MATH0390 Foundation Pure and Applied

Mathematics

Professor Serguei Komissarov

2023-10-16
2
Contents

Preface 5

1 Linear and quadratic functions 9

1.1 Linear functions . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2 Linear supply and demand functions . . . . . . . . . . . . . . . . 14
1.3 Quadratic functions . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.4 Optimisation of revenue . . . . . . . . . . . . . . . . . . . . . . . 28

2 Exponentiation and power functions 33

2.1 Natural exponents . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2 Integer exponents . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.3 Rational exponents . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.4 Irrational exponents . . . . . . . . . . . . . . . . . . . . . . . . . 42

3
4 CONTENTS
Preface

Mathematics deals with highly abstract objects obeying strict basic rules and
other rules that can be derived from these basic rules using logic. These objects
and operations with them are denoted using various symbols, and mathemati-
cal calculations appear as manipulation of these symbols. The most primitive
symbols are numerical, they are used to describe numbers and the arithmetic
operations. However, arithmetics is only a tiny area of mathematics. To give
you an idea, Figure 1 shows only a small subset of symbols used in other areas
of mathematics. Learning symbolic calculations is a very important part of our
course.
Basic objects of mathematics are direct abstracts from nature. Many others
emerged solely for the sake of development of new mathematics. Development
of new mathematics and its exploration is the domain of pure mathematics.
Applications of mathematics to describe and study the world around us is the
aim of applied mathematics. Many properties of the real world around us
can be quantified. This means they can be measured and expressed as numer-
ical parameters. These parameters may take different values, vary in time
or space, and for this reason can be called variables. Different variables can
be interdependent - changes in some of them are followed by changes in others.
In mathematics, such connections (relations) between variables are described
via mathematical functions. So, mathematical modelling of real world phenom-
ena often involves identification of key parameters and suitable mathematical
functions. Often, it is not practical to include all involved parameters as this
may lead to overly complicated models which are hard to comprehend and use.
Instead, especially at the beginning, the models retain only very few most im-
portant parameters and use simple functions. This explains why in this course
we focus on functions of only one variable.
In abstract mathematical notation, such functions are presented in the form

𝑦 = 𝑓(𝑥) ,

where 𝑓 stands for some rule which shows how to calculate the value of variable
𝑦 given the value of variable 𝑥. Hence, 𝑥 is called an independent variable

5
6 CONTENTS

Figure 1: A small selection of mathematical symbols

CONTENTS 7

and 𝑦 is called a dependent variable. The set of all allowed values of 𝑥 is

called the domain of 𝑓(𝑥). Often the domain is dictated by the application.
For example, 𝑥 may stand for body’s mass, in which case the domain is the set
of all positive values of 𝑥, or if it stands for a fraction of oxygen in the air then
the domain is [0, 1], the set of all numbers between 0 and 1, including the end
points. The set of all values 𝑦 can take is called the range of 𝑓(𝑥).

In this course, we will study some of the most basic, and hence most important
types of such mathematical functions, like the linear, quadratic, power, expo-
nential, logarithmic, and trigonometric functions. We will also study some key
ways of exploring functions, like their differentiation and integration. This will
be accompanied by many real world examples, mostly from economics.

It is very useful to visualise functions by means of graphs. We will be dealing

with the most common kind of graphs, where functions are represented by means
of lines on a plot which also shows a horizontal and a vertical straight lines called
axes. The point of their intersection is called the origin. Any point on the
plane of the plot is uniquely identified by two (geometric) distances from the
axes. These distances are coordinates of this point. To plot a point (𝑥, 𝑦),
where 𝑥 and 𝑦 are numerical values of two variables (representing any kind of
quantified properties, e.g. amount of money, mass, volume etc.), we choose the
distance along the horizontal axis which is proportional to 𝑥 (the distance = 𝑎𝑥,
where 𝑎 is a constant factor) and the distance along the vertical axis which is
proportional to 𝑦 (the distance = 𝑏𝑦, where 𝑏 is another constant factor). One
is free to choose any convenient values for 𝑎 and 𝑏 as their values determine
the location of numerical labels on the x and y axes. To make a graph of the
function 𝑦 = 𝑓(𝑥) for a certain range 𝑥 ∈ [𝑐, 𝑑], one plots all points (𝑥, 𝑓(𝑥))
with 𝑥 in this range. Normally, this is a continuous line.

y )
f(x
y=

f(x2)

f(x1)

x1 x2 x

Figure 2: Graph of a function

8 CONTENTS

We will be making such plots very frequently throughout the course, as they
help to develop better understanding.
Chapter 1

Linear and quadratic

functions

1.1 Linear functions

Consider the case when any change of 𝑥 by the amount Δ𝑥 is followed by a
change of 𝑦 by the amount Δ𝑦 = 𝑎Δ𝑥 where 𝑎 ≠ 0 is always the same number (
a constant). Such relation between 𝑥 and 𝑦 is called linear and it is described
by a linear function. The graph of this function is a straight line (see figure
1.1).
The constant 𝑎 is called the slope (or the gradient) of the line because it tells
us how big is the vertical change relative to the horizontal change along
the line, i.e.

Δ𝑦
𝑎= .
Δ𝑥

• If 𝑎 = 0 the line is horizontal;

• If 𝑎 > 0 the line goes upwards to the right.
• If 𝑎 < 0 the line goes downwards to the right.

This is illustrated in figure 1.2.

From this definition, it is easy to find the formula describing the linear function
in terms of arithmetic operations.
Let (𝑥0 , 𝑦0 ) be some point on the line representing the linear function 𝑓(𝑥0 ) = 𝑦0
as 𝑦0 . If (𝑥, 𝑦) is any other point of the line, then Δ𝑥 = 𝑥 − 𝑥0 , Δ𝑦 = 𝑦 − 𝑦0 ,
and

9
10 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

aDx
y0
Dx

x0 x

Figure 1.1: Graph of a linear function is a straight line.

y
a>0

y0 a=0

a<0

x0 x

Figure 1.2: Graphs of linear functions for 𝑎 > 0, 𝑎 = 0, and 𝑎 < 0.

1.1. LINEAR FUNCTIONS 11

𝑦 − 𝑦0
= 𝑎. (1.1)
𝑥 − 𝑥0

Hence,

𝑦 = 𝑦0 + 𝑎(𝑥 − 𝑥0 ) . (1.2)

This equation shows how, given the values of 𝑥0 , 𝑦0 and 𝑎, one can compute 𝑦
for any 𝑥 using the operations of addition, subtraction and multiplication. (It
is obvious that for 𝑥 = 𝑥0 the formula yields 𝑦 = 𝑦0 , as expected.) If the value
of 𝑎 is not given, but we know another point of the line, we can derive it using
(1.1). Denote the second point as (𝑥1 , 𝑦1 ). Since (1.1) is valid for any point
(𝑥, 𝑦) of the line, it must be valid for (𝑥1 , 𝑦1 ) too, and hence

𝑦1 − 𝑦 0
𝑎= . (1.3)
𝑥1 − 𝑥 0

In equation (1.2), instead of the arbitrary point (𝑥0 , 𝑦0 ) of the line we may
choose the one where the line intercepts the y axis (the y-intercept) . Since
along the y axis 𝑥 = 0, we may write (𝑥0 , 𝑦0 ) = (0, 𝑏) and obtain

𝑦 = 𝑎𝑥 + 𝑏 . (1.4)

One advantage of this form is that it is more compact, and the other is that the
value of 𝑏 shows where the line intercepts the y axis. This is the most common
for of general representation of linear functions.
Instead of the y-intercept we may use the x-intercept, the point where the
line intercepts the x axis. For this intercept we may write (𝑥0 , 𝑦0 ) = (𝑐, 0) and
equation (1.2) reduces to
12 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

𝑦 = 𝑎(𝑥 − 𝑐) . (1.5)

aDx
y0
Dx
y=y0+ a(x-x0)
b
=ax+b
=a(x-c)

c x0 x

Figure 1.3: Graphic interpretation of all parameters appearing in the three

different forms of linear function.

Traditionally, mathematicians use the symbols 𝑦 and 𝑥 to denote the dependent

and independent variables, respectively. However in applications, other symbols
may be preferable. When dealing with applications, it is also important to chose
convenient units.
Example 1.1. The global world population was 5978 million in the year 1999,
and 6707 million in the year 2008. Assuming a linear relation between the global
population and time, predict the global population in the year 2049.

Solution.
In this problem, we need to find the linear function approximating the popula-
tion growth given its values at two times. Since any straight line is completely
determined by its two points, this problem will always have a unique solution.
We have already solved this problem in general via symbolic calculations (deriva-
tions). The solution is given by equations (1.2) and (1.3). Here we simply apply
this result.
If is instead of 𝑥 and 𝑦 we use symbols 𝑡 (for time) and 𝑝 (for population)
respectively, then we have

𝑝 = 𝑝0 + 𝑎(𝑡 − 𝑡0 ) , (1.6)
1.1. LINEAR FUNCTIONS 13

where

𝑝1 − 𝑝0
𝑎= . (1.7)
𝑡1 − 𝑡0

To proceed with numerical calculations, we need to choose how we measure the

population and time. The data shows that the population is best measured in
millions (billions will also be good). In these units,

𝑝0 = 5978 , 𝑝1 = 6707 .

The time is best measured in years. Moreover, if we decide to measure the time
not starting not from the birth of Christ but from the year 1999, the numbers
will be smaller (and hence easier to deal with). Using these definitions, our key
times are

𝑡0 = 0 , 𝑡1 = 2008 − 1999 = 9 , 𝑡2 = 2049 − 1999 = 50 .

Substituting the data for 𝑡0 and 𝑡1 into eq.(1.7), we find

6707 − 5978
𝑎= = 81 .
9

and hence eq.(1.6) reads

𝑝 = 81𝑡 + 5978 . (1.8)

Now we can compute the population in the year 2049,

𝑝2 = 𝑝(50) = 81 × 50 + 5978 = 10028

in millions of people. (If you forgot about the units and wrote that
the predicted population was 10028 people, this would be a major
mistake!!! Unfortunately, quite a large number of students made
mistakes of this type in previous examinations.)
Comment. This linear model also predicts 𝑝 = −22 for 𝑡 = −75 (the year
1925). This negative number does not make sense and shows the limitations of
linear approximations! Normally, the linear function is the least accurate ap-
proximation one can make. Nonetheless, it is very simple and can be sufficiently
accurate for a narrow range of the independent variable, where the graph of the
actual data function is close to a fragment of straight line.
14 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

1.2 Linear supply and demand functions

As an application of linear functions, we now consider simplified models of
demand and supply.
The demand is the amount (or quantity) of a product which consumers would
buy per certain period of time. The supply is the amount (or quantity) of this
product provided on the market by manufactures per same period of time. Both
the demand and supply can be influenced by various factors. Here we will ignore
all these factors but the price, probably the most important factor. Moreover,
we will describe the dependence of demand and supply on the price using linear
functions. Although this model is an over-simplification, it has the merit of
being both simple and insightful.

1.2.1 Linear model of demand

Denote the market price of a product as 𝑃 , and the demand as 𝑄𝐷 . Obviously,

only 𝑃 > 0 and 𝑄𝐷 > 0 make sense in this application. Typically, 𝑄𝐷 (𝑃 ) de-
creases with 𝑃 and above a certain price 𝑃𝑚𝑎𝑥 the demand dries out completely.

Linear demand function:

Like any linear function, the linear demand function can be written in the
standard form

𝑄𝐷 = 𝑎𝐷 𝑃 + 𝑏𝐷 . (1.9)

To reflect the decrease of 𝑄𝐷 with 𝑃 we require 𝑎𝐷 < 0, and to reflect the

always existing demand at vanishing price, 𝑄𝐷 (0) > 0, we require 𝑏𝐷 > 0.
These conditions ensure that 𝑄𝐷 = 0 at 𝑃𝑚𝑎𝑥 = −𝑏𝐷 /𝑎𝐷 > 0.
In economics a different form is preferred,

𝑄𝐷 = 𝛼(𝑃𝑚𝑎𝑥 − 𝑃 ) , (1.10)

where 𝛼 = −𝑎𝐷 > 0 and 𝑃𝑚𝑎𝑥 = −𝑏𝐷 /𝑎𝐷 > 0. The advantage of this form
is that 1) all its parameters are positive, 2) it is clear that the demand is a
decreasing function of price, and 3) it is also clear that the demand dries out
when the price reaches the value of 𝑃𝑚𝑎𝑥 .
1.2. LINEAR SUPPLY AND DEMAND FUNCTIONS 15

QD

bD

Pmax P

Figure 1.4: Graph of a linear demand function 𝑄𝐷 = 𝑎𝐷 𝑃 + 𝑏𝐷 .

1.2.2 Linear model of supply

Denote as 𝑄𝑆 be the quantity of the same product supplied to the market by

manufacturers. Typically, it increases with the market price 𝑃 of the product
(The higher is the price, the higher is the expected profit). Moreover, below
a certain price 𝑃𝑚𝑖𝑛 , the proceeds from the sales will be below the associated
costs and no profit can be made. Hence at this price the productions stops and
the supply dries out.

Linear supply function:

The linear supply function can also be written in the standard form

𝑄𝑆 = 𝑎 𝑆 𝑃 + 𝑏 𝑆 , (1.11)

where the slope 𝑎𝑆 > 0 and 𝑏𝑆 < 0. This model captures the growth of 𝑄𝑆 with
𝑃 and the existence of minimal price, 𝑃𝑚𝑖𝑛 = −𝑏𝑆 /𝑎𝑆 > 0. In economics, the
linear supply function is often written as

𝑄𝑆 = 𝛽(𝑃 − 𝑃𝑚𝑖𝑛 ) , (1.12)

where 𝛽 > 0 is the same as 𝑎𝑆 . The advantage of this form is that 1) all its
parameters are positive, 2) it is clear that the demand is an increasing function
of price, and 3) it is also clear that the supply dries out when the price drops
to the value of 𝑃𝑚𝑖𝑛 .
16 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

QS

Pmin P

Figure 1.5: Graph of a linear supply function 𝑄𝑆 = 𝑎𝑆 𝑃 + 𝑏𝑆 .

1.2.3 Equilibrium price

Obviously, the model based on the linear supply and demand functions deter-
mines the meaningful price range as 𝑃𝑚𝑖𝑛 < 𝑃 < 𝑃𝑚𝑎𝑥 . Only for prices within
this range both 𝑄𝐷 > 0 and 𝑄𝑆 > 0. Note that this range exists only when
𝑃𝑚𝑖𝑛 < 𝑃𝑚𝑎𝑥 , the top price at which the demand dries out is higher than the
bottom price at which the supply dries out.
But what determines the actual price at which the product is sold?
The market forces normally drive the price of a product towards a particular
equilibrium value 𝑃𝑒𝑞 . When the price 𝑃 < 𝑃𝑒𝑞 , the demand exceeds supply
(𝑄𝐷 > 𝑄𝑆 ) and the stock get sold very quickly, leaving empty shelves and
unhappy customers. If the price 𝑃 > 𝑃𝑒𝑞 , the supply exceeds demand (𝑄𝐷 <
𝑄𝑆 ), and unsold stock accumulates at shops and warehouses, increasing costs.
When 𝑃 = 𝑃𝑒𝑞 , the supply matched the demand exactly, 𝑄𝑆 = 𝑄𝐷 . This is the
optimum sales regime.
Given the model functions 𝑄𝐷 (𝑃 ) and 𝑄𝑆 (𝑃 ), the equilibrium price can be
found by solving the balance equation

𝑄𝐷 (𝑃𝑒𝑞 ) = 𝑄𝑆 (𝑃𝑒𝑞 )

for 𝑃𝑒𝑞 . For the linear demand and supply functions given in the forms (1.9)
and (1.11), this balance equation reads

𝑎𝐷 𝑃𝑒𝑞 + 𝑏𝐷 = 𝑎𝑆 𝑃𝑒𝑞 + 𝑏𝑆 ,
1.2. LINEAR SUPPLY AND DEMAND FUNCTIONS 17

QD ,QS

QD QS

Pmin Peq Pmax P

Figure 1.6: At equilibrium price 𝑃𝑒𝑞 , the product supply matches the demand
for it.

and yields

𝑏𝐷 − 𝑏 𝑆
𝑃𝑒𝑞 = . (1.13)
𝑎𝑆 − 𝑎𝐷

One can also use 𝑄𝐷 (𝑃 ) and 𝑄𝑆 (𝑃 ) in the forms (1.10) and (1.12). In this case,
the balance equation reads

𝛼(𝑃𝑚𝑎𝑥 − 𝑃𝑒𝑞 ) = 𝛽(𝑃𝑒𝑞 − 𝑃𝑚𝑖𝑛 ) ,

and yields,

𝛼𝑃𝑚𝑎𝑥 + 𝛽𝑃𝑚𝑖𝑛
𝑃𝑒𝑞 = . (1.14)
𝛼+𝛽

This result looks different but only because it is given in terms of different
parameters. In fact, the results (1.13) and (1.14) are equivalent.
Example 1.2.

a) When the price of a book on sale is £15, bookshops sell 10,000 copies per
month. When the price is £9, they sell 16,000 copies per month. Use the
linear demand function to determine the upper limit 𝑃𝑚𝑎𝑥 on the price at
which this book can still be sold.
b) For the book price of £22, there is a supply of 26,000 copies per month.
For the price below 𝑃𝑚𝑖𝑛 = £2, the book is no longer supplied. Determine
the corresponding linear supply function.
18 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

c) Use the linear supply and demand models obtained in a) and b) to deter-
mine the equilibrium book price. Hence determine the rate of book sales
at this price.

Solution.
In this example it is convenient to measure the rate of sale in thousands per
month and the price in pounds and write the provided data as
For parts a) and b), we again can use the general result (1.2) and (1.3).
a) So we write

𝑄𝐷 = 𝑄𝐷0 + 𝑎𝐷 (𝑃 − 𝑃0 ) , (1.15)

where

𝑄𝐷1 − 𝑄𝐷0
𝑎𝐷 = . (1.16)
𝑃1 − 𝑃 0

In the selected units,

(𝑃0 , 𝑄𝐷0 ) = (15, 10) , (𝑃1 , 𝑄𝐷1 ) = (9, 16) .

Substituting these numbers, we obtain

16 − 10
𝑎𝐷 = = −1 ,
9 − 15
and

𝑄𝐷 = 10 − (𝑃 − 15) = 25 − 𝑃 .

This result is already in the form

𝑄𝐷 = 𝛼(𝑃 − 𝑃𝑚𝑎𝑥 ) ,
with 𝑃𝑚𝑎𝑥 = 25 and 𝛼 = 1. (The demand vanishes when the price reaches £25.)
b) Here we can proceed with exactly the same calculations as in part a), but
because we already know that 𝑃𝑚𝑖𝑛 = 2 it is better to write the supply function
in the form

𝑄𝑆 = 𝛽(𝑃 − 𝑃𝑚𝑖𝑛 ) .

Here 𝛽 is the only unknown parameter and we can find it substituting the data
(𝑃 , 𝑄𝑆 ) = (22, 26). This yields
1.3. QUADRATIC FUNCTIONS 19

26 = 𝛽(22 − 2) ,

and hence

𝛽 = 1.3 .

So, the solution to this part is

𝑄𝑆 = 1.3(𝑃 − 2) .

c) To find the equilibrium price, we can simply use the general result (1.14)

𝛼𝑃𝑚𝑎𝑥 + 𝛽𝑃𝑚𝑖𝑛
𝑃𝑒𝑞 = .
𝛼+𝛽

Hence

1 × 24 + 1.3 × 2 27.6
𝑃𝑒𝑞 = = = 12 .
1 + 1.3 2.3

Thus, the equilibrium price of books is £12.

For 𝑃 = 12,

𝑄𝐷 = 25 − 12 = 13 .

Hence the at this price the shops sale 13, 000 copies per month. □
Take Quiz 1 on Minerva (Learning Resources > Quizzes).

1.3 Quadratic functions

Linear approximations are the simplest but also least accurate. When it becomes
clear that this approximation in not sufficiently accurate, e.g. the data points
deviate significantly from a straight line, we begin looking for a better one. The
next easiest choice is the quadratic approximation, where we utilise quadratic
functions instead of linear ones. It is more flexible by allowing curvature of
model graphs and it includes the linear approximation as a limiting case.
Usually, quadratic functions are presented in the following standard form
20 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

Quadratic function

𝑦 = 𝑎𝑥2 + 𝑏𝑥 + 𝑐 , (1.17)

where 𝑎 ≠ 0, 𝑏 and 𝑐 are constants (parameters of this function).

One can see that in the limit 𝑎 → 0 this function reduces to the linear function
𝑦 = 𝑏𝑥 + 𝑐. This is why we consider here only the case where 𝑎 ≠ 0.

Parabola. The most basic quadratic function is

𝑦 = 𝑥2 . (1.18)

The graph of this function is called a parabola.

y=x2

Figure 1.7: The graph of basic parabola 𝑦 = 𝑥2 .

It is symmetric relative to the y axis

𝑦(−𝑥) = 𝑦(𝑥) ,

because (−𝑥)2 = 𝑥2 . It opens up, and its turning point (the vertex) is located
at (0, 0), the origin.
Scaled parabola. Multiplying (1.18) by 𝑎 ≠ 0, we obtain the quadratic func-
tion

𝑦 = 𝑎𝑥2 . (1.19)

It described the parabola stretched along the y axis by the factor |𝑎|.
1.3. QUADRATIC FUNCTIONS 21

y=ax2 y=x2

a>0

Figure 1.8: The graph of scaled parabola 𝑦 = 𝑎𝑥2 .

If 𝑎 > 0, it still opens up and its vertex is still at (0, 0). It rises faster than the
parabola if 𝑎 > 1 and slower if 𝑎 < 1.
If 𝑎 < 0, the parabola is not only scaled but also reflected about the x axis - it
opens down. The graphs of 𝑦 = 𝑎𝑥2 is the graph of 𝑦 = −𝑎𝑥2 reflected about
the x axis.
Scaled parabola shifted along the x axis.
Shifting all points of the scaled parabola along the x axis by the same amount
𝑥𝑒 leads to the curve described by the quadratic function

𝑦 = 𝑎(𝑥 − 𝑥𝑒 )2 . (1.20)

If 𝑥𝑒 > 0, the parabola is shifted to the right, and if 𝑥𝑒 < 0, it is shifted to the
left. It’s vertex is now at the point (𝑥𝑒 , 0).

y=ax2 y=a(x-xe )2

a>0
xe>0

xe x

Figure 1.9: The graph of 𝑦 = 𝑎(𝑥 − 𝑥𝑒 )2 , scaled parabola shifted by 𝑥𝑒 along

the x axis.

Scaled parabola shifted along both the x and y axes.

Shifting all points of the scaled parabola along the x axis by 𝑥𝑒 and along the
y axis by 𝑦𝑒 leads to the curve described by the quadratic function
22 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

𝑦 = 𝑎(𝑥 − 𝑥𝑒 )2 + 𝑦𝑒 . (1.21)

If 𝑦𝑒 > 0, the parabola is shifted upwards, and if 𝑦𝑒 < 0, it is shifted downwards.

Its vertex is now at the point (𝑥𝑒 , 𝑦𝑒 ).

y=ax2

y=a(x-xe )2 +ye

ye a>0
xe>0
ye>0

xe x

Figure 1.10: The graph of 𝑦 = 𝑦𝑒 + 𝑎(𝑥 − 𝑥𝑒 )2 , scaled parabola shifted by 𝑥𝑒

along the x axis and by 𝑦𝑒 along the y axis.

Equation (1.21) is known as the vertex form of quadratic function. Ex-

panding the square, we obtain

𝑦 = 𝑎𝑥2 − (2𝑎𝑥𝑒 )𝑥 + (𝑦𝑒 + 𝑎𝑥2𝑒 ) = 0 .

Comparing this with the standard form (1.17) we find the formulae connecting
the parameters of both forms,

𝑏 = −2𝑥𝑒 𝑎 , 𝑐 = 𝑦𝑒 + 𝑎𝑥2𝑒 , (1.22)

and

𝑏
𝑥𝑒 = − , 𝑦𝑒 = 𝑐 − 𝑎𝑥2𝑒 . (1.23)
2𝑎

Vertex form

𝑦 = 𝑎(𝑥 − 𝑥𝑒 )2 + 𝑦𝑒 . (1.24)

In this form, 𝑎, 𝑥𝑒 , 𝑦𝑒 are considered as parameters of the quadratic function.

1.3. QUADRATIC FUNCTIONS 23

X-intercepts of quadratic function are solutions to the quadratic equa-

tion

𝑎𝑥2 + 𝑏𝑥 + 𝑐 = 0 .

To find them, we utilise the vertex form of the equation

𝑦𝑒
𝑎(𝑥 − 𝑥𝑒 )2 + 𝑦𝑒 = 0 ⇒ (𝑥 − 𝑥𝑒 )2 = − .
𝑎
Now, it is easy to see that

• if −𝑦𝑒 /𝑎 < 0, there are no real solutions (no x-intercepts);

• if −𝑦𝑒 /𝑎 = 0, there is only one real solution, 𝑥𝑒 (the vertex);
• if −𝑦𝑒 /𝑎 > 0, there are two real solutions

𝑦𝑒
𝑥± = 𝑥𝑒 ± √− .
𝑏
Substituting the expressions (1.23) for 𝑥𝑒 and 𝑦𝑒 , we immediately obtain the
following well-known result

Solution to quadratic equation

𝑎𝑥2 + 𝑏𝑥 + 𝑐 = 0

exists only if its discriminant

𝐷 = 𝑏2 − 4𝑎𝑐 ≥ 0 . (1.25)

If 𝐷 = 0, there is only one solution,

𝑏
𝑥𝑒 = − .
2𝑎
If 𝐷 > 0, there are two solutions
√ √
−𝑏 ± 𝐷 𝐷
𝑥± = = 𝑥𝑒 ± . (1.26)
2𝑎 2𝑎
Notice that the vertex is exactly between the x-intercepts of the quadratic
function.
24 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

Figures 1.11-1.13 illustrate these results.

y
a>0, D>0

x
a<0, D>0

Figure 1.11: For 𝐷 > 0 there are two x-intercepts.

y
a>0, D=0

x
a<0, D=0

Figure 1.12: For 𝐷 = 0 there is only one x-intercept.

Factorised form
Quadratic functions with two x-intercepts can be written in the form

𝑦 = 𝑎(𝑥 − 𝑥− )(𝑥 − 𝑥+ ) . (1.27)

In this form, not only 𝑎 but also the intercepts 𝑥− and 𝑥+ are considered as
function’s parameters.
Quadratic functions with only one x-intercept can be written in the form
1.3. QUADRATIC FUNCTIONS 25

y
a>0, D<0

x
a<0, D<0

Figure 1.13: For 𝐷 < 0 there are no x-intercepts.

𝑦 = 𝑎(𝑥 − 𝑥𝑒 )2 ,

where 𝑎 and 𝑥𝑒 are their parameters.

Example 1.3. Find the x- and y-intercepts of the quadratic function 𝑦(𝑥) =
2𝑥2 − 11𝑥 + 9. Then find its vertex.

Solution.
We are dealing with a quadratic function given in the form

𝑦(𝑥) = 𝑎𝑥2 + 𝑏𝑥 + 𝑐 ,

with parameters 𝑎 = 2, 𝑏 = −11 and 𝑐 = 9.

- Y-intercept: Since 𝑦(0) = 𝑐, the y-intercept is 𝐴 = (0, 9).
- X-intercepts: To find these, we need to solve the quadratic equation

2𝑥2 − 11𝑥 + 9 = 0 .

Its discriminant

𝐷 = 𝑏2 − 4𝑎𝑐 = 112 − 4 × 2 × 9 = 49 .

Since 𝐷 > 0, there are two solutions

√
−𝑏 ± 𝐷 11 ± 7
𝑥± = = = {1, 9/2} .
2𝑎 4
26 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

Hence the x-intercepts are 𝐵 = (1, 0) and 𝐷 = (9/2, 0).

- Vertex: Along the x axis the vertex is right in the middle between the x-
intercepts, at

𝑥+ + 𝑥− 𝑏
𝑥𝑒 = =− = 11/4 .
2 2𝑎

Its y coordinate can be found via equation (1.23)

𝑏2 (−11)2 72 − 121 49
𝑦𝑒 = 𝑐 − =9− = =− ,
4𝑎 4×2 8 8

or via direct calculations

11 2 11 49
𝑦𝑒 = 𝑦(𝑥𝑒 ) = 2 ( ) − 11 ( ) + 9 = ⋯ = − ,
4 4 8

which is a little bit longer. Hence the vertex is 𝐶 = (11/4, −49/8).

Given these results, the graph of this function can be sketched as shown in
Figure 1.14 □

B D x
C

Figure 1.14: Graph of the quadratic function 𝑦 = 2𝑥2 −11𝑥+9. Here 𝐴 = (0, 9),
𝐵 = (1, 0), 𝐶 = (11/4, −49/8), and 𝐷 = (9/2, 0).

Finding quadratic function given its three points. The linear function
has two parameters. They are fully determined by the values of the function
at two points. The quadratic function has three parameters, and to determine
them we need to know values of the function at three points. Here we show how
this is done.
The starting point is the general formula of quadratic function
1.3. QUADRATIC FUNCTIONS 27

𝑦 = 𝑎𝑥2 + 𝑏𝑥 + 𝑐 ,

and we need to determine its parameters 𝑎, 𝑏, and 𝑐, given the values of the
function at three points.
To simplify the calculations, we consider only the case of points equidistant
along the x axis. Denote them as 𝑥1 , 𝑥2 = 𝑥1 + Δ𝑥, 𝑥3 = 𝑥2 + Δ𝑥, where
Δ𝑥 is the distance between the points. Denote the corresponding values of the
functions respectively as 𝑦1 , 𝑦2 , and 𝑦3 . Hence

⎧𝑦1 = 𝑎𝑥21 + 𝑏𝑥1 + 𝑐 ,

{
2 (1.28)
⎨𝑦2 = 𝑎𝑥2 + 𝑏𝑥2 + 𝑐 ,
{𝑦 = 𝑎𝑥2 + 𝑏𝑥 + 𝑐 .
⎩ 3 3 3

This is a system of three (linear) equations for the three unknowns, 𝑎, 𝑏, and 𝑐.
To solve it, we can apply the Gauss elimination method. First, we eliminate 𝑐.
Subtracting the first equation from the second and the second from the third,
we obtain

𝑦 − 𝑦1 = 𝑎(𝑥22 − 𝑥21 ) + 𝑏(𝑥2 − 𝑥1 ) ,

{ 2
𝑦3 − 𝑦2 = 𝑎(𝑥23 − 𝑥22 ) + 𝑏(𝑥3 − 𝑥2 ) .

Since 𝑥3 − 𝑥2 = 𝑥2 − 𝑥1 = Δ𝑥, 𝑥22 − 𝑥21 = (𝑥2 − 𝑥1 )(𝑥2 + 𝑥1 ) = Δ𝑥(𝑥2 + 𝑥1 ),

and similarly 𝑥23 − 𝑥22 = Δ𝑥(𝑥3 + 𝑥2 ), this reduced system reads

𝑦2 − 𝑦1 = 𝑎Δ𝑥(𝑥2 + 𝑥1 ) + 𝑏Δ𝑥 ,
{ (1.29)
𝑦3 − 𝑦2 = 𝑎Δ𝑥(𝑥3 + 𝑥2 ) + 𝑏Δ𝑥 .

Next, we eliminate 𝑏 by subtracting the two equations of this system and ob-
taining

𝑦3 − 2𝑦2 + 𝑦1 = 𝑎Δ𝑥(𝑥3 − 𝑥1 ) = 2𝑎Δ𝑥2 .

From this equation, we find 𝑎

𝑦3 − 2𝑦2 + 𝑦1
𝑎= . (1.30)
2Δ𝑥2

Having found 𝑎, we can find 𝑏 from one of the equations in (1.29), e.g.

𝑦2 − 𝑦 1
𝑏 = −𝑎(𝑥2 + 𝑥1 ) + . (1.31)
Δ𝑥
28 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

Having found both 𝑎 and 𝑏, we can find 𝑐 from one of the equations in (1.28),
e.g.

𝑐 = 𝑦1 − 𝑎𝑥21 − 𝑏𝑥1 . (1.32)

Example 1.4. Find the quadratic function given its three values, 𝑦(1) = 1,
𝑦(2) = 3, and 𝑦(3) = 2.

Solution.
In this example, (𝑥1 , 𝑦1 ) = (1, 1), (𝑥2 , 𝑦2 ) = (2, 3), (𝑥2 , 𝑦3 ) = (3, 2) and Δ𝑥 = 1.
Applying (1.30)-(1.32), we find

2−2×3+1 3
𝑎= =− ;
2 2
3 3−1 13
𝑏 = (1 + 2) + = ;
2 1 2
3 13
𝑐 = 1 + × 12 − × 1 = −4 .
2 2
Thus, the quadratic function is

3 13
𝑦 = − 𝑥2 + 𝑥 − 4 ,
2 2
To check the result for computational errors, we can compute the values of the
derived function at 𝑥2 = 2 and 𝑥3 = 3. For example,

3 13
𝑦(𝑥2 ) = − 22 + 2 − 4 = −6 + 13 − 4 = 3 .
2 2
This is indeed 𝑦2 as it should be in the case of error-free computations. □

1.4 Optimisation of revenue

Revenue is the amount of money obtained from sales of a product over some
time interval. Here we denote is as 𝑅. Given the demand 𝑄𝐷 for the product,
which is the quantity of product sold per time interval, and the price of the
product 𝑃 , the revenue per the same time interval is

𝑅 = 𝑃 𝑄𝐷 . (1.33)

For example, if 𝑄𝐷 is measured as the quantity of items sold per day, and 𝑃 is
measured in pounds, then 𝑅 is be measured in pounds per day. The key target
for a seller is optimization (maximization) of the revenue.
1.4. OPTIMISATION OF REVENUE 29

For the linear model (1.10), the demand

𝑄𝐷 = 𝛼(𝑃𝑚𝑎𝑥 − 𝑃 ) with 𝛼 > 0,

and the revenue is

𝑅 = 𝛼𝑃 (𝑃𝑚𝑎𝑥 − 𝑃 ) = −𝛼𝑃 2 + 𝛼𝑃𝑚𝑎𝑥 𝑃 .

Thus, in this model the revenue is a quadratic function of 𝑃 . Since the coefficient
of 𝑃 2 is negative, the corresponding parabola opens down, and at the vertex
point the revenue reaches its highest possible value. The x-intercepts of the
revenue function are 𝑃− = 0 and 𝑃+ = 𝑃𝑚𝑎𝑥 , and hence the optimal price

𝑃𝑚𝑎𝑥
𝑃𝑜𝑝𝑡 = (𝑃− + 𝑃+ )/2 = ,
2
and the optimal revenue

𝑃𝑚𝑎𝑥 𝑃 𝑃2
𝑅𝑜𝑝𝑡 = 𝑅(𝑃𝑜𝑝𝑡 ) = 𝛼 (𝑃𝑚𝑎𝑥 − 𝑚𝑎𝑥 ) = 𝛼 𝑚𝑎𝑥 .
2 2 4

Pmax/2 Pmax P

Figure 1.15: Typical graph of the revenue function 𝑅(𝑃 ) = 𝛼𝑃 (𝑃𝑚𝑎𝑥 − 𝑃 ).

Optimal revenue
For the linear model (1.10) of the demand, the revenue 𝑅 is a quadratic function
of price:
30 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS

𝑅(𝑃 ) = 𝛼𝑃 (𝑃𝑚𝑎𝑥 − 𝑃 ) . (1.34)

The corresponding optimal price is

𝑃𝑚𝑎𝑥
𝑃𝑜𝑝𝑡 = , (1.35)
2
and the optimal revenue

2
(𝑃𝑚𝑎𝑥 )
𝑅𝑜𝑝𝑡 = 𝑅(𝑃𝑜𝑝𝑡 ) = 𝛼 . (1.36)
4

Example 1.5. In example 1.2, the demand for a book at a bookstore is

𝑄𝐷 = 25 − 𝑃 (1.37)

in thousands of books per month, where 𝑃 is the price of a single book in pounds.
Determine the optimal price for the book and the corresponding maximised
mounthly revenue.

Solution.
The revenue function is

𝑅 = 𝑃 𝑄𝐷 = 𝑃 (25 − 𝑃 ) . (1.38)

Its maximum price 𝑃𝑚𝑎𝑥 = 25 pounds and hence the optimal price

25
𝑃𝑜𝑝𝑡 = = 12.5
2

pounds (see equation (1.35)). The corresponding optimized (maximum) revenue

can be found using equation (1.36), but it is easier just to substitute 𝑃𝑜𝑝𝑡 into
(1.38) and obtain

25 25 25 2
𝑅𝑜𝑝𝑡 = (25 − ) = ( ) = 156.25
2 2 2

in thousands of pounds per month.

Comment: If we measured 𝑄𝐷 in books per month, we would have
1.4. OPTIMISATION OF REVENUE 31

𝑄𝐷 = 1000(25 − 𝑃 ) .

Hence the revenue would be

𝑅 = 𝑃 𝑄𝐷 = 1000𝑃 (25 − 𝑃 )

and measured in pounds per month (Note that now 𝛼 = 1000.). The maximum
revenue would be

25 2
𝑅 = 1000 ( ) = 156, 250
2

in pounds per month. Obviously, this is the same results but in different units.
The disadvantage of using these units is longer numbers appearing in the cal-
culations. Here, the amount of calculations is relatively small and hence this
disadvantage is rather minor, but other applications can be much more compu-
tationally intense, and choosing good units can make a difference.
Take Quiz 2 on Minerva (Learning Resources > Quizzes).
32 CHAPTER 1. LINEAR AND QUADRATIC FUNCTIONS
Chapter 2

Exponentiation and power

functions

Exponentiation is a mathematical operation, written as 𝑎𝑏 , where the base 𝑎

and the exponent 𝑏 (also called the power index or just power) are two real
numbers. It is pronounced as 𝑎 raised to the power of 𝑏.
Ultimately, it is based on arithmetic operations, but goes a bit beyond. We
start with the simplest case where 𝑏 is a natural number and then gradually
expand it to the other types of numbers.

2.1 Natural exponents

Natural numbers, 𝑁 = {1, 2, 3, … }, are usually denoted via Latin letters
𝑛, 𝑚, 𝑘, 𝑖 or 𝑗.

By definition,

𝑎𝑛 = 𝑎
⏟×⏟⏟𝑎⏟×⋯⏟⏟×⏟
𝑎. (2.1)
n times

Thus, 𝑎 is multiplied by itself (n-1) times.

For example,

33
34 CHAPTER 2. EXPONENTIATION AND POWER FUNCTIONS

34 = 3 × 3 × 3 × 3 .

According to this definition

𝑎1 = 𝑎 . (2.2)

Using the properties of multiplication, it is easy to prove that

Rules of bases

(𝑎𝑏)𝑛 = 𝑎𝑛 𝑏𝑛 . (2.3)

𝑎 𝑛 𝑎𝑛
( ) = 𝑛. (2.4)
𝑏 𝑏

and

Rules of exponents (power indices)

Addition:
𝑎𝑚 × 𝑎𝑛 = 𝑎𝑛 × 𝑎𝑚 = 𝑎𝑚+𝑛 . (2.5)

Subtraction:

⎧𝑎𝑚−𝑛 if 𝑚 > 𝑛;
𝑎𝑚 {
𝑛
= ⎨1 if 𝑚 = 𝑛; (2.6)
𝑎 {1/𝑎𝑛−𝑚
⎩ if 𝑚 < 𝑛.

Multiplication:
2.1. NATURAL EXPONENTS 35

(𝑎𝑚 )𝑛 = 𝑎𝑚𝑛 . (2.7)

For example,

𝑎3 × 𝑎2 = (𝑎 × 𝑎 × 𝑎) × (𝑎 × 𝑎)
= 𝑎 × 𝑎 × 𝑎 × 𝑎 × 𝑎 = 𝑎5 .

Since there are five 𝑎’s multiplied together, the result is 𝑎5 .

Similarly, for any natural 𝑛 and 𝑚,

𝑎𝑛 × 𝑎𝑚 = 𝑎
⏟×⏟⏟𝑎⏟
×⋯ ×⏟
⏟⏟ 𝑎 = 𝑎𝑛+𝑚 ,
n+m times

which proves the addition law for all natural exponents.

Comment: Exponentiation to small natural powers is most common and em-
ploys special names. For example, 𝑦2 is called 𝑦 squared, and 𝑥3 is called 𝑥
cubed.
Examples of application,

53 = 5 × 5 × 5 = 125 .

2 4 24 16
( ) = 4 = .
3 3 81

𝑎 × 2𝑎 × 4𝑎 × 3𝑎 = (1 × 2 × 4 × 3) × (𝑎 × 𝑎 × 𝑎 × 𝑎) =
= 24𝑎4 .

𝑎5 𝑎3 × 𝑎2 𝑎5
2
= = 𝑎3 or just = 𝑎5−2 = 𝑎3 .
𝑎 𝑎2 𝑎2
𝑎2 𝑎2 1 𝑎2 1 1
5
= 2 = 3 or just = 5−2 = 3 .
𝑎 𝑎 × 𝑎3 𝑎 𝑎5 𝑎 𝑎

(𝑎2 )3 = 𝑎2 × 𝑎2 × 𝑎2 = 𝑎2+2+2 = 𝑎2×3 = 𝑎6 .

36 CHAPTER 2. EXPONENTIATION AND POWER FUNCTIONS

(𝑥4 )3 = 𝑥4×3 = 𝑥12 .

Extensions to other types of exponents is carried out in such a way

that all the rules for natural exponents remain true.

2.2 Integer exponents

The set of integer numbers 𝑍, includes all natural numbers, 0, and all negative
whole numbers, −𝑛 where 𝑛 ∈ 𝑁 . The extension is carried out via the definitions

Zero exponent:
For any 𝑎 ≠ 0,
𝑎0 = 1 . (2.8)

Negative exponent:
For any 𝑛 ∈ 𝑁 and 𝑎 ≠ 0,

1
𝑎−𝑛 = . (2.9)
𝑎𝑛

The condition 𝑎 ≠ 0 is introduce to avoid the undefined division by zero.

Using this definitions one can prove that exponentiation with the negative expo-
nents satisfies the same rules as naturals. For example, let us prove the extension
of rule (2.3) for positive exponents to negative exponents, (𝑎𝑏)−𝑛 = 𝑎−𝑛 𝑏−𝑛 . By
the definition of negative exponent,

1
(𝑎𝑏)−𝑛 = .
(𝑎𝑏)𝑛
By the rule (2.3),

1 1 1 1
𝑛
= 𝑛 𝑛 = 𝑛 𝑛.
(𝑎𝑏) 𝑎 𝑏 𝑎 𝑏

Next, we apply the definition of negative exponent again and obtain

2.2. INTEGER EXPONENTS 37

1 1
= 𝑎−𝑛 𝑏−𝑛 .
𝑎𝑛 𝑏 𝑛

Hence,

(𝑎𝑏)−𝑛 = 𝑎−𝑛 𝑏−𝑛 .

This is the required extension of rule (2.3) to negative exponents.

The subtraction rule (2.6), can now be written in the more compact form

Subtraction rule:

𝑎𝑚
= 𝑎𝑚−𝑛 , (2.10)
𝑎𝑛

for any 𝑚, 𝑛 ∈ 𝑁 , provided 𝑎 ≠ 0.

This is because for 𝑎 ≠ 0

1 1 1
= −(𝑚−𝑛) = 1 = 𝑎𝑚−𝑛 .
𝑎𝑛−𝑚 𝑎 𝑎𝑚−𝑛

and because for 𝑚 = 𝑛

𝑎𝑚−𝑛 = 𝑎0 = 1 .

Moreover, when 𝑛, 𝑚 ∈ 𝑍, the subtraction rule simply follows from the addition
rule

𝑎𝑚 𝑎𝑛 = 𝑎𝑛+𝑚 .

Indeed,

𝑎𝑚
= 𝑎𝑚 𝑎−𝑛 = 𝑎𝑚+(−𝑛) = 𝑎𝑚−𝑛 ,
𝑎𝑛

where we applied the addition rule at the second step.

Examples of application.
38 CHAPTER 2. EXPONENTIATION AND POWER FUNCTIONS

1. By the definition of non-positive exponents and the rules of natural

exponents

𝑎5
𝑎5 𝑎−3 = = 𝑎5−3 = 𝑎2 .
𝑎3
By direct application of the addition rule to integer exponents

𝑎5 𝑎−3 = 𝑎5+(−3) = 𝑎2 .

2. By the definition on non-positive exponents of the rules of natural

exponents

𝑎3
𝑎−5 𝑎3 = = 𝑎3−5 = 𝑎−2 .
𝑎5
By direct application of the addition rule to integer exponents

𝑎−5 𝑎3 = 𝑎−5+3 = 𝑎−2 .

25
3. 52 3−2 = .
9

52
4. = 25 × 9 = 225 .
3−2

2.3 Rational exponents

The set of rational numbers 𝑄, is made out of all numbers which can be expressed
as 𝑖/𝑗, where 𝑖, 𝑗 ∈ 𝑍 and 𝑗 ≠ 0.
Any such number 𝑝 can also be written as

𝑝 = 𝑙 + 𝑚/𝑛 , (2.11)

where

1) 𝑙 ∈ 𝑍 (integer);
2) 0 < 𝑚/𝑛 < 1 ;
3) 𝑛, 𝑚 ∈ 𝑁 (natural) and have no common factors.
2.3. RATIONAL EXPONENTS 39

For example,

11/4 = (8 + 3)/4 = 2 + 3/4 ;

−11/4 = (−12 + 1)/4 = −3 + 1/4 ;
10/4 = (8 + 2)/4 = 2 + 2/4 = 2 + 1/2.

Demanding applicability of the addition rule, we write

𝑎𝑙+𝑚/𝑛 = 𝑎𝑙 × 𝑎𝑚/𝑛 . (2.12)

We know how to compute 𝑎𝑙 . Hence, all what is left to figure out is the expo-
nentiation of 𝑎𝑚/𝑛 .
We start with the exponents of the form 𝑝 = 1/𝑛 where 𝑛 ∈ 𝑁 . If the multipli-
cation rule is preserved then

(𝑎1/𝑛 )𝑛 = 𝑎𝑛/𝑛 = 𝑎 . (2.13)

This suggests to define 𝑎1/𝑛 as the real number which gives us 𝑎 when it is
raised to the power of 𝑛. However, some further qualifications are required.
First, there are no such numbers when 𝑎 < 0 and 𝑛 is even. (For example,
𝑏 × 𝑏 ≥ 0 for any real number 𝑏. Hence, (𝑎)1/2 does not exits for 𝑎 < 0.) Hence,
for even 𝑛, 𝑎1/𝑛 is defined only for 𝑎 > 0 .
Second, there are two such numbers when 𝑎 > 0 and 𝑛 is even. For example,
22 = 4 and (−2)2 = 4, and hence both -2 and 2 could be taken as 41/2 . To avoid
this non-uniqueness, we accept only positive numbers for 𝑎1/𝑛 in such
cases.
√
When 𝑛 ∈ 𝑁 , 𝑎1/𝑛 is also called 𝑛-th root of 𝑎 and denoted as 𝑛 𝑎.
40 CHAPTER 2. EXPONENTIATION AND POWER FUNCTIONS

For any natural number 𝑛 ≥ 2,

1 √
𝑎𝑛 ≡ 𝑛
𝑎. (2.14)

For even 𝑛, it is defined only for 𝑎 ≥ 0.

Examples:

• Since 16 = 2 × 2 × 2 × 2 = 24 , we have 161/4 = 2.

• Since 27 = 3 × 3 × 3 = 33 , we have 271/3 = 3.

• Since −27 = (−3) × (−3) × (−3) = (−3)3 , we have (−27)1/3 = −3.

Comment:
√
𝑚
𝑎 of most numbers is an irrational number with an infinite non-repeating dec-
imal expansion. As a result, the decimal expansion of any such square root can
only be computed approximately. This is a topic of computational mathematics.
For example,

√
5 = 2.23606797749978969640917366873 … ,

where we show only the first 30 digits. □

Based on the multiplication rule, we may write

1 √
𝑎3/4 = 𝑎 4 ×3 = (𝑎1/4 )3 = ( 4 𝑎)3 .

and

1
√
4
𝑎3/4 = 𝑎3× 4 = (𝑎3 )1/4 = 𝑎3 .

This shows how to exponentiate with the exponents √ in the form 𝑚/𝑛. Note that
in this example, it is required that 𝑎 ≥ 0 so that 4 𝑎 is applied to a non-negative
number.
2.3. RATIONAL EXPONENTS 41

For any rational number 𝑚/𝑛, where 𝑛, 𝑚 ∈ 𝑁 have no common factors and
0 < 𝑚/𝑛 < 1,

𝑚 √ √
𝑎 𝑛 = ( 𝑛 𝑎)𝑚 = 𝑛 𝑎𝑚 . (2.15)

For even 𝑛, the exponentiation is defined only when 𝑎 ≥ 0.

√ √
Example 2.1. While the expressions ( 𝑛 𝑎)𝑚 √ and = 𝑛 𝑎𝑚 are identical, when
calculating by hand it is often easier to use ( 𝑛 𝑎)𝑚 . This because it involves
taking 𝑛-th root of a smaller number.
3
Evaluate 16 4 using this avise and going against it.

Solution

3 √
16 4 = ( 16)3 = 23 = 8 .
4

If, however, we proceeded the other way, we would obtain

3
√
4
√
4
16 4 = 163 = 4096 .

At this point, most of us would reach for a calculator. □

Example 2.2. Simplify as much as possible the expressions

(3𝑎 + 2)3
a) √ ;
3𝑎 + 2
√
(𝑎2 + 4𝑎 + 4) 𝑎 + 2
b) .
(𝑎 + 2)3

Solution

a)

(3𝑎 + 2)3 (3𝑎 + 2)3

√ = =
3𝑎 + 2 (3𝑎 + 2)1/2
1 5
= (3𝑎 + 2)3− 2 = (3𝑎 + 2) 2 ;
42 CHAPTER 2. EXPONENTIATION AND POWER FUNCTIONS

√ 1
b) Here, we first notice that 𝑎2 + 4𝑎 + 4 = (𝑎 + 2)2 and 𝑎 + 2 = (𝑎 + 2) 2 .
Hence

√
(𝑎2 + 4𝑎 + 4) 𝑎 + 2 (𝑎 + 2)2 (𝑎 + 2)1/2
= =
(𝑎 + 2)3 (𝑎 + 2)3
1
= (𝑎 + 2)2+ 2 −3 =
1
= (𝑎 + 2)−1/2 = √ .
𝑎+2

2.4 Irrational exponents

One can also extend exponentiation to the case of irrational exponents (𝑎𝑏 where
𝑏 ∈ 𝑃 ). This is the most involved case, which we cannot discuss here at depth.
To say something, one may use

𝑎𝑏 = 𝑒𝑏 ln(𝑎) , (2.16)

where 𝑒 is Euler’s number (the base of natural power functions and logarithms)
and ln(𝑎) is the natural logarithm of 𝑎 > 0. We will discuss them at some length
later.
Example 2.3. Solve the equation

23𝑥−1 = 4 .

Solution
Since 4 = 22 , this equation reads

23𝑥−1 = 22 .

This is satisfied only if the exponents of the both sides are the same. Hence

3𝑥 − 1 = 2 and hence 𝑥 = 1. □