If the variables are allowed
to take any value that .......... the constraints, the optimization
model is said to be continuous. If the variables are allowed to take
on only .......... values, the optimization model is called integer or
discrete. Finally, when there are some integer variables and some
continuous .......... in the problem, the optimization model is said
to be mixed. In general, problems with integer variables are sig-
nificantly more difficult to solve than those with .......... variables.
Network models are a .......... of linear programming models that
are an exception to this rule, as their special .......... results in inte-
ger optimal solutions. Although significant ......... has been made
in the general area of mixed and integer linear .........., there is
still no algorithm that can .......... solve all general medium-size
integer linear programs in a reasonable .......... of time though, for
special problems, adequate computational techniques have been
developed.
11
Selected Topics in Analysis

11.1 Compact space

1
Intuitively speaking, a space is said to be compact if whenever
one takes an infinite number of “steps” in the space, eventually one
must get arbitrarily close to some other point of the space. Thus,
while disks and spheres are compact, infinite lines and planes are
not, nor is a disk or a sphere with a missing point. In the case
of an infinite line or plane, one can set off making equal steps in
any direction without approaching any point, so that neither space
is compact. In the case of a disk or sphere with a missing point,
one can move towards the missing point without approaching any
point within the space. This demonstrates that the disk or sphere
with a missing point are not compact either.
Compactness generalizes many important properties of closed
and bounded intervals in the real line; that is, intervals of the
form [a, b] for real numbers a and b. For instance, any continuous
function defined on a compact space into an ordered set (with
the order topology) such as the real line is bounded. Thus, what
is known as the extreme value theorem in calculus generalizes to
compact spaces. In this fashion, one can prove many important
theorems in the class of compact spaces, that do not hold in the
context of non-compact ones.
Various definitions of compactness may apply, depending on
the level of generality. A subset of Euclidean space in particular is
1
This chapter has been quoted from:
A Course in Functional Analysis, John B. Conway , Springer-Verlag New York
Berlin Heidelberg Tokyo, 1985.
96 11 Selected Topics in Analysis

called compact if it is closed and bounded. This implies, by the

Bolzano-Weierstrass Theorem, that any infinite sequence from the
set has a subsequence that converges to a point in the set. This
puts a fine point on the idea of taking “steps” in a space. Various
equivalent notions of compactness, such as sequential compactness
and limit point compactness, can be developed in general metric
spaces.
In general topological spaces, however, the different notions of
compactness are not equivalent, and the most useful notion of
compactness originally called bicompactness involves families of
open sets that “cover” the space in the sense that each point of
the space must lie in some set contained in the family. Specifically,
a topological space is compact if, whenever a collection of open
sets covers the space, some subcollection consisting only of finitely
many open sets also covers the space. That this form of compact-
ness holds for closed and bounded subsets of Euclidean space is
known as the Heine-Borel Theorem. Compactness, when defined
in this manner, often allows one to take information that is known
locally - in a neighborhood of each point of the space - and to ex-
tend it to information that holds globally throughout the space.
An example of this phenomenon is Dirichlet’s theorem, to which
it was originally applied by Heine, that a continuous function on a
compact interval is uniformly continuous: here continuity is a local
property of the function, and uniform continuity the corresponding
global property.

11.2 Hilbert space

A Hilbert space is the abstraction of the finite-dimensional Eu-

clidean spaces of geometry. Its properties are very regular and
contain few surprises, though the presence of an infinity of dimen-
sions guarantees a certain amount of surprise. Historically, it was
the properties of Hilbert spaces that guided mathematicians when
they began to generalize.
Definition 11.1 If X is a vector space over F, a semi-inner
product on X is a function u : X × X → F such that for all α, β
in F and x, y, z in H, the following are satisfied:
 11.3 Orthogonal Sets of Vectors and Basis 97

(a) u(αx + βy, z) = αu(x, z) + βu(y, z),

(b) u(x, αy + βz) = αu(x, y) + βu(x, z),
(c) u(x, x) ≥ 0,
(d) u(x, y) = u(y, x).
An inner product on H is a semi-inner product that also satisfies
the following: If u(x, x) = 0, then x = 0. An inner product here
will be denoted by hx, yi = u(x, y).
A Hilbert space is a vector space H over F together with an
inner product h., .i such that relative to the metric d(x, y) = ||x −
y|| induced by the norm, H is a complete metric space.
The greatest advantage of a Hilbert space is its underlying con-
cept of orthogonality .
Definition 11.2 If H is a Hilbert space and f, g ∈ H , then f
and g are orthogonal if hf, gi = 0. In symbols, f ⊥g. If A, B ⊆ H
, then A⊥B if f ⊥g for every f in A and g in B.
Definition 11.3 If H is any vector space over F and A ⊆ H,
then A is a convex set if for any x and y in A and 0 ≤ t ≤ 1,
tx + (1 − t)y ∈ A.
Note that{tx + (1 − t)y : 0 ≤ t ≤ 1} is the straight-line segment
joining x and y. So a convex set is a set A such that if x and y in
A, the entire line segment joining x and y is contained in A.

11.3 Orthogonal Sets of Vectors and Basis

It cab be shown that, as in Euclidean space, each Hilbert space

can be coordinatized. The vehicle for introducing the coordinates is
an orthogonal basis.
Definition 11.4 An orthogonal subset of a Hilbert space H is
a subset E having the properties: (a) for e in E, ||e|| = 1; (b) if
el , e2 ∈ E and e1 6= e2 , then e1 ⊥ e2 .
A basis for H is a maximal orthogonal set.
Actually, the sum that appears in the following can be given
a better interpretation- a mathematically precisePone that will be
useful later. The question is, what is meant by {hi : i ∈ I} if
hi ∈ H and I is an infinite, possibly uncountable set? Let F be
98 11 Selected Topics in Analysis

the collection of all finite subsets of I and order F by inclusion ,

so F becomes a directed set. For each F in F, define
X
hF = {hi : i ∈ F }.

Since this is a finite sum, hF is a well-defined element of H. Now

{hF : F ∈ F}is a net in H. P
With the notation above, the sum {hi : i ∈ I} converges
if the net {hF : F ∈ F} converges; the value of the sum is the
limit of the net. If H = F, the definition above gives meaning to
an uncountable sum of scalars. If the set I is countable, then this
definition of convergent sum is not the usualPone. That is, if {hn }
is a sequence in H then the convergence of {hn : n ∈ N} is not
X∞
equivalent to the convergence of hn .
n=1

11.4 Isomorphic Hilbert Spaces

Every mathematical theory has its concept of isomorphism. In

topology there is homeomorphism and homotopy equivalence; al-
gebra calls them isomorphisms. The basic idea is to define a map
which preserves the basic structure of the spaces in the category.
Definition 11.5 If H and K are Hilbert spaces, an isomor-
phism between H and K is a linear surjection U : H → K such
that
hU h, U gi = hh, gi.
for all h, g in H. In this case H and K are said to be isomorphic.
It is easy to see that if U : H → K is an isomorphism, then so
is U −1 : K → H. Similar such arguments show that the concept of
isomorphic is an equivalence relation on Hilbert spaces. It is also
certain that this is the correct equivalence relation since an inner
product is the essential ingredient for a Hilbert space and isomor-
phic Hilbert spaces have the same inner product. One might object
that completeness is another essential ingredient in the definition
of a Hilbert space. So it is! However, this too is preserved by an
isomorphism. An isometry between metric spaces is a map that
preserves distance.
 11.5 Operators on Hilbert Space 99

Note that an isometry between metric spaces maps Cauchy se-

quences into Cauchy sequences. Thus an isomorphism also pre-
serves completeness. That is, if an inner product space is isomor-
phic to a Hilbert space, then it must be complete.
A word about terminology. Many call what we call an isomor-
phism a unitary operator. We shall define a unitary operator as
a linear transformation U : H → K that is a surjective isometry.
That is, a unitary operator is an isomorphism whose range coin-
cides with its domain. This may seem to be a minor distinction,
and in many ways it is. But experience has taught me that there
is some benefit in making such a distinction, or at least in being
aware of it.
Theorem 11.1 Two Hilbert spaces are isomorphic if and only
if they have the same dimension.

11.5 Operators on Hilbert Space

There is a marked contrast here between Hilbert spaces and the

Banach spaces. Essentially all of the information about the geom-
etry of Hilbert space is contained in the preceding chapter. The
geometry of Banach space lies in darkness and has attracted the at-
tention of many talented research mathematicians. However, the
theory of linear operators (linear transformations) on a Banach
space has very few general results, whereas Hilbert space opera-
tors have an elegant and well-developed general theory. Indeed,
the reason for this dichotomy is related to the opposite status of
the geometric considerations. Questions concerning operators on
Hilbert space don’t necessitate or imply any geometric difficulties.
Proposition 11.1 Let H and K be Hilbert spaces and A : H →
K a linear transformation. The following statements’ are equiva-
lent.
(a) A is continuous.
(b) A is continuous at 0.
(c) A is continuous at some point.
(d) There is a constant c > 0 such that ||Ah|| ≤ c||h|| for all h in
H.
100 11 Selected Topics in Analysis

By virtue of the preceding proposition, d(A, B) = ||A − B||

defines a metric on B(H, K). So it makes sense to consider B(H, K)
as a metric space.
Definition 11.6 If H and K are Hilbert spaces, a function
u : H × K :→ F is a sesquilinear form if for h, g in H , k, f in K,
and α, β in F,
(a) u(αh + βg, k) = αu(h, k) + βu(g, k);
(b) u(h, αk + βf ) = αu(h, k) + βu(h, f ).
The prefix sesqui is used because the function is linear in one
variable but (for F = C) only conjugate linear in the other. (Sesqui
means One-and-a-half. )
a sesquilinear form is bounded, if there is a constant M such
that |u(h, k)| ≤ M ||h||||k|| for all h in H and k in K. The constant
M is called a bound for u.
If K : L2 (µ) → L2 (µ) is defined by
Z
(Kf )(x) = k(x, y)f (y)dtµ(y),

then K is a bounded linear operator. The operator described above

is called an integral operator and the function k is called its kernel.
Definition 11.7 If A ∈ B(H, K), then the unique operator B
in B(K, H) satisfying u(h, k) = (Ah, k) = (h, Bk) is called the
adjoint of A and is denoted by B = A∗ .

11.6 Exercises

1. Translate the following text.

The complex numbers are the field C of numbers of the form
x + iy, where x and y are real numbers
√ and i is the imaginary unit
equal to the square root of −1, −1. When a single letter z =
x + iy is used to denote a complex number, it is sometimes called
an “affix”. In component notation, z = x+iy can be written (x, y).
The field of complex numbers includes the field of real numbers as
a subfield.
Complex numbers are useful abstract quantities that can be
used in calculations and result in physically meaningful solutions.
 11.6 Exercises 101

However, recognition of this fact is one that took a long time for
mathematicians to accept. For example, John Wallis wrote, “These
Imaginary Quantities (as they are commonly called) arising from
the Supposed Root of a Negative Square (when they happen) are
reputed to imply that the Case proposed is impossible”.
Through the Euler formula, a complex number z = x + iy may
be written in “phasor” form

z = |z|(cos θ + i sin θ) = |z|eiθ .

Here, |z| is known as the complex modulus (or sometimes the com-
plex norm) and theta is known as the complex argument or phase.
Historically, the geometric representation of a complex number
as simply a point in the plane was important because it made
the whole idea of a complex number more acceptable. In partic-
ular,“imaginary” numbers became accepted partly through their
visualization.
Unlike real numbers, complex numbers do not have a natural
ordering, so there is no analog of complex-valued inequalities. This
property is not so surprising however when they are viewed as
being elements in the complex plane, since points in a plane also
lack a natural ordering.
2. Write down other forms of the following words.

Noun Verb adjective adverb

coordinates
equivalently
linearize
sequential
necessitate
operation
uniformly
compact
distinction
102 11 Selected Topics in Analysis

3. Use the correct form of the word.

1. Show that if we (calculation) ..........with symbols x + iy, where
x and y are real numbers, according to the usual rules for
adding and multiplying numbers and in addition use i2 = −1,
then all the (require) .......... for a field are satisfied.
2. After (introduction) .......... complex numbers we can, for any
given real number, find a real or complex number whose square
is the given number.
3. A map is called conformal if it preserves angles and their (ori-
ent) ..........
4. The angle between two smooth curves in an oriented surface at
a point of (intersect) ..........of the curves is the angle between
the tangents at the point.
5. So far we have said nothing about (converge) .......... on the
boundary of the circle of convergence. There is a good reason
for this; nothing much can be said in general. One can have
divergence at every point of the circle, convergence at some
points and (diverge).......... at others or one can have absolute
convergence at every point of the circle.
4. Fill gaps with one of the following words
1. properties 2. integrals 3. fundamental 4. complex
5. numbers 6. analytic 7. applications 8. derivative
9. regions 10. nice 11. derivative 12. extremely
Complex analysis is the study of complex .......... together with
their derivatives, manipulation, and other ........... Complex analy-
sis is an .......... powerful tool with an unexpectedly large number
of practical .......... to the solution of physical problems. Contour
integration, for example, provides a method of computing difficult
.......... by investigating the singularities of the function in ..........
of the complex plane near and between the limits of integration.
The key result in complex analysis is the Cauchy integral theo-
rem, which is the reason that single-variable complex analysis has
so many .......... results. A single example of the unexpected power
of .......... analysis is Picard’s great theorem, which states that an
........ function assumes every complex number, with possibly one
 11.6 Exercises 103

exception, infinitely often in any neighborhood of an essential sin-

gularity!
A .......... result of complex analysis is the Cauchy-Riemann
equations, which give the conditions a function must satisfy in
order for a complex generalization of the .........., the so-called
complex derivative, to exist. When the complex .......... is defined
“everywhere” the function is said to be analytic.
12
Selected Topics in Number Theory

12.1 What is number theory?

1
Number theory (or arithmetic or higher arithmetic in older usage)
is a branch of pure mathematics devoted primarily to the study of
the integers and integer-valued functions. German mathematician
Carl Friedrich Gauss (1777-1855) said,“Mathematics is the queen
of the sciences-and number theory is the queen of mathematics”
Number theorists study prime numbers as well as the properties
of objects made out of integers (for example, rational numbers)
or defined as generalizations of the integers (for example, algebraic
integers). Number theory seeks to understand the properties of
integer systems in spite of their apparent complexity.

12.2 Main subdivisions

12.2.1 Elementary tools

The term elementary generally denotes a method that does not use
complex analysis. For example, the prime number theorem was
first proven using complex analysis in 1896, but an elementary
proof was found only in 1949 by Erdős and Selberg.[76] The term
is somewhat ambiguous: for example, proofs based on complex
Tauberian theorems (for example, Wiener-Ikehara) are often seen
as quite enlightening but not elementary, in spite of using Fourier
1
This chapter has been quoted from:
https://en.wikipedia.org/wiki/Number theory#Elementary tools
106 12 Selected Topics in Number Theory

analysis, rather than complex analysis as such. Here as elsewhere,

an elementary proof may be longer and more difficult for most
readers than a non-elementary one.

12.2.2 Analytic number theory

Some subjects generally considered to be part of analytic number

theory, for example, sieve theory, are better covered by the second
rather than the first definition: some of sieve theory, for instance,
uses little analysis, yet it does belong to analytic number theory.
The following are examples of problems in analytic number the-
ory: the prime number theorem, the Goldbach conjecture (or the
twin prime conjecture, or the Hardy-Littlewood conjectures), the
Waring problem and the Riemann hypothesis. Some of the most
important tools of analytic number theory are the circle method,
sieve methods and L-functions (or, rather, the study of their prop-
erties). The theory of modular forms (and, more generally, auto-
morphic forms) also occupies an increasingly central place in the
toolbox of analytic number theory.

12.2.3 Algebraic number theory

An algebraic number is any complex number that is a solution to

some polynomial equation f (x) = 0 with rational coefficients; for
example, every solution x of x5 + 11 2
x3 − 7x2 + 9 = 0 (say) is an
algebraic number. Fields of algebraic numbers are also called al-
gebraic number fields, or shortly number fields. Algebraic number
theory studies algebraic number fields. Thus, analytic and alge-
braic number theory can and do overlap: the former is defined by
its methods, the latter by its objects of study.
It could be argued that the simplest kind of number fields (viz.,
quadratic fields) were already studied by Gauss, as the discussion
of quadratic forms in Disquisitiones arithmeticae can be restated
in terms of ideals and norms in quadratic fields.
√ (A quadratic field
consists of all numbers of the form a + b d, where a and b are
rational numbers and d is a fixed rational number whose square
root is not rational.) For that matter, the 11th-century chakravala
method amounts-in modern terms-to an algorithm for finding the
units of a real quadratic number field. However, neither Bhāskara
nor Gauss knew of number fields as such.
 12.2 Main subdivisions 107

12.2.4 Diophantine geometry

The central problem of Diophantine geometry is to determine

when a Diophantine equation has solutions, and if it does, how
many. The approach taken is to think of the solutions of an equa-
tion as a geometric object.
For example, an equation in two variables defines a curve in
the plane. More generally, an equation, or system of equations,
in two or more variables defines a curve, a surface or some other
such object in n-dimensional space. In Diophantine geometry, one
asks whether there are any rational points (points all of whose
coordinates are rationals) or integral points (points all of whose
coordinates are integers) on the curve or surface. If there are any
such points, the next step is to ask how many there are and how
they are distributed. A basic question in this direction is if there
are finitely or infinitely many rational points on a given curve (or
surface).

12.2.5 Probabilistic number theory

Much of probabilistic number theory can be seen as an important

special case of the study of variables that are almost, but not quite,
mutually independent. For example, the event that a random in-
teger between one and a million be divisible by two and the event
that it be divisible by three are almost independent, but not quite.
It is sometimes said that probabilistic combinatorics uses the
fact that whatever happens with probability greater than 0 must
happen sometimes; one may say with equal justice that many ap-
plications of probabilistic number theory hinge on the fact that
whatever is unusual must be rare. If certain algebraic objects (say,
rational or integer solutions to certain equations) can be shown
to be in the tail of certain sensibly defined distributions, it fol-
lows that there must be few of them; this is a very concrete non-
probabilistic statement following from a probabilistic one.

12.2.6 Computational number theory

While the word algorithm goes back only to certain readers of al-
Khwārizmī, careful descriptions of methods of solution are older
than proofs: such methods (that is, algorithms) are as old as any
108 12 Selected Topics in Number Theory

recognisable mathematics-ancient Egyptian, Babylonian, Vedic,

Chinese-whereas proofs appeared only with the Greeks of the clas-
sical period.
An interesting early case is that of what we now call the Eu-
clidean algorithm. In its basic form (namely, as an algorithm for
computing the greatest common divisor) it appears as Proposi-
tion 2 of Book VII in Elements, together with a proof of correct-
ness. However, in the form that is often used in number theory
(namely, as an algorithm for finding integer solutions to an equa-
tion ax + by = c, or, what is the same, for finding the quantities
whose existence is assured by the Chinese remainder theorem) it
first appears in the works of Aryabhata (5th-6th century CE) as
an algorithm called kuttaka (“pulverise”), without a proof of cor-
rectness.

12.3 Modular arithmetic

2
In mathematics, modular arithmetic is a system of arithmetic
for integers, where numbers “wrap around” when reaching a cer-
tain value, called the modulus. The modern approach to modular
arithmetic was developed by Carl Friedrich Gauss in his book Dis-
quisitiones Arithmeticae, published in 1801.
A familiar use of modular arithmetic is in the 12-hour clock,
in which the day is divided into two 12-hour periods. If the time
is 7:00 now, then 8 hours later it will be 3:00. Usual addition
would suggest that the later time should be 7 + 8 = 15, but this
is not the answer because clock time “wraps around” every 12
hours. Because the hour number starts over after it reaches 12,
this is arithmetic modulo 12. In terms of the definition below, 15
is congruent to 3 modulo 12, so the (military) time called “15:00”
has the equivalent clock form“3:00”.

12.3.1 Definition of congruence relation

Modular arithmetic can be handled mathematically by introduc-

ing a congruence relation on the integers that is compatible with
2
This section has been quoted from:
https://en.wikipedia.org/wiki/Modular arithmetic
 12.3 Modular arithmetic 109

the operations on integers: addition, subtraction, and multiplica-

tion. For a positive integer n, two numbers a and b are said to be
congruent modulo n, if their difference a − b is an integer multi-
ple of n (that is, if there is an integer k such that a − b = kn).
This congruence relation is typically considered when a and b are
integers, and is denoted:

a ≡ b (mod n)
The parentheses mean that (mod n) applies to the entire equation,
not just to the right-hand side (here b). Sometimes, = is used
instead of ≡ ; in this case, if the parentheses are omitted, this
generally means that “mod” denotes the modulo operation applied
to the righthand side, and the equality implies thus that 0 ≤ a < n.
The number n is called the modulus of the congruence.

12.3.2 Applications

In theoretical mathematics, modular arithmetic is one of the foun-

dations of number theory, touching on almost every aspect of its
study, and it is also used extensively in group theory, ring theory,
knot theory, and abstract algebra. In applied mathematics, it is
used in computer algebra, cryptography, computer science, chem-
istry and the visual and musical arts.
A very practical application is to calculate checksums within se-
rial number identifiers. For example, International Standard Book
Number (ISBN) uses modulo 11 (for 10 digit ISBN) or modulo 10
(for 13 digit ISBN) arithmetic for error detection. Likewise, Inter-
national Bank Account Numbers (IBANs), for example, make use
of modulo 97 arithmetic to spot user input errors in bank account
numbers. In chemistry, the last digit of the CAS registry number
(a unique identifying number for each chemical compound) is a
check digit, which is calculated by taking the last digit of the first
two parts of the CAS registry number times 1, the previous digit
times 2, the previous digit times 3 etc., adding all these up and
computing the sum modulo 10.
In cryptography, modular arithmetic directly underpins public
key systems such as RSA and Diffie-Hellman, and provides finite
fields which underlie elliptic curves, and is used in a variety of
110 12 Selected Topics in Number Theory

symmetric key algorithms including Advanced Encryption Stan-

dard (AES), International Data Encryption Algorithm (IDEA),
and RC4. RSA and Diffie-Hellman use modular exponentiation.

12.4 Exercises
1. Translate the following text.
A continued fraction is an expression obtained through an it-
erative process of representing a number as the sum of its inte-
ger part and the reciprocal of another number, then writing this
other number as the sum of its integer part and another reciprocal,
and so on. In a finite continued fraction (or terminated continued
fraction), the iteration/recursion is terminated after finitely many
steps by using an integer in lieu of another continued fraction. In
contrast, an infinite continued fraction is an infinite expression. In
either case, all integers in the sequence, other than the first, must
be positive. The integers ai are called the coefficients or terms of
the continued fraction.
Continued fractions have a number of remarkable properties
related to the Euclidean algorithm for integers or real numbers.
Every rational number pq has two closely related expressions as a
finite continued fraction, whose coefficients ai can be determined
by applying the Euclidean algorithm to (p, q). The numerical value
of an infinite continued fraction is irrational; it is defined from its
infinite sequence of integers as the limit of a sequence of values for
finite continued fractions. Each finite continued fraction of the se-
quence is obtained by using a finite prefix of the infinite continued
fraction’s defining sequence of integers. Moreover, every irrational
number α is the value of a unique infinite continued fraction, whose
coefficients can be found using the non-terminating version of the
Euclidean algorithm applied to the incommensurable values α and
1. This way of expressing real numbers (rational and irrational) is
called their continued fraction representation.
2. Use the correct form of the word.
1. Number theory is a vast and fascinating field of mathematics,
sometimes called “higher arithmetic,” (consist) ..........of the
study of the properties of whole numbers.
 12.4 Exercises 111

2. The Euclidean algorithm is an (effect) .......... method for com-

puting the greatest common divisor of two integers, without
explicitly factoring the two integers.
3. Sum of squares theorems are theorems in additive number
theory concerning the (express) .......... of integers as sums of
squares of other integers.
4. Sum of squares theorems have found various (apply) ..........s
in applied number theory, such as cryptography and integer
factoring algorithms.
5. Arithmetic functions are real- or complex-valued functions de-
fined on the set Z+ of positive integers. They (description)
.......... arithmetic properties of numbers and are widely used
in the field of number theory.

3. Fill gaps with one of the following words

1.geometrical 2. states 3. figurate 4. equation
5. congruences 6. possess 7. factored 8. square
9. dealing 10. residue 11.finding 12. ranges
1. A Diophantine equation is an .......... in which only integer so-
lutions are allowed.
2. Arithmetic is the branch of mathematics .......... with integers
or, more generally, numerical computation.
3. The fundamental theorem of arithmetic ..........that every posi-
tive integer (except the number 1) can be represented in exactly
one way apart from rearrangement as a product of one or more
primes.
4. An abnormal number is a hypothetical number which can be
.......... into primes in more than one way.
5. Interval arithmetic is the arithmetic of quantities that lie within
specified .......... (i.e., intervals) instead of having definite known
values.
6. Modular arithmetic is the arithmetic of .........., sometimes
known informally as ”clock arithmetic.”
7. The word ..........is used in a number of different contexts in
mathematics. Two of the most common uses are the complex
residue of a pole, and the remainder of a congruence.
112 12 Selected Topics in Number Theory

8. A cubic number is a ........... number of the form n3 with n a

positive integer.
9. A square number, also called a perfect ..........., is a figurate
number of the form Sn = n2 , where n is an integer.
10. A figurate number, also known as a figural number, is a number
that can be represented by a regular .......... arrangement of
equally spaced points.
11. A set of real numbers x1 , . . . , xn is said to .......... an integer
relation if there exist integers ai such that a1 x1 + a2 x2 + ... +
an xn = 0, with not all ai = 0.
12. The subset sum problem is the problem of .......... what subset
of a list of integers has a given sum, which is an integer relation
problem where the relation coefficients ai are 0 or 1.
13
Famous problems in Math history

13.1 Fermat’s Last Theorem

1
Pierre de Fermat died in 1665. Today we think of Fermat as a
number theorist, in fact as perhaps the most famous number the-
orist who ever lived. It is therefore surprising to find that Fermat
was in fact a lawyer and only an amateur mathematician. Also
surprising is the fact that he published only one mathematical pa-
per in his life, and that was an anonymous article written as an
appendix to a colleague’s book.
Because Fermat refused to publish his work, his friends feared
that it would soon be forgotten unless something was done about
it. His son, Samuel undertook the task of collecting Fermat’s letters
and other mathematical papers, comments written in books, etc.
with the object of publishing his father’s mathematical ideas. In
this way the famous ‘Last theorem’ came to be published. It was
found by Samuel written as a marginal note in his father’s copy of
Diophantus’s Arithmetica.
Fermat’s Last Theorem states that
xn + y n = z n
has no non-zero integer solutions for x, y and z when n > 2. Fermat
wrote
I have discovered a truly remarkable proof which this margin
is too small to contain.
1
This section has been quoted from:
http://www-groups.dcs.st-and.ac.uk/h̃istory/HistTopics/Fermat’s last theorem.html
114 13 Famous problems in Math history

Fermat almost certainly wrote the marginal note around 1630,

when he first studied Diophantus’s Arithmetica. It may well be
that Fermat realized that his remarkable proof was wrong, how-
ever, since all his other theorems were stated and restated in
challenge problems that Fermat sent to other mathematicians.
Although the special cases of n = 3 and n = 4 were issued as
challenges (and Fermat did know how to prove these) the general
theorem was never mentioned again by Fermat.
In fact in all the mathematical work left by Fermat there is only
one proof. Fermat proves that the area of a right triangle cannot
be a square. Clearly this means that a rational triangle cannot be
a rational square. In symbols, there do not exist integers x, y, z
with x2 + y 2 = z 2 such that xy/2 is a square. From this it is easy
to deduce the n = 4 case of Fermat’s theorem.
It is worth noting that at this stage it remained to prove Fer-
mat’s Last Theorem for odd primes n only. For if there were inte-
gers x, y, z with xn + y n = z n then if n = pq,
(xq )p + (y q )p = (z q )p .
Euler wrote to Goldbach on 4 August 1753 claiming he had a proof
of Fermat’s Theorem when n = 3. However his proof in Algebra
(1770) contains a fallacy and it is far from easy to give an alterna-
tive proof of the statement which has the fallacious proof. There
is an indirect way of mending the whole proof using arguments
which appear in other proofs of Euler so perhaps it is not too
unreasonable to attribute the n = 3 case to Euler.
Euler’s mistake is an interesting one, one which was to have a
bearing on later developments. He needed to find cubes of the form
p2 + 3q 2 and Euler shows that, for any a, b if we put p = a3 − 9ab2 ,
q = 3(a2 b − b3 ) then p2 + 3q 2 = (a2 + 3b2 )3 .
This is true but he then tries to show that, if p2 + 3q 2 is a cube
then an a and b exist such that p and q are as above. His
√ method is
imaginative, calculating with numbers of the form a+b 3. However
numbers of this form do not behave in the same way as the integers,
which Euler did not seem to appreciate.
The next major step forward was due to Sophie Germain. A
special case says that if n and 2n + 1 are primes then xn + y n = z n
implies that one of x, y, z is divisible by n. Hence Fermat’s Last
Theorem splits into two cases.
 13.1 Fermat’s Last Theorem 115

Case 1: None of x, y, z is divisible by n.

Case 2: One and only one of x, y, z is divisible by n.
Sophie Germain proved Case 1 of Fermat’s Last Theorem for all n
less than 100 and Legendre extended her methods to all numbers
less than 197. At this stage, Case 2 had not been proved for even
n = 5 so it became clear that Case 2 was the one on which to
concentrate. Now Case 2 for n = 5 itself splits into two. One of
x, y, z is even and one is divisible by 5. Case 2(i) is when the
number divisible by 5 is even; Case 2(ii) is when the even number
and the one divisible by 5 are distinct.
Case 2(i) was proved by Dirichlet and presented to the Paris
Académie des Sciences in July 1825. Legendre was able to prove
Case 2(ii) and the complete proof for n = 5 was published in
September 1825. In fact Dirichlet was able to complete his own
proof of the n = 5 case with an argument for Case 2(ii) which was
an extension of his own argument for Case 2(i).
In 1832 Dirichlet published a proof of Fermat’s Last Theorem
for n = 14. Of course he had been attempting to prove the n = 7
case but had proved a weaker result. The n = 7 case was finally
solved by Lamé in 1839. It showed why Dirichlet had so much diffi-
culty, for although Dirichlet’s n = 14 proof used similar (but com-
putationally much harder) arguments to the earlier cases, Lamé
had to introduce some completely new methods. Lamé’s proof is
exceedingly hard and makes it look as though progress with Fer-
mat’s Last Theorem to larger n would be almost impossible with-
out some radically new thinking.
The year 1847 is of major significance in the study of Fermat’s
Last Theorem. On 1 March of that year Lamé announced to the
Paris Académie that he had proved Fermat’s Last Theorem. He
sketched a proof which involved factorizing xn + y n = z n into lin-
ear factors over the complex numbers. Lamé acknowledged that
the idea was suggested to him by Liouville. However, Liouville ad-
dressed the meeting after Lamé and suggested that the problem of
this approach was that uniqueness of factorization into primes was
needed for these complex numbers and he doubted if it were true.
Cauchy supported Lamé but, in rather typical fashion, pointed out
that he had reported to the October 1847 meeting of the Académie
an idea which he believed might prove Fermat’s Last Theorem.
116 13 Famous problems in Math history

Much work was done in the following weeks in attempting to

prove the uniqueness of factorization. Wantzel claimed to have
proved it on 15 March but his argument
It is true for n = 2, n = 3 and n = 4 and one easily sees
that the same argument applies for n > 4
was somewhat hopeful. [Wantzel is correct about n = 2 (ordinary
integers), n = 3 (the argument Euler got wrong) and n = 4 (which
was proved by Gauss).]
On 24 May Liouville read a letter to the Académie which settled
the arguments. The letter was from Kummer, enclosing an off-
print of a 1844 paper which proved that uniqueness of factorization
failed but could be ‘recovered’ by the introduction of ideal complex
numbers which he had done in 1846. Kummer had used his new
theory to find conditions under which a prime is regular and had
proved Fermat’s Last Theorem for regular primes. Kummer also
said in his letter that he believed 37 failed his conditions.
By September 1847 Kummer sent to Dirichlet and the Berlin
Academy a paper proving that a prime p is regular (and so Fer-
mat’s Last Theorem is true for that prime) if p does not divide
the numerators of any of the Bernoulli numbers B2 , B4 , . . . , Bp−3 .
The Bernoulli number Bi is defined by
∞
X Bi xi
x
= .
ex − 1 i−0
i!

Kummer shows that all primes up to 37 are regular but 37 is not

regular as 37 divides the numerator of B32 .
The only primes less than 100 which are not regular are 37, 59
and 67. More powerful techniques were used to prove Fermat’s Last
Theorem for these numbers. This work was done and continued to
larger numbers by Kummer, Mirimanoff, Wieferich, Furtwängler,
Vandiver and others. Although it was expected that the number
of regular primes would be infinite even this defied proof. In 1915
Jensen proved that the number of irregular primes is infinite.
Despite large prizes being offered for a solution, Fermat’s Last
Theorem remained unsolved. It has the dubious distinction of be-
ing the theorem with the largest number of published false proofs.
For example over 1000 false proofs were published between 1908
 13.1 Fermat’s Last Theorem 117

and 1912. The only positive progress seemed to be computing re-

sults which merely showed that any counter-example would be very
large. Using techniques based on Kummer’s work, Fermat’s Last
Theorem was proved true, with the help of computers, for n up to
4, 000, 000 by 1993.
In 1983 a major contribution was made by Gerd Faltings who
proved that for every n > 2 there are at most a finite number of
coprime integers x, y, z with xn + y n = z n . This was a major step
but a proof that the finite number was 0 in all cases did not seem
likely to follow by extending Faltings’ arguments.
The final chapter in the story began in 1955, although at this
stage the work was not thought of as connected with Fermat’s
Last Theorem. Yutaka Taniyama asked some questions about el-
liptic curves, i.e. curves of the form y 2 = x3 +ax+b for constants a
and b. Further work by Weil and Shimura produced a conjecture,
now known as the Shimura-Taniyama-Weil Conjecture. In 1986
the connection was made between the Shimura-Taniyama-Weil
Conjecture and Fermat’s Last Theorem by Frey at Saarbrücken
showing that Fermat’s Last Theorem was far from being some
unimportant curiosity in number theory but was in fact related to
fundamental properties of space.
Further work by other mathematicians showed that a counter-
example to Fermat’s Last Theorem would provide a counter-
example to the Shimura-Taniyama-Weil Conjecture. The proof of
Fermat’s Last Theorem was completed in 1993 by Andrew Wiles,
a British mathematician working at Princeton in the USA. Wiles
gave a series of three lectures at the Isaac Newton Institute in
Cambridge, England the first on Monday 21 June, the second on
Tuesday 22 June. In the final lecture on Wednesday 23 June 1993
at around 10.30 in the morning Wiles announced his proof of Fer-
mat’s Last Theorem as a corollary to his main results. Having writ-
ten the theorem on the blackboard he said I will stop here and sat
down. In fact Wiles had proved the Shimura-Taniyama-Weil Con-
jecture for a class of examples, including those necessary to prove
Fermat’s Last Theorem.
This, however, is not the end of the story. On 4 December 1993
Andrew Wiles made a statement in view of the speculation. He
said that during the reviewing process a number of problems had
118 13 Famous problems in Math history

emerged, most of which had been resolved. However one problem

remains and Wiles essentially withdrew his claim to have a proof.
He states
The key reduction of (most cases of) the Taniyama-Shimura
conjecture to the calculation of the Selmer group is correct.
However the final calculation of a precise upper bound for
the Selmer group in the semisquare case (of the symmetric
square representation associated to a modular form) is not
yet complete as it stands. I believe that I will be able to
finish this in the near future using the ideas explained in
my Cambridge lectures.
In March 1994 Faltings, writing in Scientific American, said
If it were easy, he would have solved it by now. Strictly
speaking, it was not a proof when it was announced.
Weil, also in Scientific American, wrote
I believe he has had some good ideas in trying to construct
the proof but the proof is not there. To some extent, proving
Fermat’s Theorem is like climbing Everest. If a man wants
to climb Everest and falls short of it by 100 yards, he has
not climbed Everest.
In fact, from the beginning of 1994, Wiles began to collaborate
with Richard Taylor in an attempt to fill the holes in the proof.
However they decided that one of the key steps in the proof, using
methods due to Flach, could not be made to work. They tried a
new approach with a similar lack of success. In August 1994 Wiles
addressed the International Congress of Mathematicians but was
no nearer to solving the difficulties.
Taylor suggested a last attempt to extend Flach’s method in
the way necessary and Wiles, although convinced it would not
work, agreed mainly to enable him to convince Taylor that it could
never work. Wiles worked on it for about two weeks, then suddenly
inspiration struck.
In a flash, I saw that the thing that stopped it [the extension
of Flach’s method] working was something that would make
another method I had tried previously work.
 13.2 The Four Color Problem 119

On 6 October Wiles sent the new proof to three colleagues

including Faltings. All liked the new proof which was essentially
simpler than the earlier one. Faltings sent a simplification of part
of the proof.
No proof of the complexity of this can easily be guaranteed to be
correct, so a very small doubt will remain for some time. However
when Taylor lectured at the British Mathematical Colloquium in
Edinburgh in April 1995 he gave the impression that no real doubts
remained over Fermat’s Last Theorem.

13.2 The Four Color Problem

2
The Four Colour Conjecture first seems to have been made by
Francis Guthrie. He was a student at University College London
where he studied under De Morgan. After graduating from London
he studied law but by this time his brother Frederick Guthrie
had become a student of De Morgan. Francis Guthrie showed his
brother some results he had been trying to prove about the coloring
of maps and asked Frederick to ask De Morgan about them.
De Morgan was unable to give an answer but, on 23 October
1852, the same day he was asked the question, he wrote to Hamil-
ton in Dublin. De Morgan wrote:
A student of mine asked me today to give him a reason
for a fact which I did not know was a fact - and do not
yet. He says that if a figure be anyhow divided and the
compartments differently colored so that figures with any
portion of common boundary line are differently coloured -
four colours may be wanted, but not more - the following
is the case in which four colors are wanted. Query cannot
a necessity for five or more be invented. ...... If you retort
with some very simple case which makes me out a stupid
animal, I think I must do as the Sphynx did....
Hamilton replied on 26 October 1852 (showing the efficiency of
both himself and the postal service):

2
This section has been quoted from:
http://www-groups.dcs.st-and.ac.uk/h̃istory/HistTopics/The four colour theorem.html
120 13 Famous problems in Math history

I am not likely to attempt your quaternion of color very

soon.
Before continuing with the history of the Four Color Conjec-
ture we will complete details of Francis Guthrie. After practising
as a barrister he went to South Africa in 1861 as a Professor of
Mathematics. He published a few mathematical papers and be-
came interested in botany. A heather (Erica Guthriei) is named
after him.
De Morgan kept asking if anyone could find a solution to
Guthrie’s problem and several mathematicians worked on it. Charles
Peirce in the USA attempted to prove the Conjecture in the 1860’s
and he was to retain a lifelong interest in the problem. Cayley also
learnt of the problem from De Morgan and on 13 June 1878 he
posed a question to the London Mathematical Society asking if the
Four Color Conjecture had been solved. Shortly afterwards Cayley
sent a paper “On the coloring of maps” to the Royal Geographical
Society and in was published in 1879. The paper explains where
the difficulties lie in attempting to prove the conjecture.
On 17 July 1879 Alfred Bray Kempe announced in Nature that
he had a proof of the Four Color Conjecture. Kempe was a Lon-
don barrister who had studied mathematics under Cayley at Cam-
bridge and devoted some of his time to mathematics throughout
his life. At Cayley’s suggestion Kempe submitted the Theorem
to the American Journal of Mathematics where it was published
in 1879. Story read the paper before publication and made some
simplifications. Story reported the proof to the Scientific Associa-
tion of Johns Hopkins University in November 1879 and Charles
Peirce, who was at the November meeting, spoke at the Decem-
ber meeting of the Association of his own work on the Four Color
Conjecture.
Kempe used an argument known as the method of Kempe
chains. If we have a map in which every region is colored red,
green, blue or yellow except one, say X. If this final region X is
not surrounded by regions of all four colors there is a colour left
for X. Hence suppose that regions of all four colours surround X.
If X is surrounded by regions A, B, C, D in order, colored red,
yellow, green and blue then there are two cases to consider.
 13.2 The Four Color Problem 121

(i) There is no chain of adjacent regions from A to C alternately

colored red and green.
(ii) There is a chain of adjacent regions from A to C alternately
colored red and green.
If (i) holds there is no problem. Change A to green, and then
interchange the colour of the red/green regions in the chain joining
A. Since C is not in the chain it remains green and there is now
no red region adjacent to X. Colour X red.
If (ii) holds then there can be no chain of yellow/blue adjacent
regions from B to D. [It could not cross the chain of red/green
regions.] Hence property (i) holds for B and D and we change
colours as above.
Kempe received great acclaim for his proof. He was elected a
Fellow of the Royal Society and served as its treasurer for many
years. He was knighted in 1912. He published two improved ver-
sions of his proof, the second in 1880 aroused the interest of P.
G. Tait, the Professor of Natural Philosophy at Edinburgh. Tait
addressed the Royal Society of Edinburgh on the subject and pub-
lished two papers on the (what we should now call) Four Colour
Theorem. They contain some clever ideas and a number of basic
errors.
The Four Colour Theorem returned to being the Four Colour
Conjecture in 1890. Percy John Heawood, a lecturer at Durham
England, published a paper called Map Colouring Theorem. In it
he states that his aim is
rather destructive than constructive, for it will be shown
that there is a defect in the now apparently recognized
proof.
Although Heawood showed that Kempe’s proof was wrong he
did prove that every map can be 5-coloured in this paper. Kempe
reported the error to the London Mathematical Society himself
and said he could not correct the mistake in his proof. In 1896
de la Vallée Poussin also pointed out the error in Kempe’s paper,
apparently unaware of Heawood’s work.
Heawood was to work throughout his life on map colouring,
work which spanned nearly 60 years. He successfully investigated
the number of colours needed for maps on other surfaces and gave
122 13 Famous problems in Math history

what is known as the Heawood estimate for the necessary number

in terms of the Euler characteristic of the surface.
Heawood’s other claim to fame is raising money to restore
Durham Castle as Secretary of the Durham Castle Restoration
Fund. For his perseverance in raising the money to save the Castle
from sliding down the hill on which it stands Heawood received
the O.B.E.
Heawood was to make further contributions to the Four Colour
Conjecture. In 1898 he proved that if the number of edges around
each region is divisible by 3 then the regions are 4-colourable. He
then wrote many papers generalizing this result.
To understand the later work we need to define some concepts.
Clearly a graph can be constructed from any map the regions
being represented by the vertices and two vertices being joined by
an edge if the regions corresponding to the vertices are adjacent.
The resulting graph is planar, that is can be drawn in the plane
without any edges crossing. The Four Colour Conjecture now asks
if the vertices of the graph can be coloured with 4 colours so that
no two adjacent vertices are the same colour.
From the graph, a triangulation can be obtained by adding edges
to divide any non-triangular face into triangles. A configuration is
part of a triangulation contained within a circuit. An unavoidable
set is a set of configurations with the property that any triangu-
lation must contain one of the configurations in the set. A config-
uration is reducible if it cannot be contained in a triangulation of
the smallest graph which cannot be 4-coloured.
The search for avoidable sets began in 1904 with work of
Weinicke. Renewed interest in the USA was due to Veblen who
published a paper in 1912 on the Four Colour Conjecture general-
ising Heawood’s work. Further work by G. D. Birkhoff introduced
the concept of reducibility (defined above) on which most later work
rested.
Franklin in 1922 published further examples of unavoidable sets
and used Birkhoff’s idea of reducibility to prove, among other re-
sults, that any map with ≤ 25 regions can be 4-coloured. The
number of regions which resulted in a 4-colourable map was slowly
increased. Reynolds increased it to 27 in 1926, Winn to 35 in 1940,
Ore and Stemple to 39 in 1970 and Mayer to 95 in 1976.
 13.2 The Four Color Problem 123

However the final ideas necessary for the solution of the Four
Colour Conjecture had been introduced before these last two re-
sults. Heesch in 1969 introduced the method of discharging. This
consists of assigning to a vertex of degree i the charge 6 − i. Now
from Euler’s formula we can deduce that the sum of the charges
over all the vertices must be 12. A given set S of configurations
can be proved unavoidable if for a triangulation T which does not
contain a configuration in S we can redistribute the charges (with-
out changing the total charge) so that no vertex ends up with a
positive charge.
Heesch thought that the Four Colour Conjecture could be
solved by considering a set of around 8900 configurations. There
were difficulties with his approach since some of his configurations
had a boundary of up to 18 edges and could not be tested for re-
ducibility. The tests for reducibility used Kempe chain arguments
but some configurations had obstacles to prevent reduction.
The year 1976 saw a complete solution to the Four Colour
Conjecture when it was to become the Four Colour Theorem for
the second, and last, time. The proof was achieved by Appel and
Haken, basing their methods on reducibility using Kempe chains.
They carried through the ideas of Heesch and eventually they con-
structed an unavoidable set with around 1500 configurations. They
managed to keep the boundary ring size down to 14, making com-
putations easier that for the Heesch case. There was a long period
where they essentially used trial and error together with unbeliev-
able intuition to modify their unavoidable set and their discharging
procedure. Appel and Haken used 1200 hours of computer time to
work through the details of the final proof. Koch assisted Appel
and Haken with the computer calculations.
The Four Colour Theorem was the first major theorem to be
proved using a computer, having a proof that could not be verified
directly by other mathematicians. Despite some worries about this
initially, independent verification soon convinced everyone that the
Four Colour Theorem had finally been proved. Details of the proof
appeared in two articles in 1977. Recent work has led to improve-
ments in the algorithm.
Index

abscissa, 46 boundary, 1, 59
absolute value, 2, 10
abstract, 9 calculate, 1
additive, 37 calculus, 1
additively, 37 canonical, 13
adjacent, 78, 121 category, 98
algebraic, 105 ceil, 22
algorithm, 47 chain, 70
analogy, 76 chaos, 29
analytic, 106 circle, 1
analytical, 47 circumference, 1
angle, 10 coarser, 66
antiderivative, 4 collection, 96
applicability, 50 column, 10
approach, 1–3 combinatorics, 76, 107
approximate, 2 commutativity, 38
approximation, 1, 45 compact, 95
arc, 78 compactness, 95
arithmetic, 105 comparable, 66
array, 9 completeness, 98
assignment, 88 complex, 105
associative, 37 complexity, 86
associativity, 38 component, 11
assume, 47 componentwise, 11
assumption, 89 composition, 37
asymptotic, 33 compute, 3
average, 3 conclusion, 76
axiomatically, 9 conditional, 76
confidence, 23
barycentric, 69 configuration, 122
bases, 9 conjecture, 106
basis, 97 conjugate, 11, 100
orthogonal-, 97 conjunction, 75
126 Index

connective, 76 element, 10
constraint, 88 eliminate, 56
functional-, 88 elliptic, 60
nonnegativity-, 88 endomorphism, 39
continuous, 4 equation
contradict, 75 characteristic, 81
converge, 2 equivalent, 99
convergent, 98 error, 1
convex, 97 estimate, 122
coordinates, 97 evaluating, 5
coordinatize, 97 event, 30
corollary, 117 execute, 47
correctness, 75 experiment, 29
countable, 98 explicit, 77
counter-example, 117 extreme, 95
cover, 96
cryptography, 109 factorization, 115
curve, 5 fallacious, 114
cycle, 70 fallacy, 114
false, 75
decomposition, 9 feasible, 86
degenerate, 10 region, 88
denominator, 1 field, 10
derangement, 77 figure, 45
derivative, 3 finer, 66
partial-, 58 floating-point, 10
deterministic, 29 form
diagonal, 13 bilinear-, 9
diameter, 1 quadratic-, 9
dichotomy, 99 frequency, 22, 32
difference, 46 function, 3
Differential, 4 position-, 3
differential, 55 objective, 88
differentiate, 55
differentiation, 4 geometry, 1
digit, 45 graph, 77
significant-, 45 acyclic, 79
dimension, 9, 11 bipartite, 78
disjunction, 76 complete, 78
disks, 95 planar, 79
dispersion, 23 simple-, 78
distance, 5 undirected-, 78
distributivity, 39 weighted, 79
domain, 99
homogeneous, 81
edge, 67, 78 homomorphism, 39
eigenfunction, 59 hyperbolic, 60
eigenvalue, 13 hyperplane, 69
eigenvector, 13 hypotheses, 47
 Index 127

hypothesis, 76 marginal, 113

mathematical, 86
ideal, 40 matrix, 9
image, 37 block triangular , 14
imaginary, 11 adjacency-, 78
implication, 76 mean, 20
implicit, 55 arithmetic-, 20
inaccuracy, 10 geometric-, 20
inclusion, 98 harmonic-, 20, 21
increase, 1 median, 22
independence, 30 member, 2
induced, 38, 97 middle, 22
induction, 81 mode, 22
infinite, 2 modular, 108
infinitely, 56 module, 40
initial, 58 modulus, 11
inscribe, 1 monoid, 38
integrable, 6
integral, 4 negation, 76
integration, 4 neighborhood, 96
interior, 69 net, 98
intermediate, 47 norm, 10
interpolation, 46 number
interpret, 30 complex, 9
intersection, 66 even, 115
interval, 3 integer, 105
intuitive, 47 irrational-, 2
invariant, 67 odd, 114
inverse, 38 prime, 105
invertible, 40 rational, 1
isogonal, 57 real, 9
isometry, 98 numerator, 1
isomorphic, 98
isomorphism, 39, 98 operation, 11
iterating, 31 operator
unitary-, 99
length, 5 integral-, 100
limit, 1 order, 55
line segment, 11 orientation, 70
linear oriented, 70
-combination, 12 origin, 3
-programming, 86 orthogonal, 57
-transformation, 9 orthogonality, 97
linear algebra, 9 outcome, 30
logic, 75
loop, 78 paradox, 75
partition, 78
magnitude, 45 path, 79
margin, 113 percentile, 23
128 Index

perimeter, 1 sample, 31
permutation, 77 sampling, 31
phenomenon, 19 scalar, 9
plane, 95 scientific, 85
polygon, 1 secant, 3
polyhedron, 67 sentence, 75
polynomial declarative-, 75
characteristic-, 13 sequence, 1, 47
interpolating-, 48 sequential, 96
root, 13 sesqui, 100
predictability, 29 sesquilinear, 100
premise, 76 set
probabilistic, 107 ordered-, 95
probability, 19, 29 orthogonal-
conditional-, 31 maximal-, 97
process, 1 ordered, 66
proof, 108 side, 1
propagation, 49 simplex
proportional, 90 method, 87
proposition, 75 simplicial, 70
compound-, 75 skew, 22
slope, 1
quadratic, 106 solution, 88
quadrature, 49 feasible-, 88
quantitatively, 19 optimal-, 88
quartile, 24 space, 95
quasi triangular, 14 Euclidean-, 95
quaternion, 120 metric-, 96
quotient, 30 complete-, 97
topological-, 96
randomness, 29 vector-, 12
range, 22 linear-, 12
crude-, 22 quotient-, 69
rate, 3 standard deviation, 23
ratio, 1 statistical, 19
recurrence, 81 statistics, 19
Recursion, 80 descriptive-, 19
reducibility, 122 inductive-, 19
reducible, 122 inferential-, 19
reformulate, 91 stochastic, 30
regular, 1 subgraph, 78
relation subgroup, 40
equivalence-, 98 subinterval, 5
congruence, 108 submodule, 41
equivalence, 69 submonoid, 38
replacement, 31 subsequence, 96
restrictive, 91 subsequent, 45
ring, 39 subspaces, 9
rounding, 45 subtract, 22
 Index 129

successive, 46 true, 75
surface, 69
surjective, 99 uncountable, 98
symmetric, 78 uniform, 22
uniformly, 96
union, 66
tangent, 3
unoriented, 70
term, 2
unpredictability, 29
terminology, 66
tetrahedron, 68
variability, 23
topology, 10, 65
variable, 2
discrete-, 66 decision-, 87
indiscrete, 66 random-, 20
trajectory, 57 variance, 24
transitive, 45 variation
transportation, 88 coefficient of- , 23
tree, 79 vector, 9
rooted-, 79 space, 10
spanning-, 79 velocity, 3
minimal, 79 instantaneous-, 3
triangle verification, 38
right, 114 vertex, 67, 78
triangulable, 70 degree of a-, 78
triangular, 13 isolated-, 78
triangulation, 70, 122 volume, 5
trivial, 38
trivially, 39 weighting, 24