CHAPTER 2

LINEAR ALGEBRA
2.1 INTRODUCTION
The main objective of this chapter is the development of various techniques that can be
employed to solve a system of simultaneous equations such as :

2x 3y 2z = 9
8x 5y z = 5

You might recognize the notion of simultaneous equations from your school years or possibly from
a previous course in Algebra.

Before progressing to the techniques we first discuss the concept of a matrix and the
determinant of a matrix.

2.2 MATRICES
A matrix can be thought of as an ordered rectangular arrangement of numbers into horizontal
rows and vertical columns. To indicate that we are working with a matrix the numbers are
enclosed by square brackets - […….]

Examples of matrices:

; ; ; ; ;[ ]

Matrices are classified according to the size, i.e. the number of rows and columns. The first matix
above has size “2 by 3”, writen as 2 3 , as it has 2 rows and 3 columns. The number of rows
(horizontal) is always stated first and the number of columns (vertical) second. The sizes of the
other matrices above are: 3 × 2; 2 × 2; 3 × 3; 3 × 1 and 1 × 5 respectively.

Consider the system given above.

The coefficients of x , y and z can be described by the matrix:
2 3 2
A=
8 5 1
in which the coefficients of x appear in the first column, the coefficients of y in the second
column and the coefficients of z in the third column. This matrix is called the coefficient matrix
of the system. This matrix will be used later on in the solving of systems of linear equations.
Because A consists of two rows and three columns, we say A has size “2 by 3”, and write 2 3
for the size of A .

At this point we formally define what is meant by the term matrix.

DEFINITION

Let ,
An ordered rectangular arrangement of real numbers into m rows and n columns

is called an m n (read as “m by n”) matrix over .

The aij appearing in the matrix are referred to as the entries of the matrix. Specifically, aij refers

to the entry appearing in the i th row and j th column of A.

Remarks

We use “matrix over ” to express the fact that the entries of the matrix are real numbers.
Matrices are usually denoted with capital letters and the entries by small letters (lower case
letters).

Referring back to the coefficient matrix A,

2 3 2
A = aij
2 3 8 5 1

the entries a11 = 2 , a12 = 3 , a13 = 2 , etc.

DEFINITION

Two matrices A = ai j and B = bi j are equal if they have the same size and entries in
corresponding positions are equal.
 Example
2 3 2 3
Matrix A = and Matrix B = are different because although they have the same
5 2 2 5
size, not all the entries in corresponding positions are equal.

Specifically,
a21 = 5 but b21 = 2
a22 = 2 but b22 = 5
Therefore, A B .

2.3 TYPES OF MATRICES

Various important types of matrices exist and we consider a few types.

1. If a matrix has the same number of rows as columns, we call it a square matrix. Thus
an m m matrix is a square matrix for all values m .

2. If A = a i j is an m n matrix for which all the entries ai j = 0 , then A is called the m n

m n

zero matrix.

3. A matrix with precisely one row is called a row matrix.

Thus A = a i j = a11 a12 a1n is a row matrix.
1 n

4. A matrix with precisely one column is called a column matrix.

b 11
b 21
Thus B = b i j = is a column matrix.
n 1

bn1

5. A diagonal matrix is a square matrix A = ai j where ai j = 0 if i j.

m m

a11 0 0
0 a22 0
A diagonal matrix is thus of the form

0 0 amm
The diagonal from the upper left to bottom right, is called the main diagonal of the
matrix.
 Note. If a matrix is not square, then we cannot speak of a diagonal.

6. An identity matrix I n is an n n diagonal matrix in which all the entries along the
main diagonal equals 1. Thus aii = 1 for all i = 1, 2,..., n and aij = 0 for i j.
1 0 0
The 3 3 identity matrix I 3 is, therefore, I3 0 1 0
0 0 1

7. An m m matrix A = aij is symmetric if aij = a ji for all i = 1, 2,..., m and all

m m

j = 1, 2,..., m .
1 1 5
0 1
For example, the 2 2 matrix and the 3 x 3 matrix 1 4 3 are symmetric.
1 3
5 3 2
8. An m m matrix A = aij is said to be upper triangular if aij = 0 for i > j and
m m

lower triangular if aij = 0 for i < j . Also, A is said to be triangular if it is either

upper triangular or lower triangular. For example, the 3 3 matrices:

3 2 1 1 0 0
0 1 2 and 1 2 0
0 0 1 1 1 1

are both triangular. The first is upper triangular and the second is lower triangular

DEFINITION

The transpose of an m n matrix A = aij , denoted by A T , is the n m matrix

m n

B = b ji defined by b ji = aij for j = 1,..., n and i = 1,..., m .

n m

Thus the jth row of AT has the same entries, respectively, as the jth column of A
T
and the ith. column of A has the same entries, respectively ,as the ith row of A .

Example
1 4
1 2 3
If A = , then A = 2 5 .
T
4 5 6
3 6
 DEFINITION

The sum of all the diagonal entries of an m m matrix A = aij m m

is called the trace of A

and is denoted by tr A .

Example
1 2
If A = , then tr A = 1 4 = 5.
3 4

2.4 OPERATIONS WITH MATRICES

DEFINITION

Let A and B both be m n matrices. The sum of A and B is the matrix

defined by :
A B aij bij
m n m n

aij bij
m n

a11 b11 a12 b12 a1n b1n

a 21 b 21 a 22 b 22 a 2n b 2n

a m 1 bm 1 a m 2 b m 2 a mn bmn

Note: We can only add matrices of the same size and the sum is found by adding the entries
appearing in corresponding positions. If A and B are matrices with differing sizes, then A B is
not defined.

Example
2 3 1 5 1 8 2 5 3 1 1 8 7 2 9
a). = =
5 1 2 2 3
2 1 0 2 3
5 2 1 1 2 0 2 3
7 2 2
 2 1 2 5 1
b). is not defined because the matrices have different sizes.
0 1 2 2
8 9 7 2 3

DEFINITION
Let A be a m n matrix and let k be any scalar (i.e. real number).
The scalar multiple kA is an m n matrix, given by :
kA = k a ij
m n

= k aij m n

k a11 k a12 k a 1n
k a 21 k a 22 k a 2n
=

k am 1 k am 2 k a mn

Note: Clearly A and kA are of the same size and kA is found by simply multiplying every entry
of A by k .

Example
1 2 3 5
Let A= and B = .
5 0 7 6
Determine 2 A 3B .

Solution

Since A and B are of the same size, viz 2 2 , 2A and 3B will also be of the size 2 2 .
Therefore, 2 A 3B is defined and
1 2 3 5
2 A 3B = 2 3
5 0 7 6
2 4 9 15
=
10 0 21 18
7 11
=
11 18
 Example

1 1 0 1 0 0 0 0 1 2
Let r s t u = .
0 0 0 0 1 0 1 1 3 2
Determine the value of r , s , t and u .

Solution

From the Definition of scalar multiple, we have :

r r 0 s 0 0 0 0 1 2
= .
0 0 0 0 t 0 u u 3 2

From the Definition of matrix addition, we obtain :

r r s 1 2
= .
t u u 3 2

Using the Definition of equality of matrices, we deduce :

r = 1, r s = 2, u =2 t u =3
1 s =2 t 2=3
s =1 t =1
r 1, s 1, u 2, t 1

The following THEOREM lists some important algebraic properties of matrix addition and scalar
multiples.

THEOREM

Let A , B and C be m n matrices and let 0 denote the zero m n matrix. For any scalar k
we have :

• A 0= A=0 A ( 0 is the additive identity)

• A B =B A (matrix addition is commutative)

• A B C =A B C (matrix addition is associative)

• k A B = kA kB

• 1 A=A 1

• 0A = 0 (0 on the left hand side is the number zero)

 DEFINITION

The additive inverse of an m n matrix A is the matrix A= a ij

m n

Remark: Clearly, if A is an m n matrix, then A A = 0m n.

DEFINITION

Let A and B be m n matrices. The difference, A from B is the matrix

B A = bij aij = bij aij = B A

m n m n m n

Example
Use the various algebraic properties of matrices to solve for the unknown matrix X appearing
in the matrix equation 2 X 3 A = B where X , A and B are of the same size.

Solution
2X 3A = B
2X 3A 3 A = B 3 A
2X 3A 3 A = B 3A
2X 0A = B 3A
2X 0 = B 3A
2X = B 3A
1
2 2X = 1
2 B 3A
1X = 1
2 B 3A
X = 1
2 B 3A

Before defining the product of two matrices, we first define what is called the inner product of a
row matrix with a column matrix.
 DEFINITION

Let A = aij be a row matrix with n entries, and B = bij a column matrix with n entries.
1 n n 1

The inner product of A with B is given by:

b11
b21
A B = a11 a12 ... a1n = a11b11 a12 b21 ... a1n bn1

bn1
which is a real number .

Note: The row and column matrix in the inner product must have the same number of entries,
otherwise the inner product is not defined.

Example
2
1
Let A = 2 3 4 5 and B = . Determine A 2B
2
5

Solution: A and B have the same number of entries, thus:

4
2
A 2B = A 2B = 2 3 4 5 = 2 4 3 2 4 4 5 10 = 48
4
10

Example
3
Let A = 2 3 7 and B = 2 . Determine A B A
1
Solution
3
A B 2 3 7 2 2 3 3 2 7 1 19
1
Thus A B A 19 A 19 2 3 7 38 57 133
 DEFINITION

Let A be an m n matrix and B an n p matrix. The matrix product of A and B is an

m p matrix C = cij , where cij is the inner product of the i th row of A and the j th
m p

column of B . That is :
Am n Bn p = Cm p = cij
m p

b1 j
b2 j
where cij = [ai 1 ai 2 ... ain ] = ai 1b1 j ai 2 b2 j ... ain bnj

bnj

Note: For the product AB to be defined, the number of columns of A must equal the number of
rows of B.

Example
2
Let A = and B = 7 3 . Determine AB if this product is defined.
1

Solution: First consider AB . Because the number of columns of A is the same as the number
of rows of B , the produce AB = C exists. Now :
c11 c12
A 2 1 B 1 2 = C 2 2 = c ij =
2 2 c 21 c 22
where cij is the inner product of the i th row of A with the j th column of B . Thus :
c11 2 7 2 7 14
c12 2 3 2 3 6
c21 1 7 1 7 7
c22 1 3 13 3
2 14 6
7 312 =
1 2 1
7 3 2 2
 Example
2 1 1 3 2
Let A = and B = . Determine AB and BA if these products exist and
5 8 4 2 0
comment on whether AB = BA .

Solution: Since the number of columns of A and the number of rows of B are the same, the
product A2 2 B2 3 = cij = C exists. From the definition of matrix multiplication cij is the inner
2 3

product of the i th row of A with the j th column of B , thus:

1 1
c11 2 1 2 1 1 4 2 c21 5 8 51 8 4 37
4 4
3 3
c12 2 1 2 3 1 2 4 c22 5 8 5 3 8 2 31
2 2
2 2
c13 2 1 2 2 1 0 4 c23 5 8 5 2 8 0 10
0 0

2 1 1 3 2 2 4 4
Therefore, AB
5 8 4 2 0 37 31 10

The product bij aij does not exist because the number of columns of B differs from
2 3 2 2

the number of rows of A . Therefore, AB BA .

In fact, in general AB BA and we say matrix multiplication is not commutative. Just in case
you thought this was the only surprising property the next example shows it is possible that AB = 0
where A 0 and B 0 .

Example

1 1 1 1
Let A = and B = . Calculate AB .
1 1 1 1

Solution
1 1 1 1 0 0
AB = = = 0.
1 1 1 1 0 0
Example
2 1 0 3
2 1 6
Let C = and D = 4 1 5 2 . Determine CD if it exists.
3 5 0
5 1 0 6
30 9 5 44
Solution: The product CD = A will exist and CD =
26 2 25 1

Example
2x 3y 2z = 8
Describe the system of equations : 5x y = 3 by means of a matrix equation.
2y 7z = 10

2 3 2 x 8
Solution: Let A = 5 1 0 , X = y , B= 3 .
0 2 7 z 10

2 3 2 x
AX = 5 1 0 y
0 2 7 3 3
z 3 1
Then:
2x 3y 2z 2x 3y 2z 8
= 5x y 0z = 5x y = 3 = B
0x 2y 7z 3 1
2y 7z 10

Therefore the equation AX = B represents the system.

THEOREM
For matrices A , B and C of the appropriate sizes, so that the products are defined, and scalars
k and c we have :

1. 0 A = 0, A0 = 0

2. IA = A , AI = A where I = I n is an identity matrix of suitable size

3. AB C = A BC (associative law of multiplication)

4. A B C = AB AC , A B C = AC BC are the distributive laws

 5. kA cB = kc AB

6. AT T
=A
T
7. kA = kAT
T
8. A B = AT BT
T
9. AB = B T AT

2.5 DETERMINANTS
With every square matrix we can associate a real number, referred to as the determinant of the
matrix. We start with square matrices of size 2 2 .

DEFINITION
a a12
Let A = 11 be a 2 2 matrix. The determinant of A is denoted by det A or A ,
a21 a22
a11 a12
and is defined to be: det A = = a11a 22 a21a12
a 21 a 22

Example 1.15
2 3
a) Find det A if A = .
5 1
2 3
Solution: A = =2 1 5 3 = 17.
5 1

a 0
b) Find det A where A = , where a, b .
0 b
a 0
Solution: A = = ab 0 = ab .
0 b
x2 x
c) Determine x such that = 0.
4 2
x2 x
Solution: Suppose = 0. Then we find that:
4 2
2 x 2 ( 4) x 0 2 x2 4x 0 2 x( x 2) 0 x 0 or x 2
The definition of the determinant of an n n matrix is done inductively. By this we mean we use
the Definition of the 2 2 case for the 3 3 case, we use the Definition of the 3 3 to define the
4 4 case, etc.

DEFINITION
a11 a12 a13
The determinant of the 3 3 matrix A = a21 a22 a23 is given by :
a31 a32 a33

a11 a12 a13

det a 21 a 22 a 23 a 22 a 23 a 21 a 23 a 21 a 22
= a11 det a12 det a13 det
a 31 a 32 a 33 a 32 a 33 a 31 a 33 a 31 a 33

Example
2 1 0
a) Let A = 5 2 2 . Evaluate det A .
1 0 5

Solution: From the Definition of the 3 3 case :

2 2 5 2 5 2
det A = 2 1 0 2 10 0 1 25 2 0 = 43
0 5 1 5 1 0

a 0 0
b) Find 0 b 0 .
0 0 c

a 0 0
b 0 0 0 0 b b 0
Solution: 0 b 0 = a 0 0 =a = a bc 0 = abc
0 c 0 c 0 0 0 c
0 0 c

Remark: We saw that :

a11 a12 a13

a 22 a 23 a 21 a 23 a 21 a 22
det a 21 a 22 a 23 = a11 det a12 det a13 det
a 32 a 33 a 31 a 33 a 31 a 32
a 31 a 32 a 33
and here we see that:

a22 a23
• is the matrix obtained by deleting the first row and first column
a32 a33

a21 a23
• is the matrix obtained by deleting the first row and second column
a31 a33

a21 a22
• is the matrix obtained by deleting the first row and third column
a31 a32

DEFINITION

Let A be an n n matrix and let M ij be the determinant of the n 1 n 1 matrix obtained

by deleting the i th row and the j th column of A . Then M i j is called the ij th minor of A.

Then we can write the determinant of a 3 3 matrix as :

a11 a12 a13
det a21 a22 a23
a31 a32 a33
a22 a23 a21 a23 a21 a22
= a11 det a12 det a13 det
a32 a33 a31 a33 a31 a32
= a11 M11 a12 M 12 a13 M 13
= a11 M11 a12 1 M 12 a13 M 13

Example
2 3 4
A 2 0 1 . Calculate det A .
5 6 7

Solution
 det A = 2 M 11 3M 12 4 M 13 , where
0 1
M 11 = =0 6= 6
6 7
2 1
M 12 = = 14 5 = 19
5 7
2 0
M 13 = = 12 0 = 12
5 6

det A = 2 6 3 19 4 12 = 12 57 48 = 93

DEFINITION

i j
Let A be an n n matrix. The ij th co-factor of A is given by Aij 1 M ij .

The last description is referred to as the determinant obtained by cofactor expansion along the
first row, because the a11 , a12 and a13 are the entries of the first row.

Example
2 3 1
Let A = 1 0 4 .
7 3 5
Find det A by cofactor expansion of : (a) the first row (b) the second column.

Solution

a) Using the cofactor expansion of the first row :

det A = a11A11 a12 A12 a13A13
= 2A11 3A12 1A13
1 1 1 2 1 3
= 2 1 M 11 3 1 M 12 1 1 M 13
0 4 1 4 1 0
= 2 3 1
3 5 7 5 7 3
= 2 0 12) 3(5 28 13 0
= 24 69 3 = 48
 b) Using the cofactor expansion of the second column :
det A = a12 A12 a22 A 22 a32 A 32
= 3A12 0A 22 3A 32
= 3A12 3A 32
1 2 3 2
= 3 1 M 12 3 1 M 32
= 3M 12 3M 32
1 4 2 1
= 3 3
7 5 1 4
= 3 5 28 3 8 1
= 3 23 7 = 3 16 = 48

Notice that we obtained the same determinant value although we used different cofactor expansions.
In fact, in general the cofactor expansion can be done according to any row or column and the
determinants value will always be the same. We now define the determinant of an n n matrix.

DEFINITION
a11 a12 a1n
a a22 a2 n
Let A = 21 be an n n matrix. The determinant of A by cofactor

an1 an 2 ann
expansion of the first row is given by:
n
det A = a11 A11 a12 A12 ... a1n A1n = a1k A1k
k =1

Remark: Evaluate the determinant by choosing a suitable row or column

We select the row or column containing the most zeros.

Example
2 3 0 4
5 1 2 0
Evaluate det A , where A =
3 1 0 1
0 3 0 2
 Solution: We use cofactor expansion along the third column:
det A = 0A13 2A 23 0A33 0A 43
= 2A 23
2 3 4
2 3
= 2 1 3 1 1
0 3 2
2 3 4
Let B = 3 1 1 . We now progress along the first column as follows:
0 3 2
2 3 4
|A | = 23 1 1
0 3 2
= 2 2B 11 3B 21
= 4B 11 6B 21
1 1 1 1 2 1 3 4
= 4 1 6 1
3 2 3 2
= 4 2 3 6 6 12 = 20 36 = 16

Clearly, the more zeros we can obtain in a row or column, the easier a “large'' determinant is to
evaluate.

2.6 PROPERTIES OF DETERMINANTS

The following properties of determinants can simplify computation to a great extent.
Let A be an n n matrix.

1. If any row or column of A consists entirely of zeros, then det A = 0 .

2. det AB = det A det B for any n n matrices A and B .
3. det( A) det( AT )
4. The determinant of a triangular matrix is the product of the entries on
the diagonal.
5. The interchanging of two rows (or columns) of A changes the sign of
det A .
6. If any non-zero multiple of a row (or column) is added to a different row (or column),
then det A is unchanged.
2.7 SYSTEMS OF LINEAR EQUATIONS : TERMINOLOGY
Having discussed matrices and determinants of matrices, we can now progress to the actual
solution techniques that could be employed to solve a system of linear equations.
This section deals with terminology and some general results concerning systems of linear
equations.

DEFINITIONS

Let m and n be natural numbers. Then :

a) An equation a1 x1 a2 x2 ... an xn = b where a1 , a2 ,..., an , b are real numbers and the

x1 , x2 ,..., xn are unknowns, is called a linear equation in n unknowns.

b) A system
a11 x1 a12 x2 ... a1n xn = b1
a21 x1 a22 x2 ... a2 n xn = bn

am1 x1 am 2 x2 ... amn xn = bn

of m linear equation in n unknowns x1 , x2 ,..., xn is called a homogeneous system if
bi = 0 for every i = 1, 2,..., n.
If bi 0 for at least one i = 1, 2,..., n then the system is called a non-homogeneous
system of linear equations.

c) An ordered n -tuple c1 , c2 ,..., cn is a solution of the system if the substitution xi = ci for

i = 1, 2,..., n satisfies every equation in the system.
The set of all solutions of the system is called the solution set of the system.

Remark: A homogeneous system of m equations in n variables x1 , x2 , , xn always has at least

one solution namely ( x1 , x2 , , xn ) = (0, 0, , 0) .

You would probably have encountered a linear equation in two variables before. Such an equation
can be represented graphically as a straight line on the Cartesian plane.

Consider a system of two equations in the variable x1 and x2 :

a11 x1 a12 x2 = b1
a21 x1 a22 x2 = b2 .
The solution of this system is an ordered tuple c1 , c2 such that the substitution x1 = c1 , x2 = c2
satisfies both equations.
 Geometrically c1 , c2 corresponds to a point x1 , x2 = c1 , c2 in the Cartesian plane where the
straight lines are represented by :
L1 : a11 x1 a12 x2 = b1
L2 : a21 x1 a22 x2 = b2
intersect.

Therefore, a solution of the system is quite simply a point of intersection of the straight lines.
The solution set is the set of all points of intersection. Naturally the question ''How many
solutions could there be?'', comes to mind.
The following graphs indicate the possible nature of the solution set of the system :
No intersection Infinitely many One intersection
points intersection points point L1
L1 = L2
L1
L2
L2

Therefore, the system :

a11 x1 a12 x2 = b1
a21 x1 a22 x2 = b2

could have no solution, an infinite number of solutions or a unique solution.

In general it is true that any system of m linear equations in n unknowns has either no solution,
a unique solution or an infinite number of solutions. A system with no solution is called an
inconsistent system.

2.8 CRAMER’S RULE

This technique can be used if the system of linear equations is such that :

• the coefficient matrix A is square, and

• det A 0.
If these conditions are met, then the system has a unique solution. We state Cramer's
THEOREM for a system of three equations in three unknowns. However, this can be
generalized to a system of n linear equations in n unknowns.

THEOREM

(Cramer's Rule).
a11 x1 a12 x2 a13 x3 = b1
Let a21 x1 a22 x2 a23 x3 = b2
a31 x1 a32 x2 a33 x3 = b3
be a system of equations with coefficient matrix A = aij .

If det A 0 , then the system has a unique solution x1 , x2 , x3 , given by

det Aj
xj = ; j = 1, 2, 3
det A
where Aj is obtained from A by replacing the j -th column of A with the column of constants :
b1
b2
b3

det Aj
Remark: When you consider x j = , do realize that Aj is obtained by removing the column
det A
of coefficients of x j from A and replacing them with the constants.
Clearly, in the THEOREM we have :
b1 a12 a13 a11 b1 a13 a11 a12 b1
A1 = b 2 a22 a23 , A 2 = a21 b 2 a23 , A 3 = a21 a22 b 2
b3 a32 a33 a31 b3 a33 a31 a32 b3

An advantage of Cramer's Rule is that you are able to solve for a specific unknown.

Example 1.23
Use Cramer's Rule to solve for the unknown y appearing in the non-homogeneous system :
x y z = 0
y z = 1
2x z = 3
 1 1 1
A = 0 1 1
2 0 1
1 1 1
Solution: = 0 1 1 R1,3 2
0 2 1
1 1 1 1
= 1 1
2 1
= 11 2 = 1
det A2
Because A 0 , we can employ Cramer's Rule and y=
det A
where A2 is formed by replacing the coefficients of y appearing
in the coefficient matrix A with the constants. Thus :
1 0 1
A2 = 0 1 1
2 3 1
1 0 1
= 0 1 1 R1,3 2
0 3 1
1 1 1 1
= 1 1
3 1
= 1 1 3 =2
det A2 2
y= = = 2.
det A 1

One disadvantage of Cramer's Rule is that it can only be applied to a restricted class of linear
systems.

The following technique can be employed to solve any system of linear equations.

2.9 THE INVERSE MATRIX

One of the most important properties that a square matrix can have, is invertibility.

DEFINITION

An n n matrix A is invertible if there exists an n n matrix B such that AB = BA = I n . If

such a B exists, it is called the inverse of A and we denote it by A 1.

Naturally, for a given matrix A two questions need to be answered, viz.

• When does a given n n matrix A have an inverse?

1
• If such an inverse A exists, how is it calculated?

The answer to the first question is given by the following THEOREM.

THEOREM

Let A be an n n matrix.
1
A is invertible (i.e. A exists) det A 0.

1
Therefore, for a given n n matrix A , if det A 0 , we know A exists.
To find the inverse of a n n matrix A , if det A 0 , you can use your scientific calculator or
software (simply search for “matrix calculator”).

To find the inverse without softtware or the calculator, we use a process called Gaussian
Elimination, this process will be explained in the next secion.

One application of the inverse matrix is its use in solving a system of equations with
• square coefficient matrix A and
• det A 0.
Notice that these are the same conditions required when we were discussing Cramer's Rule.
Suppose we had matrix equation representation of a system of equation in which the coefficient
matrix satisfies the two conditions stated above.
Then A 1 exists and
AX = B
1
A A X = A 1B
A 1A X = A 1B
I n X = A 1B , where A is an n n matrix
X = A 1B .
Because matrix multiplication is NOT commutative, we need to ensure we use
X = A 1B and NOT X = BA 1 .
 Example
2x 3y = 2
Solve the system using the inverse matrix method.
x 2y = 4

Solution: The matrix equation of the system is AX = B , where

2 3 x 2
A= , X = , B = .
1 2 y 4
2 3 1
Because A = = 4 3 = 1 , we have det A 0 which implies A exists.
1 2
2 3
Using software we have A 1 = .
1 2
Therefore X = A 1 B

x 2 3 2
=
y 1 2 4
4 12
=
2 8

x 8
=
y 6

x = 8, y = 6 is the solution of the system.

2 8 3 6 = 16 18 = 12
Check :
8 2 6 = 8 12 = 4
And x = 8, y = 6 satisfies both equations.

One advantage of this method of solving a system of linear equations is that should we replace
the constant matrix B in AX = B , but kept A unchanged, the solution is easily obtained.

Example
Consider the same system as in example 30, but with different constant terms
2x 3y = 6
x 2y = 0
Use the inverse matrix to solve the system.

2 3 x 6
Solution: Here AX = B , with A = ,X = ,B= .
1 2 y 0
 2 3
As before A 1 = , so:
1 2
x
= A 1B
y
2 3 6
=
1 2 0
x 12
=
y 6

x = 12, y = 6 is the unique solution of the system.

Example
z = 2 2x
Solve the system : y z = 3
x z = 4
using the inverse matrix if possible.

Solution: The matrix equation representing the system is AX = B where

2 0 1 x 2
A = 0 1 1 , X = y , B = 3 and
1 0 1 z 4
2 0 1
A = 0 1 1 = 1( 1) 2 2
2 1 = 3
1 0 1
1
Since A 0, A exists and we can use the inverse matrix method.
Then:
1
3 0 1
3 1 0 1
1 1
A = 1
3 1 2
3 = 1 3 2 .
3
1
3 0 2
3 1 0 2
Therefore, X = A 1B gives :

x 1 0 1 2 2 4 2 2
3
1 1 1
y = 1 3 2 3 2 9 8 = 1 = 1
3
3 3 3
z 1 0 2 4 2 8 10 10
3

x = 2
3 , y = 1
3 , z = 103 is the solution of the system.
2.10 GAUSSIAN ELIMINATION AND THE AUGMENTED
MATRIX
The basic idea used in this technique is that we replace a given system with a simpler system of
equations having the same solution set. To obtain a simpler system with the same solution set as the
original system, we restrict manipulation of the equation to the following elementary three row
operations:
• Replacement of an equation with the sum of the equation and any multiple of
a different equation.
• Replacement of an equation with a nonzero multiple of itself.
• Interchanging of any two equations.
Since the equations are in rows, the above elementary row operations correspond to the meaning
associated with + . , ( ), ) encountered earlier.

It can be proved formally that performing these operations on a system does not change the solution
set. NOTE: This is only applicable to elementary row operations and not elementary column
operations.

Systems of equations having the same solution set are called equivalent systems.

Example
Solve the system, using only elementary row operations :

2 x1 4 x2 6 x3 = 18
4 x1 5 x2 6 x3 = 24
3x1 x2 2 x3 = 4

Solution: Notice that the variables x1 are listed beneath one another and the same applies to the
variables x2 and x3 . For the solution we find :

2 x1 4 x2 6 x3 = 18
4 x1 5 x2 6 x3 = 24
3x1 x2 2 x3 = 4
1x1 2 x2 3 x3 9 1
2 R1 x1 2 x2 3x3 9
4 x1 5 x2 6 x3 24 3 x2 6 x3 12 R2 R2 4 R1
3 x1 x2 2 x3 4 3 x1 x2 2 x3 4

x1 2 x2 3 x3 9 x1 2 x2 3x3 9
3x2 6 x3 12 x2 2 x3 4 1
3 R2
5 x2 11x3 23 R3 R3 3 R1 5 x2 11x3 23

x1 2 x2 3x3 9 x1 1x3 1 R1 R1 2 R2
x2 2 x3 4 x2 2 x3 4
1x3 3 R3 R3 5 R2 x3 3 1 R3

x1 4 R1 R1 R3
x2 2 R2 R2 2 R3
x3 3

Therefore, the system has a unique solution with solution set :

x 1, x 2 , x 3 | x 1 = 4, x 2 = 2, x 3 = 3
DEFINITION

Consider the following system of m equations in n unknowns :

a11 x1 a12 x2 a1n xn = b1
a21 x1 a22 x2 a2 n xn = b2

am1 x1 a2 m x2 amn xn = bm

a11 a12 a 1n b1
a 21 a 22 a 2n b2
Then is the augmented matrix of the system.
am 1 am 2 amn bm

Example
Solve the system
2 x1 4 x2 6 x3 = 18
4 x1 5 x2 6 x3 = 24
3x1 x2 2 x3 = 4
using Gaussian elimination and the augmented matrix of the system.

Solution: This is the system discussed in the previous example and its augmented matrix is :

2 4 6 18
4 5 6 24
3 1 2 4

Proceeding with the solution :

2 4 6 18 1
2
R1 1 2 3 9 1 2 3 9
4 5 6 24 4 5 6 24 R2 R2 4 R1 0 3 6 12
3 1 2 4 3 1 2 4 3 1 2 4

1 2 3 9 1 2 3 9 1 2 3 9
0 3 6 12 1
3 R2 0 1 2 4 0 1 2 4
R3 R3 3 R1 0 5 11 23 0 5 11 23 R3 R3 5R2 0 0 1 3

R1 R1 2 R2 1 0 1 1 R1 R1 R3 1 0 4
0 1 2 4 R2 R2 2 R3 0 1 0 2
1 R3 0 0 1 3 0 0 1 3

This augmented matrix represents the system :

 x1 = 4
x2 = 2
x3 = 3

which gives the solution set ( x 1 , x 2 , x 3 ) | x 1 = 4, x 2 = 2, x 3 = 3

Note how the Gaussian Elimination technique is a systematic attempt to reduce the coefficient matrix
of the system to an identity matrix.
Example 1.24
Solve :
x1 2 x2 x3 = 2
2 x1 3x2 5 x3 = 5
x1 3 x2 8 x3 = 1

Solution: Consider the augmented matrix

1 2 1 2
2 3 5 5
1 3 8 1

1 2 1 2 R1 R1 2 R2 1 0 13 4
R2 R2 2 R1 0 1 7 1 0 1 7 1
R3 R3 R1 0 1 7 1 R3 R3 R2 0 0 0 0 Note row of zeroes

1 0 13 4
1 R2 0 1 7 1
0 0 0 0
The last augmented matrix represents the system :
x1 13x3 = 4
x2 7 x3 = 1
Therefore, all solutions x1 , x2 , x3 of the system will satisfy
x1 = 4 13x3
x2 = 1 7 x3
x3 ,
The system, therefore, has an infinite number of solutions and the solution set is :
x 1 , x 2 , x 3 | x 1 = 4 13x 3 , x 2 = 1 7 x 3 , x 3 .
We can obtain particular solutions by assigning a specific value for x3 , for example, if
x3 = 1 then x1 = 9, x2 = 6, x3 = 1 is a particular solution of the system.

Example 1. 25
Determine the solution set of :

x y 2z = 1
x 2z = 0
x y 4z = 2
4x 3y 4z = 2

Solution:
1 1 2 1 1 1 2 1 1 1 2 1
1 0 2 0 R2 R2 1 R1 0 1 4 1 0 1 4 1
1 1 4 2 R3 R3 1 R1 0 0 2 1 0 0 2 1
4 3 4 2 R4 R4 4 R1 0 1 4 2 R4 R4 R2 0 0 0 1

The last row of this augmented matrix states 0 x1 0 x2 0 x3 = 1, which is impossible.

The system is, therefore, an inconsistent system and the solution set is the empty
set or .

Example 1.26
Solve :
x y 2z = 1
x 2z = 0
x y 4z = 2
4x 3y 4z = 3
Solution
 1 1 2 1 1 1 2 1 1 1 2 1
1 0 2 0 R2 R2 1 R1 0 1 4 1 0 1 4 1
1 1 4 2 R3 R3 1 R1 0 0 2 1 1
2 R3 0 0 1 1
2

4 3 4 3 R4 R4 4 R1 0 1 4 1 R4 R4 R2 0 0 0 0 Not contradictory

R1 R1 2 R3 1 1 0 0 R1 R1 R2 1 0 0 1
R2 R2 4 R3 0 1 0 1 0 1 0 1
0 0 1 1
2 0 0 1 1
2

0 0 0 0 0 0 0 0

Therefore, the solution is represented by x 1, y 1 and z 1

2 .

The solution set consists of an unique : x; y; z | x 1, y 1, z 1

2 .

1
For a given n n matrix A , if det A 0 , we know A exists.
We employ Gaussian Elimination to perform the reduction :

1
A | In Gaussian Elimination In | A
where A is an n n matrix with det A 0.

Example 1.27
2 3
Determine A 1 if it exists, where A = .
1 2
2 3 1
Solution: Because A = = 4 3 = 1 , we have det A 0 which implies A exists.
1 2

1 1
Determine A by doing the reduction : A | I 2 G.E. I 2 | A .

Hence,
 2 3 1 0
1 2 0 1

R1 R2 1 2 0 1 1 2 0 1
2 3 1 0 R2 R2 2 R1 0 1 1 2

R1 R1 2 R2 1 0 2 3 1 0 2 3
0 1 1 2 1 R2 0 1 1 2
2 3
Therefore, A 1 =
1 2
Check that this is correct : A A 1 = I 2 and A 1 A = I 2 must be true.

1 2 3 2 3
AA =
1 2 1 2

2(2) 3 1 2 3 3 2 1 0
= = = I2
1 2 2 1 1 3 2 2 0 1

Similarly, you can verify that A 1 A = I 2 also holds.

 CHAPTER 2 EXERCISES
1.1 Write down one example of each of the following :

a) a 1 3 matrix A
b) the zero 2 3 matrix 0
c) a 3 3 diagonal matrix C such that C I3
d) I 4

e) a column matrix D = dij

2 1

f) a row matrix E = eij

1 2

2 0 2
2 1 3 3 2 1 3 2
1. 2 Let A = , B= , C= and D = 1 3 0 .
0 5 1 1 0 3 1 1
1 1 1

Determine the following if they are defined :

T T
a) B A , C B , AD , tr CC , BD , 2 A 4 B , tr AD
T T
b) A B D , AD BD , tr CAD , tr 2C
c) D 2 where D n = D D ... D for n factors of D , where n

1 1 1
3
1.3. Show that A = I 3 , where A = 2 1 0 .
1 0 0

1.4 Determine det A using cofactor expansion according to the row or column of your choice :

0 6 0 3 1 0 1 3 7
a) A = 8 6 8 b) A = 2 4 3 c) A = 2 0 8
3 2 2 5 4 2 1 3 4

1.5 Determine A 1 , if it exists :

1 2 3 3 1 5
1 2
a) A= b) A = 2 5 3 c) A = 2 4 1
1 2
1 0 8 4 2 9
 1 1 1
5 11 7 3
5 5 5
2 1 4 5
d) A= 1
5
1
5
4
5
e) A =
2 1 1
3 2 8 7
5 10 10
0 0 0 0

1.6 Solve the following systems, using inverses of matrices. You may assume they exist

a) b)
2x 4y 6 z = 18
3 x1 4 x2 = 5
4x 5y 6 z = 24
2 x1 x2 = 4
3x y 2z = 4

c) d)
2 x1 3 x2 x3 = 3 2x y z = 4
x1 2 x2 x3 = 1 3x 2y 4z = 11
x1 4 x2 = 2 6x 8y 4 z = 22

1.7 Use Cramer's Rule to show that for the system :

x y z = 0
y z = 1
2x z = 3

x = 1, y = 2 , z = 1 is the unique solution of the system.

1.8 Use Cramer's Rule to determine the solution set of the system :

x1 2 x3 = 6
3x1 4 x2 6 x3 = 30.
x1 2 x2 3 x3 = 8

1.9 Use Cramer's Rule (if possible) to solve:

2 x1 4 x2 6 x3 = 18
4 x1 5 x2 6 x3 = 24
3x1 x2 2 x3 = 4
1.10 Use Gaussian elimination to solve the following systems :

a) b)
x1 2 x2 3x3 = 11 3 x1 6 x2 6 x3 = 9
4 x1 x2 x3 = 4 2 x1 5 x2 4 x3 = 6
2 x1 x2 x3 = 10 x1 16 x2 14 x3 = 3

c) d)
x1 x2 x3 = 7 x1 x2 x3 = 0
4 x1 x2 5 x3 = 4 4 x1 x2 5 x3 = 0
6 x1 x2 3 x3 = 20 6 x1 x2 3x3 = 0

e) f)
x1 2 x2 x3 x4 = 2
x1 2 x2 4 x3 = 4 3x1 2 x3 2 x4 = 8
2 x1 4 x2 8 x3 = 9 4 x2 x3 x4 = 1
x1 6 x2 2 x3 = 7

g) h)
x1 x2 = 4 x1 x2 = 4
2 x1 3x2 = 7 2 x1 3 x2 = 7
3x1 2 x2 = 8 3 x1 2 x2 = 11