Introduction to

Matrices
Introduction
When we wish to solve large systems of simultaneous linear equations, which arise for example
in the problem of finding the forces on members of a large framed structure, we can isolate
the coefficients of the variables as a block of numbers called a matrix. There are many other
applications of a matrix. In this Block we develop the terminology and basic properties of a
matrix.

Definitions
An array of numbers rectangular in shape, is called a matrix. The first matrix below has 3
rows and 2 columns and is said to be a ‘3 by 2’ matrix (written 3 × 2)
 
1 4
 −2 3  1 2 3 4
5 6 7 9
2 1

The second matrix is a ‘2 by 4’ (written 2 × 4).

The general 3 × 3 matrix can be written
 
a11 a12 a13
A =  a21 a22 a23 
a31 a32 a33

where aij denotes the element in row i, column j.

For example in the matrix:  
0 −1 −3
A= 0 6 −12 
5 7 123

then
a11 = 0, a12 = −1, ... a22 = 6, ... a32 = 7, a33 = 123
 Key Point
The General Matrix
A general m × n matrix A has m rows and n columns.
The entries in the matrix are called the elements of A
In matrix A the element in row i and column j is denoted by aij

A matrix
  with only
 one column is called a column vector (or column matrix), for example,
x1 3
 x2  and  4  are both 3 × 1 column vectors.
x3 5
A matrix with only one row is called a row vector (or row matrix). For example [2, −3, 8, 9] is
a 1 × 4 row vector (often the entries in a row vector are separated by commas for clarity).

Square matrices
When the number of rows is the same as the number of columns, i.e. m = n, the matrix is said
to be square and of order n (or m).

• In an n × n square matrix A, the leading diagonal (or principal diagonal) is the north-
west to south-east collection of elements a11 , a22 , . . . , ann . The sum of the elements in the
leading diagonal of A is called trace of the matrix, denoted by tr(A).

 
a11 a12 ... a1n
a a22 ... a2n 
A =  21 ..  tr(A) = a11 + a22 + . . . + ann
 ... ..
.
..
. . 
an1 an2 ... ann

• A square matrix in which all the elements below the leading diagonal are zero is called an
upper triangular matrix, often denoted by U .

 
u11 u12 ... u1n
 0 u22 ... u2n 
U =
 0 .. ..  uij = 0 when i > j
0 . . 
0 0 0 unn
 • A square matrix in which all the elements above the leading diagonal are zero is called a
lower triangular matrix, often denoted by L.

 
l 11 0 0 0
 l 21 l 22 0 0 
L=
 ... .. ..  lij = 0 when i < j
. . 0 
l n1 l n2 ... l nn

• A square matrix where the only non-zero elements are along the leading diagonal is called
a diagonal matrix, often denoted by D.

 
d11 0 ... 0
 0 d22 ... 0 
D=
 ... .. .. ..  dij = 0 when i = j
. . . 
0 0 . . . dnn

1 2 3
Example A = is 2 × 3. It is not square.
4 5 6
1 2
B= is 2 × 2. It is square. Also tr(B) = 1 + 4 = 5.
3 4
   
1 2 3 4 0 3
 
Matrices C = 0 −2 −5 and E = 0 −2  5  are 3 × 3, square and
0 0 1 0 0 1
upper triangular. Also tr(C) = 0 and tr(E) = 3.
   
1 0 0 −1 0 0

Matrices F = 2 −2 0 and G =   1 4 0  are 3 × 3, square and
3 −5 1 0 1 1
lower triangular.
   
1 0 0 4 0 0
Matrices D =  0 2 0  and H =  0 2 0  are 3 × 3, square and
0 0 −3 0 0 0
diagonal.
Now do this exercise
Classify the following matrices (and, where possible, find the trace):
 
    1 2 3 4
1 2 1 2 3 4  5 6 7 8  
A= 3 4  B= 5 6 7 8  C=  9 10 11 12 
5 6 −1 −3 −2 −4
13 14 15 16
Answer

Now do this exercise

Classify the following matrices:

       
1 1 1 1 0 0 1 1 1 1 0 0
A= 1 1 1  B= 1 1 0  C= 0 1 1  D= 0 1 0 
1 1 1 1 1 1 0 0 1 0 0 1
Answer

Equality of matrices
As we noted earlier, the terms in a matrix are called the elements of the matrix.
1 2
The elements of the matrix A = are 1, 2, −1, −4
−1 −4
We say two matrices A, B are equal to each other only if A and B have the same number of
rows and the same number of columns and if each element of A is equal to the corresponding
element of B. When this is the case we write A = B. For example if the following two matrices
are equal:
1 α 1 2
A= B=
−1 −β −1 −4
then we can conclude that α = 2 and β = 4.

The unit matrix

The unit matrix or the identity matrix, denoted by In (or, often, simply I), is the diagonal
matrix of order n in which all diagonal elements
 are 1. 
1 0 0
1 0
Hence, for example, I2 = 
and I3 = 0 1 0 .
0 1
0 0 1

The zero matrix

The zero matrix or null matrix is the matrix all of whose elements are zero. There is a zero
matrix for every size. For example the 2 × 3 and 2 × 2 cases are:
0 0 0 0 0
, .
0 0 0 0 0
Zero matrices, of whatever size, are denoted by 0.
The transpose of a matrix
The transpose of a matrix A is a matrix where the rows of A become the columns of the new
matrix and the columns of A become its rows. For example
 
1 4
1 2 3
A= becomes  2 5 
4 5 6
3 6

The resulting matrix is called the transposed matrix of A and denoted AT . In the previous
example it is clear that AT is not equal to A since the matrices are of different sizes. If A is
square n × n then AT will also be n × n.

   
1 2 3 1 4 7
Example If B is  4 5 6  then B T is  2 5 8 
7 8 9 3 6 9
Both matrices are 3 × 3 but B and B T are clearly different.

When the transpose of a matrix is equal to the original matrix i.e. AT = A, then we say that
the matrix A is symmetric.
In the previous example B is not symmetric.

   
1 −2 3 1 −2 3
Example If C =  −2 4 −5  then C T =  −2 4 −5 .
3 −5 6 3 −5 6
Clearly C T = C and C is a symmetric matrix. Notice how the leading diagonal
acts as a “mirror”; for example c12 = −2 and c21 = −2. In general cij = cji
for a symmetric matrix.

Now do this exercise

Find the transpose of each of the following matrices. Which are symmetric?
 
1 2
1 2 1 1 1 1 1 0
A= , B= C= D= 4 5  E=
3 4 −1 1 1 0 0 1
7 8
Answer

Addition and Subtraction of matrices

Under what circumstances can we add two matrices i.e. define A + B for given matrices A, B?
Consider
1 2 5 6 9
A= and B=
3 4 7 8 10
There is no sensible way to define A + B in this case since A and B are different sizes.
However, if we consider matrices of the same size then addition can be defined in a very natural
1 2 5 6
way. Consider A = and B = . The ‘natural’ way to add A and B is to add
3 4 7 8
corresponding elements together:
1+5 2+6 6 8
A+B = =
3+7 4+8 10 12
In general if A and B are both m × n matrices, with elements aij and bij respectively, then their
sum is a matrix C, also m × n, (written C = A + B) such that the elements of C are
cij = aij + bij i = 1, 2, . . . , m j = 1, 2, . . . , n
In the above example
c11 = a11 + b11 = 1 + 5 = 6 c21 = a21 + b21 = 3 + 7 = 10 and so on
Subtraction of matrices follows along similar lines:
1−5 2−6 −4 −4
D =A−B = =
3−7 4−8 −4 −4

Multiplication of a matrix by a number

There is also a natural way of defining the product of a matrix with a number. Using the matrix
A above, we note that
1 2 1 2 2 4
A+A= + =
3 4 3 4 6 8
What we see is that 2A (which is the shorthand notation for A + A) is obtained by multiplying
every element of A by 2.
In general if A is an m × n matrix with typical element aij then the product of a number k with
A is written kA and has the corresponding elements kaij .
Hence, again using the matrix A above,
1 2 7 14
7A = 7 =
3 4 21 28
Similarly:
−3 −6
−3A =
−9 −12
Now do this exercise
For the following matrices find, where possible, A + B, A − B, B − A, 2A.
1 2 1 1
1. A = B=
3 4 1 1
   
1 2 3 1 1 1
2. A =  4 5 6  B =  −1 −1 −1 
7 8 9 1 1 1
   
1 2 3 1 2
3. A =  4 5 6  B= 3 4 
7 8 9 5 6
Answer
Some simple matrix properties
Using the definition of matrix addition described above we can easily verify the following prop-
erties of matrix addition:

Key Point

Properties of Matrices
Matrix addition is commutative: A + B = B + A.
Matrix addition is associative: A + (B + C) = (A + B) + C.
The distributive law holds: k(A + B) = k A + k B

These keypoint results follow by the fact that aij + bij = bij + aij etc.

We can also show that the transpose of a matrix satisfies the following simple properties:

(A + B)T = AT + B T
(A − B)T = AT − B T
(AT )T = A.

1 2 3
Example Show that (AT )T = A for the matrix A = .
4 5 6

Solution
 
1 4
1 2 3
AT =  2 5  so that (AT )T = =A
4 5 6
3 6

Now do this exercise

1 2 1 −1
For the matrices A = , B= verify that
3 4 −1 1
(i) 3(A + B) = 3A + 3B (ii) (A − B)T = AT − B T .
Answer
More exercises for you to try

   
2 3 −1 1 2 3
1. If A =  4 4 0  and B = 4 5 6  determine (3AT − B)T .
2 −1 −1 0 0 1

 
−1 4
1 2 3
2. If A = and B =  0 1  verify that 3(AT − B) = (3A − 3B T )T .
4 5 6
2 7

Answer
A is 3 × 2, B is 3 × 4, C is 4 × 4 and square. The trace is not defined for A or B. However,
tr(C) = 34.

Back to the theory

A is 3 × 3 and square, B is lower triangular, C is upper triangular and D is diagonal

Back to the theory

 1 3 1 −1 1 1
AT = , BT = CT = = C, symmetric
2 4 1 1 1 0
1 4 7 1 0
DT = ET = = E, symmetric
2 5 8 0 1

Back to the theory

 2 3 0 1 0 −1 2 4
1. A + B = A−B = B−A= 2A =
4 5 2 3 −2 −3 6 8
     
2 3 4 0 1 2 0 −1 −2
2. A + B =  3 4 5  A−B = 5 6 7  B − A =  −5 −6 −7 
8 9 10 6 7 8 −6 −7 −8
 
2 4 6
2A =  8 10 12 
14 16 18
 
2 4 6
3. None of A + B, A − B, B − A, are defined. 2A =  8 10 12 
14 16 18

Back to the theory

 2 1 6 3 3 6
(i) A + B = A = ; 3(A + B) = 3A =
2 5 6 15 9 12
3 −3 6 3
3B = ; 3A + 3B =
−3 3 6 15
0 3 0 4 1 3
(ii) A − B = ; (A − B)T = , AT = ,
4 3 3 3 2 4
1 −1 0 4
BT = ; AT − B T =
−1 1 3 3

Back to the theory

    
2 4 2 6 12 6
 3 4 −1  , 3AT =  9 12 −3 
1. AT =
−1 0 −1 −3 0 −3
   
5 10 3 5 5 −3
3AT − B =  5 7 −9  (3AT − B)T =  10 7 0 
−3 0 −4 3 −9 −4
     
1 4 2 0 6 0
2. AT =  2 5  , AT − B =  2 4  , 3(AT − B) =  6 12 
3 6 1 −1 3 −3
−1 0 2 3 6 9 −3 0 6 6 6 3
BT = , 3A − 3B T = − = .
4 1 7 12 15 18 12 3 21 0 12 −3

Back to the theory