1 Matrices 1

Matrices
As for everything else, so for a mathematical theory – beauty can be
perceived but not explained.
Arthur Cayley, 1883

Figure 1.1 shows a pack of cards. Initially the cards are piled up neatly forming a
cuboid, but two other arrangements are shown.

Figure 1.1

●
? Describe what has happened in words. Can you suggest ways of representing this
symbolically?

In this chapter you will learn about matrices, and how they give you the ability to
use algebraic techniques in geometrical and other situations. Matrices are often
used when creating effects on TV – particularly when the picture is rotated,
flipped, enlarged or reduced.

Matrices

Figure 1.2 shows the number

Dover
of ferry crossings per hour 1
offered by a certain ferry Folkstone 1
2 Zeebrugge
company from English to
Continental ports during 2
Calais
1
peak times.
Boulogne

Figure 1.2 1
 This information can be represented as an array of numbers, called a matrix
1 (plural matrices).
To

Matrices
Z C B
D 1 2 1
From
F ( 0 2 1 )
●
? What information does the zero in the matrix give?

Any rectangular array of numbers is known as a matrix. It is usual to represent

matrices by capital letters, often in bold print. (In handwriting, use a capital letter
with a wavy line under it.)

A matrix consists of rows and columns, and the entries in the various cells are
1 2 1
known as elements. The matrix M =
0 2 1 ( )
representing the ferry crossings has
six elements, arranged in two rows and three columns. We describe M as a 2 × 3
matrix, and this is known as the order of the matrix. It is important to note that
you state the number of rows first, then the number of columns.
1 3 –2
(
Matrices such as –7 5 4 and
9 0 6
)
3 2
5 3 ( )
which have the same number of rows

as columns are known as square matrices.

The matrix O = (00 00) is known as the 2 × 2 zero matrix.

Working with matrices

The number of crossings offered per hour by the ferry company is the same for
every hour in the ten-hour period from 9 am until 7 pm.

●
? What matrix represents the total number of ferry crossings offered by this
company between the hours of 9 am and 7 pm?

The number of ferry crossings per hour offered by a second ferry company
between the same ports is shown by the following matrix.

(32 1 1
1 0 )
2
 ●
? What matrix represents the total number of ferry crossings per hour offered by 1
the two companies?

Exercise 1A
This example shows that matrices can be multiplied by a number or added
together.
To multiply a matrix by a number, multiply each element by that number.

3× ( 12 34 ) = (36 129 )
You can add or subtract matrices of the same order element by element:

( 12 34 ) + ( –25 07 ) = (60 113 )

1 4 4 –2 –3 6
( )( )( )
2 5 – –3 7 = 5 –2
3 6 5 6 –2 0

But ( 12 34 ) + (10) cannot be evaluated because the matrices are not of the same
order. They are incompatible, or do not conform.

●
? Explain why matrix addition is:
(i) commutative, i.e. A + B = B + A
(ii) associative, i.e. A + (B + C) = (A + B) + C.

Equality of matrices

Two matrices are equal if and only if they have the same order and each element
in one matrix is equal to the corresponding element in the other matrix. If, for
example,

A= ( 12 34 ) B= ( 13 24 ) C= ( 12 3 0
4 0 ) D= ( 12 34 )
then A and D are equal but no other pair of matrices from A, B, C and D is equal.

EXERCISE 1A For Questions 1 and 2 use:

A= ( 13 –12 ) B= ( –21 –43) C= ( 23 1 5

0 1 )
5 2 –1 6
D= ( –3 6 0
1 2 –1 ) E=
( )
0 –1
–2 4
F=
( )
3 2
8 –4 3
 1 Write down the order of these matrices.
1 (i) A (ii) C (iii) E
2 Calculate, if possible, the following.

Matrices
(i) A + B (ii) C – D (iii) A+F
(iv) 3B (v) 3E – 2F (vi) 2D – A
3 The diagram below shows the number of direct flights on one day offered by
an airline between cities P, Q, R and S.
The same information is also given in the partly-completed matrix X.
2
Q

To
P 1
2 P Q R S
1
From P 0 2 1 0

( )
2

1
1 Q 1
X=
R
R
S 1 S

(i) Copy and complete the matrix X.

A second airline also offers flights between these four cities. The
following matrix represents the total number of direct flights offered
by the two airlines.
0 2 3 2

( ) 2
2
1
0
2
0
2
0
3
1
3
0
(ii) Find the matrix Y representing the flights offered by the second airline.
(iii) Draw a diagram similar to the one above, showing the flights offered by
the second airline.
4 Four local football teams took part in a competition in which they each played
each other twice, once at home and once away.
Here is the results matrix after half of the games had been played.
Goals Goals
Win Draw Lose for against
City 2 1 0 6 3
Rangers
Town
United
( 0
2
1
0
0
1
3
1
1
2
4
5
8
3
3
)
4
 (i) The results of the next three matches are as follows.
City 2 Rangers 0 1
Town 3 United 3

Exercise 1A
City 2 Town 4
Find the results matrix for these three matches and hence find the
complete results matrix for all the matches so far.

(ii) Here is the complete results matrix for the whole competition.
4 1 1 12 8

( 1
3
1
1
1
3
4 5 12
2 12 10
2 10 9
)
Find the results matrix for the last three matches (City v United, Rangers v
Town and Rangers v United) and deduce the result of each of these three
matches.

5 A mail-order clothing company stocks a jacket in three different sizes and four
different colours.
17 8 10 15
( )
The matrix P = 6 12 19 3 represents the number of jackets in stock at
24 10 11 6
the start of one week.
2 5 3 0
( )
The matrix Q = 1 3 4 6 represents the number of orders for jackets
5 0 2 3
received during the week.
(i) Find the matrix P – Q.
What does this matrix represent?
What does the negative element in the matrix mean?
A delivery of jackets is received from the manufacturers during the week.

( )
5 10 10 5
The matrix R = 10 10 5 15 shows the number of jackets received.
0 0 5 5
(ii) Find the matrix which represents the number of jackets in stock at the end
of the week after all the orders have been dispatched.
(iii) Assuming that this week is typical, find the matrix which represents sales
of jackets over a six-week period.
How realistic is this assumption?

5
 Transformations
1
You are probably already familiar with several different types of transformation,

Matrices
including reflections, rotations and enlargements. The original shape or point is
sometimes called the object and the new shape or point, after the transformation,
is called the image.
Some examples of transformations are illustrated in figure 1.3.

Reflection in the
y y
line y = x
A′ B′
B

O A x O x

Rotation through 90°

y y
clockwise, centre O
B

O A x O x

A′ B′

Enlargement centre O,
y y
scale factor 2 B′
B

O A x O A′ x

Figure 1.3

In this section you will also meet two other types of transformation which may
be new to you: the two-way stretch and the shear. Examples of these are illustrated
in figure 1.4.

6
 Two-way stretch, scale factor
y 3 y
horizontally, 2 vertically B′
B
2
1

Transformations
O A x O A′ x

y y
Shear parallel to the x axis
B B′

O A x O A′ x

Figure 1.4

So far you have described transformations using words. You have to include
information about mirror lines for reflections, information about centre, angle
and direction for rotations, and so on. For some transformations, all the
information about the transformation can be given in the form of a matrix.
Figure 1.5 shows a flag rotated 90° anticlockwise about the origin.

y y
4 4
Object
3 3
Image
Q
2 R 2
R′ S′
1 S 1
P
Q′ P′
–3 –2 –1 O 1 2 3 x –3 –2 –1 O 1 2 3 x

Figure 1.5

This table shows the effect of the rotation on the co-ordinates of each point on
the flag.

Point Co-ordinates Image point Co-ordinates

O (0, 0) O (0, 0)

P (0, 1) P′ (–1, 0)

Q (0, 2) Q′ (–2, 0)

R (1, 2) R′ (–2, 1)

S (1, 1) S′ (–1, 1) 7
 You can describe the effect of this rotation by saying that the x and y co-ordinates
1 have been exchanged and the sign of the original y co-ordinate has been changed.
You can express this algebraically as

Matrices
x′ = –y
{y′ = x.
Many other transformations, such as reflections and enlargements, can be
expressed as a pair of equations of the form
x′ = ax + cy A transformation of
{y′ = bx + dy
this type is called a
linear transformation.

where a, b, c and d are constants.

You can summarise a pair of equations like this by writing just the coefficients in
a c
the form of a matrix:
b d
. ( )

●
? What are the values of a, b, c and d for the transformation shown in figure 1.5?

ACTIVITY 1.1 Figure 1.6 shows the effect of two other transformations on the flag OPQRS,
shown in figure 1.5.

Reflection in the y axis Enlergement, scale factor 2, centre O

y y
Q′
4 4 R′

3 3
R′ 2 Q′ 2 S′
P′
S′ 1 P′ 1

–3 –2 –1 O 1 2 3 x –3 –2 –1 O 1 2 3 x

Figure 1.6

In each case, find the co-ordinates of the points P′, Q′, R′ and S′ and express the
x′ = ax + cy
transformation in the form {
y′ = bx + dy
.
Hence find the matrix of each transformation.

Straight lines

The diagrams in this chapter so far have shown the images of straight lines as
straight lines. You should assume that this is the case for any transformation
which can be represented by a matrix. So the image of a line through the points A
8 and B is the line through the image points A′ and B′.
EXAMPLE 1.1 A transformation (an example of a shear) maps points as follows:
● each point is moved parallel to the x axis
1

Transformations
● each point is moved twice its distance from the x axis

● points above the x axis are moved to the right

● points below the x axis are moved to the left.

Figure 1.7 shows a flag, and points P and Q and their images P′ and Q′.

P is 1 unit above the

x axis, Q is 2 units
below the x axis.

10

Figure 1.7

(i) Copy the diagram and draw the image of the flag.
(ii) Show that the image of (x, y) is the point (x + 2y, y) and find the matrix
which represents this transformation.

SOLUTION

(i) y
4 B and C are 3 units above the
B′ C′ x axis, so move 6 units to the right;
2 D and E are 1 unit above the x axis,
D′
E′ so move 2 units to the right. A is
1 unit below the x axis, so moves
–2 O 2 4 6 8 10 x 2 units to the left.
A′
–2

Figure 1.8

(ii) The distance of the point (x, y) from the x axis is y units.
The point (x, y) moves 2y units parallel to the x axis so its new position is
(x + 2y, y).
(Notice that if y is negative, then this means moving to the left.)
The transformation can be written as

{ x′y′ == 0x1x ++ 1y2y

so the matrix which represents this transformation is (10 21).
9
 EXAMPLE 1.2 Draw the triangle ABC, where A is the point (1, 0), B is the point (3, 0) and C is
1 the point (3, 1).

Draw the image A′B′C′ of the triangle under the transformation represented by

Matrices
0 1
the matrix ( )
1 0
and describe the effect of the transformation.

SOLUTION

The matrix ( 01 10 ) represents the transformation { x′y′ == xy .

The co-ordinates of A′ are therefore (0, 1), the co-ordinates of B′ are (0, 3) and
the co-ordinates of C′ are (1, 3).

y
y=x
4
B′
3 C′

2
C
1 A′
A B
O 1 2 3 4 x

Figure 1.9

The effect of the transformation is a reflection in the line y = x.

ACTIVITY 1.2 Investigate the effects of the transformation matrices ( 10 00 ) and ( 00 01 ).

Notice that substituting x = 0 and y = 0 into {y′x′ == bxax ++ dycy gives {y′x′== 00.
This means that whenever you can represent a transformation by a matrix in the way
described above, the image of the origin O is O. Thus you can only use this method
of representing a transformation if the origin is a point which does not move.
● Rotations can only be represented in this way if the centre of rotation is at O.
● Reflections can only be represented in this way if the mirror line passes
through O.

● Enlargements can only be represented in this way if the centre of enlargement

is at O.

●
? Explain why it is not possible to represent translations by matrices in this way.
10
 The columns of a matrix
1
It is useful to look at the effect of the transformation represented by the matrix
a c
( )

Transformations
M= on the points I and J with co-ordinates (1, 0) and (0, 1).
b d
x′ = ax + cy
Substituting x = 1, y = 0 in the defining equations
y′ = bx + dy {
you find that I′,
the image of I, has co-ordinates (a, b), which is the first column of matrix M.
Similarly J′, the image of J, has co-ordinates (c, d), the second column of M.
The connection between the co-ordinates of I′ and J′ and the matrix representing
a transformation provides you with a quick and usually easy way of finding
the matrix.

EXAMPLE 1.3 Find the matrix which represents a rotation through angle θ anticlockwise about
the origin.

SOLUTION

Figure 1.10 shows the points I(1, 0) and J(0, 1) and their images I′ and J′ after
rotation through angle θ anticlockwise about the origin.

y
You can see that OI′ = OI = OJ′ = OJ = 1.

J(0, 1) By symmetry, if the co-ordinates of I′ are (a, b)

b then the co-ordinates of J′ are (–b, a).
J′(–b, a)
a
I′(a, b)
a 1 1
b

O I(1, 0) x

Figure 1.10

a
From Figure 1.10, cos θ = – ⇒ a = cos θ
1
b
sin θ = – ⇒ b = sin θ.
1
Therefore I′ is (cos θ, sin θ) and J′ is (–sin θ, cos θ).
θ –sin θ
The transformation matrix is ( cos
sin θ cos θ )
.

●
? What matrix represents a rotation through angle θ clockwise about the origin?
11
 Make sure that you are able to spot a rotation matrix. Remember that either or
1 both of cos θ and sin θ may be negative for values of θ between 90° and 360°.

Matrices
ACTIVITY 1.3 Investigate rotation matrices for
(i) angles between 90° and 180°
(ii) angles between 180° and 270°
(iii) angles between 270° and 360°.

● Transformations in three dimensions

So far your work has applied to transformations of sets of points from a plane
(i.e. two dimensions) to the same plane. Similar procedures apply when you
transform a set of points within three-dimensional space.

{
x′ = x
The shear illustrated in figure 1.11 is defined by the equations y′ = y + z
z′ = z

1 0 0
( )
so you can represent this transformation by the matrix 0 1 1 .
0 0 1

Figure 1.11

Figure 1.12 illustrates a reflection in the plane y = 0 (i.e. the plane containing the
x and z axes).
1 0 0
( )
This transformation is represented by the matrix 0 –1 0 .
0 0 1

Figure 1.12
12
EXERCISE 1B 1 The diagram shows a triangle with y
vertices at O, A(1, 2) and B(0, 2). 3 1
B
2 A

Exercise 1B
1

–3 –2 –1 O 1 2 3 x

For each of the transformations below

(a) draw a diagram to show the effect of the transformation on triangle OAB
(b) give the co-ordinates of A′ and B′, the images of points A and B
(c) find expressions for x′ and y′, the co-ordinates of P′, the image of a general
point P(x, y)
(d) find the matrix of the transformation.
(i) Enlargement, centre the origin, scale factor 3
(ii) Reflection in the x axis
(iii) Reflection in the line x + y = 0
(iv) Rotation 90° clockwise about O
1
(v) Two-way stretch, scale factor 3 horizontally and scale factor –
2 vertically

2 The triangle OAB in question 1 is rotated through 20° anticlockwise about

the origin.
cos θ –sin θ
Use the transformation matrix (
sin θ cos θ )
to find the co-ordinates of A′
and B′, the images of A and B under this transformation.
Give your answers to 3 decimal places.
3 Each of the following matrices represents a rotation about the origin.
Find the angle and direction of rotation in each case.

( )
1 – ––3
–
2 2
(i)
––3 1
–
(ii) ( 0.574
0.819 0.574 )
–0.819

2 2
1 1 – ––3 1

( ) ( )
– –– –– ––
2 2 2 2
(iii) (iv)
1
– –– 1
– –– 1 – ––3
–
2 2 2 2

4 For each of the matrices below

(a) draw a diagram to show the effect of the transformation it represents on
the triangle OAB in question 1
(b) give the co-ordinates of A′ and B′
(c) give a full description of the transformation.

(i) (–10 10 ) (ii) (20 03 ) (iii) (0∑-q 0∑-q )

(iv) (10 31 ) (v) (–10 –10) (vi) (0.6
0.8 0.6 )
–0.8
13
 5 The unit square OABC has its vertices at (0, 0), (1, 0), (1, 1) and (0, 1).
4 3
1 OABC is mapped to OA′B′C′ by the transformation defined by the matrix ( )
5 4
.
Find the co-ordinates of A′, B′ and C′ and show that the area of the shape has

Matrices
not been changed by the transformation.
6 A transformation maps P to P′ as follows:
● each point is mapped on to the line y = x
● the line joining a point to its image is parallel to the y axis.
Find the co-ordinates of the image of the point (x, y) and hence show that this
transformation can be represented by means of a matrix.
What is that matrix?

● The remaining questions relate to enrichment material.

7 Find the matrices which represent the following transformations of three-

dimensional space.
(i) Enlargement, centre O, scale factor 3
(ii) Rotation 180° about the z axis
(iii) Reflection in the plane x = 0

8 Describe the transformations represented by these matrices.

1 0 0 1 0 0 2 0 0
(i)
( )
0 0 1
0 –1 0
(ii) 0 1
0 0 –1
( )
0 (iii) 0 3 0
0 0 ∑-q
( )
Multiplying matrices

The co-ordinates (x, y) of a point P in two dimensions can be written as the

x
()
column vector y . This is called the position vector of the point P. A column
vector is a 2 × 1 matrix and is sometimes called a column matrix. Column vectors
are often represented by lower case bold letters such as p or v. In handwriting, a
lower case letter with a wavy line underneath is used.
Using this notation allows you to write the effect of a transformation as a matrix
2 5
product. The transformation defined by the matrix M =
x
4 3 ( )
maps the point P
x′
()
with position vector p = y to the point P′ with position vector p′ = y .
′ ( )
2x + 5y
You can write p′ = ( )
4x + 3y
.

The 2 × 1 matrix p′ is the product of the 2 × 2 matrix M with the 2 × 1 matrix p,

in that order.

14 (24 53 ) (xy ) = (2x4x ++ 5y3y )

EXAMPLE 1.4 Calculate (13 64 )(58) . 1

Multiplying matrices
SOLUTION

1 × 5 + 6 × 8 = 53

Row from left matrix with

1 6 5 53
( 3 4 )( )
8
= ( )
47
column from right matrix.

3 × 5 + 4 × 8 = 47

Figure 1.13

A similar technique applies to all matrix multiplications. You use each row of the
first (i.e. left) matrix with each column, in turn, of the second matrix. Figure 1.14
shows the steps used when multiplying a 2 × 2 matrix by a 2 × 3 matrix. The
product is another 2 × 3 matrix.

3 × 6 + 5 × 1 = 23

3 × 7 + 5 × 8 = 61

3 × 9 + 5 × 0 = 27

( 32 54 )( 61 7 9
8 0 ) = (2316 61 27
46 18 )
2 × 6 + 4 × 1 = 16

2 × 7 + 4 × 8 = 46

2 × 9 + 4 × 0 = 18

Figure 1.14

If you multiply a 3 × 4 matrix (on the left) by a 4 × 2 matrix (on the right) similar
rules apply: the product is a 3 × 2 matrix. For example:
5 1

)( )(
1 2 4 7 39 59
( –3 5 0 1
4 2 3 5
–6 4
8 9
2 2
= –43 19 .
42 49
)
15
 Matrix products and transformations
1 Figure 1.15 shows the flag ABCD and its image A′B′C′D′ after applying the
0.8 –0.6
rotation matrix ( .)
Matrices
0.6 0.8

Figure 1.15

You can write the co-ordinates of A, B, C and D as the single 2 × 4 matrix

0 1 1 1
(2 2 1 0 )
. This provides a quick method of finding the co-ordinates of

A′, B′, C′ and D′.

ACTIVITY 1.4 By calculating the product ( 0.8

–0.6 0 1 1 1
0.6 )(
0.8 2 2 1 0 )
, check that the points A′, B′, C′
and D′ in figure 1.15 have been plotted correctly.

Other applications of matrix multiplication

The use of matrices to describe transformations is just one of the many

applications of matrices. You saw some other uses of matrices at the start of this
chapter and others are explored in Exercise 1C.

EXAMPLE 1.5 A pizza company wants to set up a computer program to keep track of orders.
Pizzas are available in three sizes: Regular, Large and Family, and four
combinations of toppings: Hawaiian, Seafood, Meat Feast and Vegetarian.
The number of pizzas of each type is entered in the form of a 4 × 3 matrix P.
One day the matrix P, representing the orders, is as follows.
Toppings
H S M V
R 2 3 0 1
Sizes L
F
( 5 7 8 4
6 4 3 3
)
(i) Find a matrix A so that the product PA gives the number of each size of base
required, and hence find the matrix representing the number of each size of
16
base required on this day.
(ii) A regular pizza requires 1 quantity of topping, a large pizza requires 1.5
quantities of topping and a family-size pizza requires 2 quantities of topping. 1
Find a matrix B so that the product BP gives the quantities of each type of

Multiplying matrices
topping required, and hence find the matrix representing the quantities of
each type of topping required on this day.
(iii) The cost of each size of pizza, in pounds, whatever the topping, is given by
the matrix C = (2.5 3.5 4).
What matrix product will give the total cost of the pizzas?
Find the total cost of this day’s pizzas.

SOLUTION

(i) You need to find a matrix A so that

2 3 0 1 2+3+0+1
( ) (
5 7 8 4 A= 5+7+8+4 .
6 4 3 3 6+4+3+3
)
1
The matrix A is therefore

1
1
1
1
().

)( ) ( )
2 3 0 1 6
(
5 7 8 4
6 4 3 3
1
1
1
= 24
16

(ii) You need to find a matrix B so that

2 3 0 1
(
6 4 3 3
)
B 5 7 8 4 = (1 × 2 + 1.5 × 5 + 2 × 6 ... ... ...).

The matrix B is therefore (1 1.5 2).

2 3 0 1
( )
(1 1.5 2) 5 7 8 4 = (21.5 21.5 18 13)
6 4 3 3
(iii) You already know that the matrix product PA gives the total number of each
size of pizza in the form of a 3 × 1 matrix. To find the total cost, you need to
pre-multiply this product by C.
The total cost is given by the product CPA.
1

( )( )
2 3 0 1
1
CPA = (2.5 3.5 4) 5 7 8 4
1
6 4 3 3
1
6

= (163)
()
= (2.5 3.5 4) 24
16

17
The total cost is £163.
 Properties of matrix multiplication
1
Matrices must be conformable for multiplication

Matrices
To be able to multiply two matrices together, the number of columns in the first
(left) matrix must equal the number of rows in the second matrix. For example, if
the first matrix is 3 × 4, the second must be 4 × something. You say that the
matrices need to conform.
Generally, if M is of order p × q, and N is of order q × r, the product MN exists
and is of order p × r. It may be helpful to think of the rules for joining dominoes
end to end.

p q q r gives p r

Matrix multiplication is not commutative

You have already seen that matrix addition is commutative, i.e. A + B = B + A.

However, matrix multiplication is not commutative.

If A = (32 17 ) and B = (41 –25 ), the product AB is given by

3 1 4 –2 13 –1
AB = (
2 7 )( 1 5 ) ( 15 31 )
=

and the product BA is given by

BA = (41 –25 )(32 17 ) = (138 –10

36 )
.

So AB ≠ BA. It matters which way round you write your matrices.

Sometimes the product MN exists but the product NM does not exist, because
the number of columns in N does not match the number of rows in M.

●
? If A =
a c
b d( )
e h
( )
, B = f i and C =
g j
k m o
l n p ( )
, which of the products AB, BA, AC,

CA, BC and CB exist?

Matrix multiplication is associative

Whenever matrices P, Q, R are conformable for multiplication, (PQ)R = P(QR).

This is the associative property of matrix multiplication.

18
ACTIVITY 1.5 Using P = ( ba dc ), Q =( ef hg ) and R = ( ij kl ), find 1
(i) PQ (ii) (PQ)R (iii) QR (iv) P(QR)

Properties of matrix multiplication

and so demonstrate that matrix multiplication is associative.

Note

Since (PQ)R = P(QR), the brackets are not needed and the product can be written
simply as PQR.

You will see how transformations justify the associativity of matrix multiplication
on page 25.

● Matrix multiplication is distributive over matrix addition

Provided the matrices P, Q and R conform so that the sums and products exist,
P(Q + R) = PQ + PR
and (P + Q)R = PR + QR.
These two properties are what we mean when we say that matrix multiplication is
distributive over matrix addition.

ACTIVITY 1.6 Using P = ( ba dc ), Q = ( ef hg ) and R = ( ij kl ), find

(i) P(Q + R) (ii) PQ + PR
(iii) (P + Q)R (iv) PR + QR
and so demonstrate the distributive property of matrix multiplication over
matrix addition.

The identity matrix

Whenever you multiply a 2 × 2 matrix M by I = (10 01 ) the product is M.

It makes no difference whether you pre-multiply by I, as in

(10 01 )(37 –25 ) = (37 –25 )

or post-multiply by I, as in

(ab dc )(10 01 ) = (ab dc ).

For multiplication of 2 × 2 matrices, I behaves in the same way as the number 1
when dealing with the multiplication of real numbers. I is known as the 2 × 2
identity matrix. 19
 1 0 0
1
( )
The 3 × 3 identity matrix, I3, is 0 1 0 .
0 0 1

Matrices
ACTIVITY 1.7 Show that I3M = MI3 = M for a 3 × 3 matrix of your choice.

●
? Why is there no such thing as a 2 × 3 identity matrix?

ACTIVITY 1.8 Find out how to input matrices into your calculator. Particularly notice whether
the calculator encourages you to input by rows or columns.
Find out how to add and multiply matrices on your calculator.

Historical note

The multiplication of matrices was first fully defined by Arthur Cayley (1821–95) in 1858. Cayley’s
mathematical talent was noticed while he was at school (in London), and his first mathematical paper
was published in 1841, while he was an undergraduate at Cambridge. He worked as a lawyer for some
14 years, refusing more cases than he accepted as he only wanted to earn sufficient to enable him to
get on with ‘his work’; during this period he published nearly 200 mathematical papers. In 1863 he
returned to Cambridge as a professor. As well as his work on matrices, he developed the geometry
of n-dimensional spaces and is known for his work on the theory of invariants, much of this in
collaboration with his life-long friend, James Joseph Sylvester (1814–97). It was Sylvester who, in
1850, coined the word ‘matrix’, Latin for ‘womb’; in geology a matrix is a mass of rock enclosing
gems, so in mathematics a matrix is a container of (valuable) information. It was at about this time that
Florence Nightingale was one of Sylvester’s students.

EXERCISE 1C In questions 1–3 use:

A= (32 14 ) B= (–32 57 ) C= (25 3 4

7 1 )
3 4 4 7 3 7 –5
D= 7 0
( )
1 –2
( )
E = 3 –2
1 5
(
F= 2 6 0
–1 4 8
)
1 Calculate, if possible, the following.
(i) AB (ii) CA (iii) BC
(iv) CD (v) DC (vi) AF
(vii) BE (viii) 4F + EC (ix) EA
(x) FE (xi) EF (xii) A2

2 By calculating, if possible, AB, BA, AD and DA demonstrate that matrix

multiplication is not commutative.
20
3 Demonstrate the associative property of matrix multiplication by calculating
both (AC)F and A(CF). 1
4 The diagram shows a flag.

Exercise 1C
y

O 1 2 x

(i) Write down the co-ordinates of the five points of the flag.
Hence write down a 2 × 5 matrix to describe the flag.
–0.6 0.8
(ii) The flag is to be transformed using the matrix (
0.8 0.6 )
.
Use matrix multiplication to find a 2 × 5 matrix to describe the image of
the flag.
(iii) Draw the original flag and its image on one diagram, and describe the
transformation in words.
5 The diagram shows a rectangle.

–2 –1 O 1 2 x

(i) Write down the co-ordinates of the four points of the rectangle.
Hence write down a 2 × 4 matrix to describe the flag.
1 0
(ii) The rectangle is to be transformed using the matrix
2 1 ( )
.
Use matrix multiplication to find a 2 × 4 matrix to describe the image of
the rectangle.
(iii) Draw the original rectangle and its image on one diagram, and describe
the transformation in words.
6 (i) Choose a shape of your own. Call it S. Write down a matrix to describe S.
2 1
(ii) Transform S using the matrix M =
diagram to show S and S′.
(
1 –2 )
to form the image S′. Draw a

(iii) Calculate M2.

(iv) Describe as simply as you can the single transformation represented by M2.

21
1 7 The matrix S = ( 57
0 1 6 2
8 4 3 9 )contains the numbers of first and second class
stamps used in an office each day last week. (Top row denotes first class.)

Matrices
(i) Find a matrix D such that DS gives the total number of stamps used each day.
(ii) Find a matrix N such that SN gives the total number of each type of stamp
used in the week.
(iii) First class stamps cost 28 pence and second class stamps cost 21 pence.
Find a way of calculating the total cost of last week’s stamps using only
matrix multiplication.
8 The stylised map below shows the bus routes in a holiday area. Lines
represent routes that run each way between the resorts. Arrows indicate
one-way scenic routes.
M is the partly completed 4 × 4 matrix which shows the number of direct
routes between the various resorts.

To
A B C D
From A 1 1 2 0
M=
B
C
D
( )
(i) Copy and complete M.
(ii) Calculate M2 and explain what information it contains.
(iii) What information would M3 contain?
9 The diagram shows the start of the plaiting process, using three strands a, b
and c.

22
 The process has only two steps, repeated alternately:
A: cross the top strand over the middle strand 1
B: cross the middle strand under the bottom strand.

Exercise 1C
a
()
At Stage 0 the order of the strands is given by s0 = b .
c
0 1 0
(i)
( )
Show that pre-multiplying s0 by the matrix A = 1 0 0 gives s1, the matrix
0 0 1
which represents the order of the strands at Stage 1.
(ii) Find the 3 × 3 matrix B which represents the transition from Stage 1 to
Stage 2.
(iii) Find matrix M = BA and show that Ms0 gives s2, the matrix which
represents the order of the strands at Stage 2.
(iv) Find M2 and hence find the order of the strands at Stage 4.
(v) Calculate M3. What does this tell you?

Composition of transformations

Notation

A single letter, such as T, is often used to represent a transformation. The matrix

T represents the transformation T. (Notice that bold italic T is used for the
transformation and bold upright T for the matrix).
The point P has position vector p. The image of the point P can be denoted by
T(P) or by P′. The image T(P) has position vector p′ = T(p). We find T(p) by
evaluating the matrix product Tp.

Two or more successive transformations

Figure 1.16 shows the effect of two successive transformations on a triangle. The
transformation X represents a reflection in the x axis. X maps the point P to the
point X(P). The transformation Q represents a rotation of 90° anticlockwise
about O. When you apply Q to the image formed by X, the point X(P) is mapped
to the point Q(X(P)). This is abbreviated to QX(P).

X Q
y y y
QX(P)
P(x, y)

x x x
X(P)

Figure 1.16 23
 The composite transformation ‘reflection in the x axis followed by rotation of 90°
1 anticlockwise about O’ is represented by QX. Notice the order: QX means ‘carry
out X, then carry out Q’.

Matrices
●
? Look at figure 1.16 and compare the original triangle with the final image after
both transformations.
What single transformation is represented by QX?

The transformation X is represented by the matrix X =( 10 –10 ).

0 –1
The transformation Q is represented by the matrix Q = (
1 0)
.
0 –1 1 0 0 1
The product QX = ( )( ) =(
1 0)
.
1 0 0 –1
This is the matrix which represents reflection in the line y = x.
Notice the order in which the matrices were arranged for the multiplication: the
matrix representing the first transformation is on the right and the matrix for the
second transformation is on the left.
In general, the matrix of a composite transformation is found by multiplying the
matrices of the component transformations, in the correct order.
Activity 1.9 shows how this is proved.
a c
ACTIVITY 1.9 The transformations T and S are represented by the matrices T =
p r
( )
b d
and
S= ( )
q s
.
x
()
T is applied to the point P with position vector p = y . The image of P is P′ with
position vector p′. S is then applied to the point P′. The image of P′ is P′′ with
position vector p′′.

Figure 1.17

(i) Find p′ by finding the matrix product Tp.

(ii) Find p′′ by finding the matrix product Sp′.
24 (iii) Find the matrix product U = ST and show that U(p) is the same as p′′.
 ●
? How can you use the idea of successive transformations to justify the associativity 1
of matrix multiplication: (PQ)R = P(QR)?

Exercise 1D
● Proving results in trigonometry

Using successive transformations allows us to prove some useful results in

trigonometry.
If you carry out a rotation about the origin through angle α, followed by a
rotation about the origin through angle β, then this is equivalent to a single
rotation through the origin through angle α + β.

ACTIVITY 1.10 (i) Write down the matrix A representing a rotation about the origin through
angle α, and the matrix B representing a rotation about the origin through
angle β.
(ii) Find the matrix BA, representing a rotation about the origin through angle α,
followed by a rotation about the origin through angle β.
(iii) Write down the matrix C representing a rotation through the origin through
angle α + β.
(iv) By equating C to BA, write down expressions for sin(α + β) and cos (α + β).
(v) Explain why BA = AB in this case.
(vi) Write down the matrix D representing a rotation about the origin through
angle –β.
Use the matrix product AD to find expressions for sin (α – β) and cos (α – β).

EXERCISE 1D Use the following matrix transformations in questions 1 to 4.

The matrix X represents reflection in the x axis.
The matrix Y represents reflection in the y axis.
The matrix Q represents rotation of 90° anticlockwise about the origin.
The matrix R represents rotation of 90° clockwise about the origin.
The matrix S represents rotation of 180° about the origin.
The matrix T represents reflection in the line y = x.
The matrix U represents reflection in the line y = –x.

1 (i) Write down the matrices Y and Q.

(ii) Find the matrix QY and describe the transformation QY as simply as you can.
(iii) Draw diagrams to show the effect of applying Y followed by Q, and check
that the result agrees with your answer to part (ii).
(iv) Find the matrix YQ and describe the transformation YQ as simply as you can.
(v) Draw diagrams to show the effect of applying Q followed by Y, and check
that the result agrees with your answer to part (iv).

25
 2 (i) Write down the matrices R and T.
1 (ii) Find the matrix RT and describe the transformation RT as simply as you can.
(iii) Draw diagrams to show the effect of applying T followed by R, and check

Matrices
that the result agrees with your answer to part (ii).
(iv) Find the matrix TR and describe the transformation TR as simply as you can.
(v) Draw diagrams to show the effect of applying R followed by T, and check
that the result agrees with your answer to part (iv).
3 (i) Write down the matrices X and Y.
(ii) Find the matrix XY and describe the transformation XY as simply as you can.
(iii) Find the matrix YX.
(iv) Explain why XY = YX in this case.
4 (i) Write down the matrices S and U.
(ii) Find the matrix SU and describe the transformation SU as simply as you can.
(iii) Find the matrix US.
(iv) Explain why SU = US in this case.
2 –1
5 The transformations R and S are represented by the matrices R =
3 0
(
1 3 )
and
S= (
–2 4
.)
(i) Find the matrix which represents the transformation RS.
(ii) Find the image of the point (2, –1) under the transformation RS.

6 R1 and R2 are rotations of the plane anticlockwise about the origin through
angles 25° and 40° respectively. The corresponding matrices are R1 and R2.
(i) By considering the effects of the rotations, explain why R1R2 = R2R1.
(ii) Write down R1 and R2 and calculate R1R2.
(iii) What single transformation is represented by R1R2?

7 There are two basic types of four-terminal electrical network, as shown in the
diagram.

(i) In Type A the output voltage V2 and current I2 are related to the input
voltage V1 and current I1 by the simultaneous equations

V2 = V1 –I1R1
I2 = I1.
V V
( ) ( )
You can write I 2 = A I 1 .
2 1

Write down matrix A.

26
 (ii) In Type B the corresponding simultaneous equations are
V2 = V1 1
V

Exercise 1D
I2 = I1 – ––1 .
R2

Write down the matrix B which represents the effect of a Type B network.
(iii) Find the matrix which represents the effect of Type A followed by Type B.
(iv) Is the effect of Type B followed by Type A the same as the effect of
Type A followed by Type B?
In questions 8 to 10 you will need to use the matrix which represents reflection in
1 1 – m2 2m
the line y = mx. This can be written as ––––––2
1+ m (
2m m2 – 1
. )
1
8 (i) Find the matrix P which represents reflection in the line y = –– x, and the
3
matrix Q which represents reflection in the line y = 3x.
(ii) Use matrix multiplication to find the single transformation equivalent to
1 x followed by reflection in the line y = 3x.
reflection in the line y = ––
3
Describe this transformation fully.
(iii) Use matrix multiplication to find the single transformation equivalent to
reflection in the line y = 3x followed by reflection in the line y = ––1 x.
3
Describe this transformation fully.
9 (i) Find the matrix R which
represents rotation through 30°
anticlockwise about the origin. 2 2
(Use exact values. You may find
the diagram helpful.)
1 1

(ii) Find the matrix M which represents reflection in the line y = 3x.
(iii) Calculate MR.
What single transformation does this matrix represent?
10 The matrix T represents a reflection in the line y = mx.
Show that T2 = I, and explain geometrically why this is the case.
11 The one-way stretch S, ×5 parallel to the line y = –12 x, can be accomplished by
A rotating the plane clockwise about O through the angle α, where
tan α = –12 , and then
B doing a one-way stretch, ×5 parallel to the x axis, and then
C rotating the plane anticlockwise about O through the angle α.

(i) Find the matrix for each of transformations A, B and C,

(ii) Hence find the matrix which represents S.
(iii) Transformation S maps figure F to figure F′.
Describe the transformation which maps F′ to F, and find its matrix. 27
 Inverse matrices
1
The transformation Q represents a rotation of 90° anticlockwise about the origin.

Matrices
0 –1
The corresponding matrix Q =
1 0 ( . )
To undo the effect of Q, you need to carry out a rotation of 90° clockwise about
the origin. This is known as the inverse of Q, and is denoted by Q–1. The matrix
0 1
Q–1 which represents this inverse transformation is (
–1 0
. )
●
? Find the matrix product QQ–1.
Is Q–1Q equal to QQ–1?
Explain your answers.

ACTIVITY 1.11 Find several pairs of transformations and their inverses, together with the
matrices that represent the transformations.
Find the product of each matrix with its inverse.
What do you notice?
1 0
If the product of two square matrices, M and N, is the identity matrix I =
–1
then N is the inverse of M. We write N = M .
0 1
, ( )
For simple transformation matrices such as the ones you looked at above, it is
easy to find the inverse matrix by considering the inverse transformation.
However, a method is needed to find the inverse of any 2 × 2 matrix.

ACTIVITY 1.12 The matrix M = (45 23 ). Let M = (wx zy ).

–1

4 2 w y 1 0
=( )( ) =(
0 1)
MM–1 .
5 3 x z
Show that 4w + 2x = 1
and 5w + 3x = 0
Solve these equations to find w and x.
Now form and solve two equations in y and z.
Hence find M–1.
Now generalise this method to find the inverse of the matrix N = (ab dc ).
w y
Let N–1 =
and z.
( )
x z
, and again find two equations in w and x, and two equations in y

Solve these to find the values of w, x, y and z in terms of a, b, c and d.

28
 1 d –c
Hence show that N–1 = ––
Δ –b a
, ( ) Notice that the elements
on the leading diagonal (from 1
where Δ = ad – bc, provided Δ ≠ 0. the top left to bottom right) are

Inverse matrices
interchanged, and the signs of
the other two elements
are changed.

The number ad – bc is known as the determinant of the 2 × 2 matrix. If the

a c
determinant is zero, the matrix
b d ( )
does not have an inverse and is described
as singular. If the determinant is not zero, then the inverse of the matrix exists,
and the matrix is described as non-singular.

EXAMPLE 1.6 Find the inverses of these matrices, if they exist.

(i) (73 84 ) (ii) (21 84 )

SOLUTION

(i) The determinant of (73 84 ) = 7 × 4 – 8 × 3 = 28 – 24 = 4.

7 8 4 –8 1 –2
The inverse of (
3 4 ) ( –3 7 )
which may be written as (–e
®- ®-u )
1
–
is 4 .

Interchange the elements on the

leading diagonal and change the sign of
the other two elements.

(ii) The determinant of (21 84 ) is 2 × 4 – 8 × 1 = 8 – 8 = 0.

So (21 84 ) has no inverse.
As matrix multiplication is not commutative, you may have been wondering if it
matters in which order you multiply the two matrices in Activity 1.12. The next
activity investigates this.

ACTIVITY 1.13 (i) Example 1.6 showed that the inverse of the matrix (73 84 ) is (–e1®- –2®-u ).
Show that(73 84 )(–e1®- –2®-u ) = I and (–e1®- –2®-u )(73 84 ) = I.
a c
The matrix M = (
b d)
(ii) .

Write down the inverse matrix M–1, and show that MM–1 = M–1M = I.

This is an important result: it means that the inverse of a matrix, if it exists, is

unique. This applies to all square matrices, not just 2 × 2 matrices. 29
 The inverse of a product
1
●
? How would you undo the effect of a rotation followed by a reflection?

Matrices
How would you write down the inverse of a matrix product MN in terms of
M–1 and N–1?

Suppose you want to find the inverse of the product AB, where A and B are
non-singular matrices. This means that you need to find a matrix X such that
X(AB) = I.
X(AB) = I ⇒ XABB–1 = IB–1 Post-multiply by B–1.
⇒ XA = B–1
⇒ XAA–1 = B–1A–1 Post-multiply by A–1.
⇒ X = B–1A–1
Thus (AB)–1 = B–1A–1, where A and B are non-singular matrices of the same
order. To undo two transformations, you must undo the second transformation
before undoing the first. You put your socks on before your shoes, but
presumably you take your shoes off before your socks!

ACTIVITY 1.14 As often happens, it is easier to prove a result when you know the answer!
Use the associative property of matrix multiplication to show that (AB)(B–1A–1)
simplifies to I and so provide an alternative proof that (AB)–1 = B–1A–1.

EXERCISE 1E 1 Where possible, find the inverses of the following matrices.

(i)(46 35 ) (ii) (–46 –3–2 ) (iii) (–64 –32 )

(52 63 )
(iv) (v) (32 –11 ) (vi) (35 47 )
(–23 –96 )
(vii) (viii) (-́-́qw 2®-e ) (ix) (eg fh )
1–k 2
2 The matrix ( )
is singular.
–1 4–k
Find the possible values of k.

3 A= (56 34 ) and B = (41 32 ).

Calculate the following.
(i) A–1 (ii) B–1 (iii) A–1B–1
(iv) B–1A–1 (v) (BA)–1 (vi) (AB)–1
Comment on your results.
30
 4 Triangle T has its vertices at (1, 0), (0, 1) and (–2, 0).
3 1
It is transformed to triangle T′ by means of the matrix M = ( )
1 1
. 1
(i) Find the co-ordinates of the vertices of T′, and show T and T′ on one

Exercise 1E
diagram.
(ii) Find the ratio of the area of T′ to the area of T, and the value of the
determinant of M.
(ii) Find M–1, and verify that this matrix maps the vertices of T′ to the vertices
of T.
a c
5 A 2 × 2 singular matrix M is given as ( )
b d
.
Find M 2 and give your answer as a multiple of M.
Hence find a formula which gives Mn in terms of M.

6 (i) Two square matrices M and N (of the same size) have inverses M–1 and N–1
respectively. Show that the inverse of MN is N–1M–1.
1 7 4 1 0 0
(ii)
( ) (
0 0 1 –1 –4 1
)
A = 0 1 2 , B = 3 1 0 and C = AB.

(a) Evaluate the matrix C.

1 a b
( )
(b) Work out the matrix product A 0 1 c .
0 0 1
1 0 0
(c)
( )
By equating the product in part (b) to 0 1 0 , find A–1.
0 0 1
(d) Using a similar method, or otherwise, find B–1.
(e) Using your results from parts (c) and (d), find C–1.
[MEI]

Using the determinant of a 2 × 2 matrix

On page 29 the determinant of the 2 × 2 matrix M = (ab dc ) was defined as the

a c
number ad – bc. It is denoted in several ways: det M, | M |, or
b d
.
| |
The following activity shows the geometrical significance of determinants.

ACTIVITY 1.15 (i) Figure 1.18 shows the unit square OIPJ and its image OI′P′J′ under a
3 1
transformation T, defined by the matrix T = ( )
1 2
.
Find the area of the image and evaluate det T.

31
1

Matrices P(1, 1)
J(0, 1)

I(1, 0)

Figure 1.18

(ii) Transformation R is defined by the matrix R = (31 51 ).

Apply R to the unit square.
Find the area of the image and evaluate det R.
(iii) What do you notice?

Ignoring for the moment the sign of the determinant, you will have found that
the determinant is the area scale factor of the transformation.
The sign of the determinant also has significance. If you move anticlockwise
around the original shape, the unit square, you come to the vertices O, I, P, J, in
that order. Moving anticlockwise around the image in figure 1.18 you come to the
vertices in the same order. However, when you apply R to the unit square, the
order of the vertices is reversed. It is this reversal of sense that is indicated by the
negative sign of det R.
a c
In the next activity you will prove that when a general matrix M = ( )
b d
is
applied to the unit square, then the area of its image is equal to det M.

ACTIVITY 1.16 Figure 1.19 shows the unit square OIPJ and its image OI′P′J′ under the
a c
transformation M defined by the matrix M = ( )
b d
.
You may assume that det M is positive.

y y

P′

J′
P(1, 1) I′
J(0, 1)

O I(1, 0) x O x

Figure 1.19

Find the co-ordinates of the image points I′, P′ and J′.

32 Hence show that the area of OI′P′J′ is equal to det M.
 In Activity 1.16 you should have proved that the determinant gives you the area
scale factor of the area of any transformation of a square using a 2 × 2 matrix. 1

Using the determinant of a 2 × 2 matrix

●
? Explain how this result can be extended to any plane shape.

So ad – bc is the area scale factor of the transformation. Strictly you should say
that det T = ad – bc is the signed scale factor, as it can be negative, signifying that
the sense (clockwise or anticlockwise) has been reversed.

ACTIVITY 1.17 Show that the matrices which represent reflections in the x axis, the y axis, the
line y = x and the line y = –x all have a determinant of –1.
In each case, draw diagrams to demonstrate that vertices labelled clockwise around
a shape are transformed to vertices labelled anticlockwise around the image.

Matrices with determinant zero

6 4
ACTIVITY 1.18 (i) The transformation T is defined by T =
Find det T.
3 2 ( )
.

(ii) Draw the rectangle with vertices (0, 0), (4, 0), (4, 3), (0, 3).
Draw the image of this rectangle under T.
(iii) Find some more 2 × 2 matrices with determinant zero.
Find the image of the rectangle in part (ii) under the transformations defined
by your matrices.
(iv) What do you notice?

If det T = 0 the matrix T is said to be singular, and the transformation

represented by T maps every point in the plane on to a single line which passes
through the origin, as shown in figure 1.20. This is called transforming the plane.

Figure 1.20

33
 Each point on the image line P′Q′ is the image of infinitely many points. All the
1 points which map to P′ fall on a straight line, labelled AB in figure 1.20, and every
point on AB maps to P′. Similarly, I′ is the image of all points on a line through I

Matrices
parallel to AB and each point on a line through Q parallel to AB maps to Q′.

●
? What happens when you map any shape using the zero matrix ( 00 00 )?
How is the plane transformed?

EXAMPLE 1.7 The plane is transformed by means of the matrix M = ( 42 –6–3 ).

(i) Show that det M = 0.
(ii) Show that the whole plane is mapped to a straight line, and find the equation
of this line.
(iii) Find the equation of the line of points that map to (6, 3).

SOLUTION

(i) det M = (4 × –3) – (–6 × 2) = –12 + 12 = 0

4 –6 x x′
(ii) ( )( ) ( )
2 –3 y
= y
′

{4x – 6y = x′
2x – 3y = y′
x′ = 2(2x – 3y) = 2y′
The plane is mapped to the line x = 2y.

(iii) ( 42 –6–3 )(xy ) = ( 63) These two statements give

{ 4x2x –– 6y3y == 63 the same information.

Points on the line 2x – 3y = 3 map to the point (6, 3).

ACTIVITY 1.19 The plane is transformed using the matrix (ab dc ), where ad – bc = 0.
Prove that the general point P(x, y) maps to P′ on the line bx – ay = 0.

34
EXERCISE 1F 1 Find the value of the determinant of each of the following matrices and decide
whether each matrix is singular or non-singular. 1
6 4 4 8 5 3 1 –2
( )
(i) ( (ii) ) (iii) ( ) (iv) ( )

Exercise 1F
2 3 –1 –2 1 †-e 2 3
5 3 3 2
M=(
4 2)
and N = (
–2 1 )
2 .
(i) Find the determinants of M and N.
(ii) Find MN and det (MN). What do you notice?
(iii) By considering the geometrical significance of the determinant, explain
why det (MN) = det M × det N.

3 The two-way stretch with matrix (a0 0d ) preserves area (i.e. the area of the
image is equal to the area of the original shape).
What is the relationship connecting a and d?
4 A shear moves each point parallel to the x axis by a distance k times its distance
from the x axis.
Find the matrix for this transformation and hence show that whatever the
value of k, the shear preserves area.
7 –3
5 The matrix M = (
–4 6 )
defines a transformation in the (x, y) plane.
A triangle S, with area 5 square units, is transformed by M into triangle T.
(i) Find the area of triangle T.
(ii) Find the matrix which transforms triangle T into triangle S.

Triangle U is obtained by rotating triangle S through 135° anticlockwise about

the origin.
(iii) Find the matrix which transforms triangle S into triangle U.
(iv) Find the matrix which transforms triangle T into triangle U.
[MEI, part]

6 The plane is transformed by means of the matrix M = (21 42 ).

(i) Show that det M = 0, and that the whole plane is mapped on to the line
x – 2y = 0.
(ii) The point P(x, y) is mapped to P′(4, 2).
Use the equation p′ = Mp to show that P could be anywhere on the line
x + 2y = 2.
(iii) Find the equation of the line of points that map to (10, 5).
7 A matrix T maps all points on the line x + 2y = 1 to the point (1, 3).
(i) Find the matrix T and show that its determinant is zero.
(ii) Show that T maps all points on the plane on to the line y = 3x.
(iii) Find the co-ordinates of the point that all points on the line x + 2y = 3 are
mapped to.

35
 8 The point P is mapped to P′ on the line 3y = x, so that PP′ is parallel to y = 3x.
1 (i) Find the equation of the line parallel to y = 3x passing through the point P
with co-ordinates (s, t).

Matrices
(ii) Find the co-ordinates of P′, the point where this line meets the line 3y = x.
(iii) Find the matrix of the transformation which maps P to P′, and show that
the determinant of this matrix is zero.
9 A shear moves each point parallel to the line y = mx.
Each point is moved k times its distance from the line y = mx.
(Points to the right of the line are moved upwards, points to the left of the line
are moved downwards).
(i) Find the images of the points I(1, 0) and J(0, 1).
(ii) Hence write down the matrix which represents the shear.
(iii) Show that whatever the value of m, the shear preserves area.

Matrices and simultaneous equations

You are already able to solve a pair of simultaneous linear equations such as

{3x4x ++ 2y5y == 95
by the method of elimination.
An alternative method, which can be used to solve any number of simultaneous
linear equations, involves the use of matrices.
The equations above can be written as a single matrix equation

( 34 25)(xy ) = ( 95).
The inverse of the matrix ( 34 25 ) is ( –45 –23 ).
1
–
7

Pre-multiply both sides

1
–
7 ( –45 –23 )( 34 25 )(xy ) = ( –45 –23 )( 95)
1
–
7 of the matrix equation by

(
1 5 –2
–
).
(xy ) = ( –2135 ) = ( –35 )
1
–
7
7 –4 3

The solution is x = 5, y = –3. As M–1Mp = p,

the left-hand side
The equation x
()
simplifies to y .

3 2 x 9
( )( ) ( )
4 5 y
=
5
has the form Mp = p′, where M and p′ are known and you are trying to find p.
This is equivalent to finding the co-ordinates of the unknown point P which is
transformed on to the known point P′ by the given matrix M.
36
An alternative geometrical interpretation of solving a pair of simultaneous equations,
which may already be familiar to you, is to think of the two equations as the equations 1
of two lines in a plane. You are looking for any points which lie on both lines.

Matrices and simultaneous equations

Whichever geometrical interpretation you use, there are three possible alternatives.

Case 1

The lines cross at a single point, as shown in figure 1.21. This is the case when
det M ≠ 0, so the inverse matrix M–1 exists and there is a unique position for P, as
shown in figure 1.22.
Both interpretations show that there is a unique solution to the equations.

Figure 1.21 Figure 1.22

Case 2

The lines are distinct parallel lines, as shown in figure 1.23. In this case det M = 0,
so the inverse matrix M–1 does not exist, and the transformation maps all points
on to a single line l through the origin. In this case P′ is not on l, so P′ is not the
image of any point (see figure 1.24).
Both interpretations show that the equations have no solution. We say that the
equations are inconsistent.

M
l

P(x, y)

Figure 1.23 Figure 1.24

Case 3

The lines are coincident lines, i.e. they are the same line, as shown in figure 1.25.
In this case det M = 0 again, so the inverse matrix M–1 does not exist and the
transformation maps all points on to a single line l through the origin. In this
case, P′ is on l, so there are infinitely many positions for P, all the points on a
particular line (see figure 1.26). 37
 Both interpretations show that the equations have infinitely many solutions,
1 which may be expressed in terms of a single parameter (see Example 1.8 part (iii)).

Matrices l

Figure 1.25 Figure 1.26

EXAMPLE 1.8 Find, if possible, the solution of the equations {2x4x –+ 3ymy == 4k
(i) when m = 1 and k = 1
(ii) when m = –6 and k = 3
(iii) when m = –6 and k = 8.

SOLUTION

The simultaneous equations may be written as the following matrix equation.

( 24 –3m )(xy ) = ( 4k )
(i) When m = 1,
| |
2 –3
4 1
1 1 3
––
=14, so the inverse matrix exists and is equal to 14 (
–4 2
. )
( xy ) = ( –41 32 )( 41 ) = ( –147 ) = ( –1∑-q )
1
––
14
1
––
14

There is a unique solution: x = –12 , y = –1.

(ii) When m = –6,

| |
2 –3
4 –6
= 0, so the inverse matrix does not exist.

When k = 3, the equations are {2x4x –– 3y6y == 43 .

These equations are inconsistent as 2x – 3y cannot be both 4 and –32 .
There are no solutions.
(iii) As for part (ii), the inverse matrix does not exist.

When k = 8, the equations are {2x4x –– 3y6y == 48 .

Both equations reduce to 2x – 3y = 4.
There are infinitely many solutions, which can be given in terms of a
2λ 4
38 parameter λ as x = λ, y = –– – – .
3 3
 The strength of the matrix method for solving simultaneous equations is that it
can be used for any number of equations. Of course, to solve a set of n 1
simultaneous equations in n unknowns, you do need to be able to find the
inverse of an n × n matrix. At this stage you will only be dealing with the case

Exercise 1G
where n = 2 unless you are given help with finding the inverse of a larger matrix,
such as 3 × 3.

ACTIVITY 1.20 Three simultaneous equations

(i) Use the matrix facility on your calculator to find the inverse of the
3 2 1

2 1 1
( )
matrix 1 4 –2 .

3x + 2y + z = 5

{
Hence find the solution of the equations x + 4y – 2z = 3 .
2x + y + z = 3
x + 3y – 2z = 7
(ii)

matrix facility on your calculator?

{
What happens if you try to solve the equations 2x – 2y + z = 3 using the
3x + y – z = 10

Why do you think this is?

Try to solve the equations algebraically. What happens?
x + 3y – 2z = 7
(iii)
{
Repeat part (ii) for the equations 2x – 2y + z = 3 .
3x + y – z = 12

●
? ● Geometrical interpretation of three simultaneous equations
What is the geometrical interpretation of each of the three situations in
Activity 1.20?

EXERCISE 1G 1 (i) ( –12 31).

Find the inverse of the matrix

2 3 x 1
(ii) Hence solve the equation ( )( ) = ( ).
–1 1 y –3
2 Use matrices to solve the following pairs of simultaneous equations.

(i)
{3x2x –+ y3y==2 5 (ii)
{3xx –+2y2y==4 4
(iii)
{x2x+–3yy == 111 (iv)
{3xx ––4y2y==–29
39
 3 For each of the following pairs of equations, decide whether the equations are
1 consistent or inconsistent.
If they are consistent, solve them, in terms of a parameter if necessary.

Matrices
In each case, describe the configuration of the corresponding pair of lines.

(i)
{3x2x ++ 5y4y == 1711 {3x2x ++ 6y4y == 1215
(ii)

(iii)
{6x2x –– 3yy ==412 {8xy =–2x4y–=411
(iv)

Find the two values of k for which the equations {

2x + ky = 3
4 do not have a
kx + 8y = 6
unique solution. Where possible, find the solution set for the equations.

5 –2 k –1 3k + 8 4k + 10
5 (i) Find AB, where A =
( –2 3 4
) 1
(
3 –4 –5 and B = –2 2k + 20 3k + 25 .
–11 –14
)
Hence write down the inverse matrix A–1, stating a necessary condition on
k for this inverse to exist.
(ii) Using the result from part (i), or otherwise, solve the equation
5 –2 k x 28
( 3 –4 –5
–2 3 4
)( ) ( )
y = 0
z m
in each of these cases.
(a) k = 8, giving x, y and z in terms of m
(b) k = 1 and m = 4
(c) k = 1 and m = 2
[MEI, part, adapted]

6 Matrices A and B are given by

1 –2 0 6 6 4
( ) ( )
A = 0 2 –2 , B = k 3 2 (where a ≠ – –32 and k ≠ 3).
a 0 3 k k 2
(i) Find the matrix AB.
(ii) Show that there is a relationship between a and k for which AB is a scalar
multiple of the identity matrix I.
6 6 4
(iii) Deduce that A =–1 1
–––––
6 + 4a ( )
–2a 3 2 and find B–1 in terms of k.
–2a –2a 2
1 –2 0 x 2
(iv)
(
a 0 3 z 3
)( ) ( )
Given that 0 2 –2 y = –2 , express each of x, y and z in terms of a.

(v) Find (AB)–1 in terms of a and k.

[MEI]
40
 7 You are given the matrices
–2 26 –16 –2 1 k 1
(
P = 1 –11
)
7 and Q = 3 5 –1 .
( )

Exercise 1G
–1 21 –13 5 8 –2
(i) Calculate the matrix product PQ.
(ii) For the case k = 3, write down the inverse matrix Q–1 and hence solve the
following equation.
x 19

z
() ( )
Q y = 4
5
(iii) For the case k = 5, you are given that Q has no inverse. Solve the equation
x 19

z
() ( )
Q y = 4 ,
5
giving x, y and z in terms of a parameter t.
[MEI, part]

Invariant points

●
? In a reflection, which points map to themselves?
In a rotation, are there any points which map to themselves?

Points which map to themselves under a transformation are called invariant

points.

The product ( 62 53 )( –22 ) = (–22 ). This means that the transformation (62 53 )
maps the point (2, –2) to itself. This is an example of an invariant point.

●
? Explain why the origin is always an invariant point in any transformation that
can be represented by a matrix.

41
1 EXAMPLE 1.9 Find the invariant points under the transformation given by the matrix (21 –10 ).
SOLUTION

Matrices
x
()
Suppose y maps to itself. Then

(21 –10 )(xy ) = (xy ) ⇔ (2x – y) (x )

x
= y
⇔ 2x – y = x and x = y
⇔ x = y.
Thus all points on the line y = x (and only points on this line) map on to
themselves. This is a line of invariant points. The invariant points can be
expressed in terms of a parameter as (λ, λ).

2 –1 x x
Notice that in Example 1.9, the matrix equation
equations which turned out to be equivalent.
(
1 0 y )( ) ( )
= y led to two

Clearly any matrix equation of this form will lead to two equations of the form
ax + by = 0, which represent straight lines through the origin. Either these two
equations are equivalent, representing the same straight line, which means that
all the invariant points lie on this line, or they are not, in which case the origin is
the only point which satisfies both equations, and so is the only invariant point.

● Invariant lines
A line AB is known as an invariant line under a transformation T if the image of
each point on AB is also on AB. It is important to note that it is not necessary for
points on AB to map on to themselves (as in Example 1.9) but merely that each
point on AB maps to a point on AB.
Sometimes it is easy to spot which lines are invariant, for example, in a reflection,
as well as the mirror line being invariant (because it is a line of invariant points),
each line perpendicular to the mirror line is invariant, as shown in figure 1.27.

Figure 1.27
42
 ●
? What lines, if any, are invariant in the following transformations? 1
(i) Enlargement, centre the origin

Exercise 1H
(ii) Rotation through 180° about the origin
(iii) Rotation through 90° about the origin
(iv) Reflection in the line y = x

EXERCISE 1H 1 Find the invariant points for the transformations with the following matrices.
0 –1 3 4 0.6 0.8
(i) (
1 2 ) (ii)
1 2 ( ) (iii)
0.8 –0.6 ( )
1 – ––
1
(iv)
( ) ––
1
2
– –– ––
2
1
2

2
x
(v)

a c x
(46 13 ) (vi) (73 –4–1 )
2 The transformation T maps y to ( ) ( )( )
b d y
.

Show that invariant points other than the origin exist if det T = a + d – 1.
– m2 –––––
1––––– 2m

3 The matrix
( 1 + m 1 + m2
2m
–––––2
2

m 2–1
–––––
1 + m 1 + m2
) represents reflection in the line y = mx.

Prove that the line y = mx is a line of invariant points.

4 (i) M is a reflection of the plane such that the point (x′, y′) is the image of
the point (x, y), where

(x′y ′ ) = ( –0.6 0.8 x

0.8 0.6 )( y )
.

Find a point, other than the origin, that is invariant under this reflection.
Hence find the equation of the mirror line.
a
(ii) T is a translation of the plane by the vector
b
. ()
The point (X, Y) is the image of the point (x, y) under the combined
transformation TM (that is M followed by T) where
–0.6 0.8 a x
() (X
Y = 0.8 0.6 b y .
1 0 0 1 1 )( )
(a) Show that if a = –4 and b = 2 then (0, 5) is an invariant point of TM.
(b) Show that if a = 2 and b = 1 then TM has no invariant point.
(c) Find a relation between a and b that must be satisfied if TM is to have
any invariant points.
[SMP]
43
 INVESTIGATIONS
1 1 Investigate the sequence of matrices I, M, M 2, M 3, ... where M = (11 10 ).
Matrices
2 Mendel’s genetic theory states that offspring inherit characteristics from their
parents through their genes. Each characteristic is controlled by a pair of
genes, which may be of two types: G or g. The three possible combinations are
GG known as homozygous dominant, and denoted by D
Gg or gG known as heterozygous, and denoted by H
gg known as homozygous recessive, and denoted by R.
Offspring inherit one gene from each parent, randomly and independently.
In a controlled study of a population of fruit flies, the only females allowed to
mate are those known to be heterozygous.

(i) The unfinished matrix shows the probability that the offspring has
combination D, H or R given the combination of genes of the male parent.

Male parent
D H R
Offspring D ∑-q ®-q 0

( )
H ... ... ...
R ... ... ...

Copy and complete the matrix M containing this information.

(ii) Initially the proportions of the D, H and R combinations in the

d
()
population are given by x 0 = h .
r
Show that x 1 = Mx 0 gives the proportions in the next generation.

(iii) Investigate what happens if the study continues over many generations.

44
KEY POINTS
1 The matrix M = (ab dc ) represents the transformation which maps the point with 1
x x′
position vector ( y ) to the point with position vector ( y ), where {
x′= ax + cy

Key points
.
′ y′ = bx + dy
1 0
2 The image of ( ) is the first column of M, the image of ( ) is the second
0 1
column of M.
3 Matrix mulitplication
Row from left matrix with
column from right matrix.

( pq rs )( ba dc ) = (paqa ++ sbrb qcpc ++ sdrd )

4 Matrix multiplication is not commutative: MN ≠ NM for all M, N.

5 Matrix multiplication is associative: N(ML) = (NM)L.

6 The matrix I = (10 01 ) is known as the identity matrix.

0 0
The matrix (
0 0)
is known as the zero matrix.

7 The composite transformation ‘M followed by N’ is represented by the

matrix product NM.
a c
8 The determinant of the matrix M = ( )
b d
is ad – bc.
a c
It is denoted by | M | , det M or
| |
b d
.

The determinant of the matrix M gives the area scale factor of the associated
transformation M.
a c 1 d –c
9 The inverse of M = ( )
b d
is M–1 = ––––––
ad – bc –b a ( )
, provided ad – bc ≠ 0.
MM–1 = M–1M = I.
10 When solving n simultaneous equations in n unknowns, the equations can
be written as a matrix equation.
x1 a1
M
()()
x2
...
xn
a
= 2
...
an
If det M ≠ 0, there is a unique solution which can be found by
pre-multiplying by the inverse matrix M–1.
If det M = 0, then there is either no solution or infinitely many solutions.
11 The point P is known as an invariant point under T if the image of P under
T is P. 45
2

Complex numbers
2 Complex numbers

…that wonder of analysis, that portent of the ideal world, that

amphibian between being and not-being, which we call the imaginary
root of negative unity.
Leibniz, 1702

The growth of the number system

The number system we use today has taken thousands of years to develop. In
primitive societies all that are needed are the counting numbers, 1, 2, 3,… (or even
just the first few of these).
The concept of a fraction was first recorded in a systematic way in an Egyptian
papyrus of about 1650 BC. By 500 BC the Greeks had developed ways of
calculating with whole numbers and their ratios (which accounts for calling
fractions rational numbers). The followers of Pythagoras believed that everything
in geometry and in applications of mathematics could be explained in terms of
rational numbers.
It came as a great shock, therefore, when one of them proved that 2 was not a
rational number. However, Greek thinkers gradually came to terms with the
existence of such irrational numbers, and by 370 BC Eudoxus had devised a very
careful theory of proportion which included both rational and irrational numbers.
It took about another thousand years for the next major development, when the
Hindu mathematician Brahmagupta (in about AD 630) described negative
numbers and gave the rules for dealing with negative signs. Surprisingly, the first
use of a symbol for zero came even later, in AD 876. This was the final element
needed to complete the set of real numbers, consisting of positive and negative
rational and irrational numbers and zero.
Figure 2.1 shows the relationships between the different types of numbers.

46