MAT 260 Module 1 PDF

MODULE
LINEAR ALGEBRA- MAT 260

MODULE 1
COPPERBELT UNIVERSITY COLLEGE

Mathematics department
Copyright
© COPPERBELT UNIVERSITY COLLEGE 2011
COPPERBELT UNIVERSITY COLLEGE

BOX 20382
KITWE
ZAMBIA
Fax: +[Add country code] [Add area code] [Add telephone #]
E-mail: cosetco@zamtel.zm
Website: www.[Add website address]
Acknowledgements
The COPPERBELT UNIVERSITY COLLEGE wishes to thank the following for their
contribution to this module.
ZACODE
DODE
Mr. P Banda
Mr. D Kambilombilo
Mr. P. Mwewa
Ms. R Mvula
Contents
About this MODULE 1
How this module is structured ....................................................................................... 1
Module overview 3
Welcome to module 1 MAT 260 .................................................................................... 3

Module 1 MAT 260 ......................................................................................................... 3
Is this module for you? ................................................................................................... 3
Module outcomes ............................................................................................................ 3
Timeframe ....................................................................................................................... 4
Study skills ....................................................................................................................... 4
Need help? ....................................................................................................................... 6
Assignments ..................................................................................................................... 6
Assessments ..................................................................................................................... 7
Getting around this Module 7
Margin icons .................................................................................................................... 7
Unit 1 8
Systems of linear equations ...............................................Error! Bookmark not defined.

1.0 Introduction ..................................................................................................... 8
1.1 Linear Equations in n unknowns. .................................................................. 9
1.2 Solutions of linear equations in n- unknowns ............................................. 21
1.3 Systems of linear equations ...............................Error! Bookmark not defined.
1.4 Solutions of systems of linear equation
1.5 Gaussian elimination method ............................Error! Bookmark not defined.
summary .............................................................................Error! Bookmark not defined.

Assignment .........................................................................Error! Bookmark not defined.
Assessment ..........................................................................Error! Bookmark not defined.
Unit 2 79
Matrix Algebra .............................................................................................................. 79

2.1 Introduction........................................................................................................... 79
Unit summary .....................................................................Error! Bookmark not defined.

Assignment .........................................................................Error! Bookmark not defined.
Assessment ..........................................................................Error! Bookmark not defined.
U LINEAR ALGEBRA- MAT 260
About this module

MODULE 1 has been produced by COPPERBELT
UNIVERSITY COLLEGECOPPERBELT UNIVERSITY
COLLEGE and structured in the same way as outlined below.
How this module is structured
The module overview

The module overview gives you a general introduction to the
module. Information contained in the module overview will help
you determine:
 If the module is suitable for you.
 What you will already need to know.
 What you can expect from the module.
 How much time you will need to invest to complete the module.
The overview also provides guidance on:
 Study skills.
 Where to get help.
 Module assignments and assessments.
 Activity icons.
 Units.
We strongly recommend that you read the overview carefully

before starting your study.
1
U LINEAR ALGEBRA- MAT 260
The module content

The module is broken down into units. Each unit comprises:
 An introduction to the unit content.
 Unit outcomes.
 New terminology.
 Core content of the unit with a variety of learning activities.
 A unit summary.
 Assignments and/or assessments, as applicable.
Resources
For those interested in learning more on contents in this module,
we provide you with a list of additional resources at the end of this
module these may be books, articles or web sites.
Your comments
After completing this module we would appreciate if you would
take a few moments to give us your feedback on any aspect of this
module. Your feedback might include comments on:
 Module content and structure.
 Module reading materials and resources.
 Module assignments.
 Module assessments.
 Module duration.
 Module support (assigned tutors, technical help, etc.)
Your constructive feedback will help us to improve and enhance

this module.
2
U
Module overview
Welcome to module 1 MAT 260

This module is based on introductory concepts in Mathematics and is part of foundation
mathematics course.
Module 1 MAT 260
Is this module for you?

This module is intended for Diploma holder teachers of Mathematics that aim to further their
teaching skills.
Module outcomes
Upon completion of module 1 you will be able to:
3
U
i) Solve systems of equations

ii) Use the Jordan Gaussian method
iii) Appreciate properties of matrices in proofs
Outcomes
Time frame
The expected study time is 100 hours
How long?
Study skills
As an adult learner your approach to learning will be different to that from

your school days: you will choose what you want to study, you will have
professional and/or personal motivation for doing so and you will most
likely be fitting your study activities around other professional or domestic
4
U
responsibilities.
Essentially you will be taking control of your learning environment. As a

consequence, you will need to consider performance issues related to time
management, goal setting, stress management, etc. Perhaps you will also
need to reacquaint yourself in areas such as essay planning, coping with
exams and using the web as a learning resource.
Your most significant considerations will be time and space i.e. the time you
dedicate to your learning and the environment in which you engage in that
learning.
We recommend that you take time now—before starting your self-study—

to familiarize yourself with these issues. There are a number of excellent
resources on the web. A few suggested links are:
 http://www.how-to-study.com/
The “How to study” web site is dedicated to study skills resources. You
will find links to study preparation (a list of nine essentials for a good
study place), taking notes, strategies for reading text books, using
reference sources, test anxiety.
 http://www.ucc.vt.edu/stdysk/stdyhlp.html
This is the web site of the Virginia Tech, Division of Student Affairs.
You will find links to time scheduling (including a “where does time
go?” link), a study skill checklist, basic concentration techniques,
control of the study environment, note taking, how to read essays for
analysis, memory skills (“remembering”).
 http://www.howtostudy.org/resources.php
Another “How to study” web site with useful links to time management,
efficient reading, questioning/listening/observing skills, getting the most
out of doing (“hands-on” learning), memory building, tips for staying
motivated, developing a learning plan.
The above links are our suggestions to start you on your way. At the time of
5
U
writing these web links were active. If you want to look for more go to
www.google.com and type “self-study basics”, “self-study tips”, “self-study
skills” or similar.
Need help?
For any help contact the University College through:
Help 1. University College Secretary - Phone +260 212293003
2. email: cosetco@zamtel.zm
3. The Mathematics HOD -
4. District Resource centre in your district -
All contacts during Office hours.
Assignments
Three, tutor marked assignments will be considered for assessment.
To be submitted to the university college through District Resource

Centres before or at stipulated times in the units.
Assignments
6
U
Assessments
There will be one Tutor marked test, at the end of this module, and
several self- marked tasks at the end of each unit.
Assessments A two hour test will be administered at the University College
during a residential school as advised
Getting around this module
Margin icons
While working through this module you will notice the frequent use of margin icons. These icons
serve to “signpost” a particular piece of text, a new task or change in activity; they have been
included to help you to find your way around this module.
A complete icon set is shown below. We suggest that you familiarize yourself with the icons and
their meaning before starting your study.
Activity Assessment Assignment Case study
Discussion Group activity Help Note it!
7
U
Outcomes Reading Reflection Study skills
Unit 1
Linear equations
1.0 Introduction
From module 1 of MAT 110, you were introduced to linear equations in one and two unknowns
such as ax  c and ax  by  c , where a, b and c are real numbers.
In this unit we will explore further the concepts covered in Linear Algebra in introductory
Mathematics, MAT 110 and look at linear equations in general, that is linear equations in several
variables or unknowns. In this unit we will expose you to the most economic way of
manipulating and solving systems of linear equations.
Upon completion of this unit you will be able to:
 State the type of solution the system of linear equation has.
 Solve systems of linear equations using the Jordan Gaussian

elimination method
Outcomes
 Perform proofs of some theorems
8
U
1.1 Definition of the general linear equation
Let us consider a linear equation in any arbitrary number of unknowns, say

x1 , x 2 , x3 , x n , then;
A linear equation in n unknowns x1 , x 2 , x3 , x n over a real field  is an equation of the form
a1 x1  a 2 x 2    a n x n  b , where a1 , a 2 , a n and b are given real numbers
and the xi are in determinants or variables. The numbers ai are coefficients of xi

and b is called the constant of the equation or simply the constant.
A set of values for the unknowns that is an n – tuple of real numbers   (k1 , k 2,  , k n ) that is an
n – component row vector, is a solution of the above equation if, on substituting k i for xi in the
equation a1 k1  a 2 k 2  a3 k 3   a n k n  b is true.
The set of values is then said to satisfy the equation.
Example 1.1
Determine the solution of the equation
x  2 y  4z  w  3
Solution
To find the solution to this equation, we are looking for real numbers   (k1 , k 2 , , k n ) such that
a1 k1  a 2 k 2    a n k n  b is a true statement.
Now   (3, 2,1, 0) is one of the solutions of the equation since

3  2(2)  4(2)  0  3 is a true statement.
On the other hand
  (1, 1, 2, 2) is not a solution of the equation since
1  2(1)  4(2)  2  3 is not a true statement
If there is no ambiguity about the position of the unknowns in the equation, then we denote this
solution by simply the n-tuple   (k1 , k 2,  , k n ) .
9
U
1.2 Solving linear equations

.
Consider the linear equation in two unknowns, x and y below.
3x  2 y  4
To solve this equation we seek to find a pair of numbers   (k1 , k 2 ) (since the equation given is
in two unknowns) which satisfies the equation, that is for which ak1  bk 2  c is a true statement
Thus, solutions of this equation can be obtained by assigning arbitrary values to x and solving
for y and vice versa.
For instance, if we substitute x  0 in the equation, we will get
3(0)  2 y  4
2y  4
y2
Hence (0, 2) is a solution.
Similarly when we substitute x  2 in the equation, we obtain
3(2)  2 y  4
2y  4  6
y  10
y5
Hence (2, 5) is a solution.
The table below gives some possible values for x and the corresponding values for y .
x y
-2 5
-1 4.5
0 2
1 0.5
2 -1
Table 1.1
These points give the graph below
10
U
Now any solution   (k1 , k 2 ) of the linear equation ax  by  c determines a point in the
Cartesian plane R2. (i.e in two dimension)
 If a and b, are not both zero, the solutions of the linear equation correspond precisely to
the point on a straight line (hence the name linear equation)
 A linear equation is said to be degenerate if the coefficients of the unknowns are all zero.
There are two cases when this is possible:
Case I
The constant is not zero that is the equation is of the form 0 x1  0 2 x2    0 n x n  b
with b  0
Then there is no solution to the linear equation.
Case II
The constant is also zero that is the equation is of the form
0 x1  0 x 2    0 x n  0
11
U
Then every n – tuple of real numbers is a solution of the equation.
1.3 System of linear equations

1.3.1Definition
A system of m linear equations in n unknowns x1 ; x 2  x n is of the form,
a11 x1  a12 x 2    a1n x n  b1
a 21 x1  a 22 x 2    a 2 n x n  b2
…………………………….....
a m1 x1  a m 2 x 2    a mn x n  bm ,
where the aij and b are real numbers. An n – tuple of numbers   (k1 , k 2 , , k n ) which satisfies
all the equations is called a solution of the system.

1.3.2 Solutions of systems of linear equations
Let us consider a system of two linear equations in two unknowns x and y
a1 x  b1 y  c1
a 2 x  b2 y  c2
A pair of numbers which satisfies both equations is called a simultaneous solution of the given
equations or a solution of the system of equations.
There are three (3) cases which can be described geometrically.
(Here we assume that the coefficient of x and y in each equation are not both zero)
Case I.
Consider the equations
x–y=–3
x + 2y = 3
Find the solution of the given system of equations.

To find the solutions to this system of equations, we can apply the knowledge on solving
simultaneous equations using the graphical method.
I m sure this is not a problem.
In case you have forgotten, refer to the graph below:
12
U
Thus the system of equations has exactly one solution, (– 1, 2).

Here the lines corresponding to the linear equations intersect in one point (– 1, 2) as shown in
figure 1.1 above.
Case II
Consider the pair of equations:
x+y=1
2x + 2y = 6
Find the solutions of the given system of equations.
To answer this question we will use the same approach used in the previous question. If we do
the correct thing, we will discover that the system of equation above has no solution.
13
U
Here the lines corresponding to the linear equations are parallel as shown in figure 1.3 above.
Therefore, the system has no solutions.
Case III
Consider the pair of equations
x  y 1
3x  3 y  3
Find the solutions to the system of equation above.
If we look at the system of equations above closely, we will discover that the two equations
represent just one equation, x  y  1 .
Refer to figure 1.4 below
14
U
As such the system has an infinite number of solutions.

Here the lines corresponding to the linear equations coincide as shown in figure 1.4 above.
Cases II and III can occur if and only if (iff) the coefficients of x and the coefficients of
y are proportional:
a1 b
 1 .
a 2 b2
There are two possibilities for this to occur:
(i) If the constant terms of the equations are in the same proportion that is
a1 b c
 1  1 ,
a 2 b2 c 2
then the lines are coincident.
Hence the system has an infinite number of solutions which correspond to the solutions of either
equation.
15
U
(ii) If the constant terms of the equations are not in the same proportion, that is:
a1 b c
 1  1 ,
a 2 b2 c 2
then the lines are parallel.
Hence the system has no solution.
Example 1.2
a) Determine the solution of the system of equation below:
3x + 6y = 9
2x + 4y = 6
Solution
To find the answer to this problem, we shall compare the ratios of the constants as given above,
that is
a1 b c
 1  1
a 2 b2 c 2
3 6 9
  
2 4 6
3 3 3
  
2 2 2
Since the ratio of the constants is in the same proportional, then the lines are coincident and the
system has an infinite number of solutions which correspond to the solutions of either equation
by (i) above.
b) Determine the solution of the system of equations below:

3x + 6y = 9
2x + 4y = 5
16
U
Solution
To answer this question we will again compare the ratios of constants as follows:
a1 b c
 1  1
a 2 b2 c 2
3 6 9
  
2 4 5
3 3 9
   ,
2 2 5
Since the ratios of the constants are not in the same proportional, then the lines are parallel and
the system has no solution by (ii) above.
If the coefficients of the unknowns are not proportional that is

a1 b
 1 ,
a 2 b2
then case I occurs, that is, the system has a unique solution.
Example
Determine the solution of the system of equation below.
x  3y  9
2x  y  4
Solution
To answer this question, we will compare the ratios of constants as follows:
a1 b1

a 2 b2
1 3
 
2 1
Now since the ratios of the constants are not equal, then the system of linear equations intersect
at one point and as such the system of linear equation has a unique or one solution.
17
U
Activity 1.1
1. Find the conditions which h and k must satisfy for the following systems of linear
equations to have:
a) A unique solution b) no solution c) an infinite number of solutions.
i) 2 x1  3 x2  x3  5 ii) x1  x 2  4 x3  5 iii) x1  x 2  x3  1
3x1  x2  kx3  2 2 x1  3 x2  x3  1 2 x1  3x 2  kx3  8
x1  7 x 2  6 x3  h 4 x1  7 x 2  kx3  h x1  hx 2  3 x3  2
2. Determine whether the following sets of equations are
a) Consistent b) inconsistent
i) x1  x 2  2 x3  1 ii) 2 x1  x 2  4 x3  4
 2 x1  x 2  x3  2 x1  2 x 2  3 x3  1
3x1  2 x 2  9 x3  4 3x1  3x3  6
iii) x1  2 x 2  x3  2
2 x1  x 2  7 x3  6
 x1  x 2  4 x3  4
x1  2 x 2  5 x3  6
We say that the system is consistent if it has a solution otherwise it is inconsistent.

1.4 Solutions of a general system of linear equation
Let’s consider the special case where all the above equations are degenerate, that is where
aij  0 . There are two cases:
18
U
i) If one of the constants bij  0 , that is the system has an equation of the form:
0 x1  0 x 2    0 x n  bi ,
with bi  0 then this equation, and hence the system, has no solution
ii) If every constant bi  0 , that is every equation in the system is of the form:
0 x1  0 x2    0 xn  0
Then this equation, and hence the system, has every n – tuple of real numbers as a solution.
Example 1.3
Determine the solution of the system of equation
0 x  0 y  0 z  0w  0
0 x  0 y  0 z  0w  0
0 x  0 y  0 z  0w  3
Solution
The system has no solution since one of the constants on the right is not zero.
Example
Determine the solution of the system of equation below
0 x  0 y  0 z  0w  0
0 x  0 y  0 z  0w  0
0 x  0 y  0 z  0w  0
Solution
The system has every 4-tuple   (k1 , k 2 , k 3 , k 4 ) as a solution since all the constants on the
right are zero.
In the usual case where the equations are not all degenerate, that is, where one of
the aij  0 , we reduce the system to a simpler system.
19
U
Example 1.4
Find the solution to the system of equation
2x  6 y  z  4
3x  2 y  z  1
5 x  9 y  2 z  12
Solution
We first observe that all the equations in the system are not degenerate, that is all the aij  0 .
Thus to find the solution, we follow the steps given below:
1
Step 1: Multiply the first equation by so that the system is replaced by
2
1
x  3y  z2
2
3x  2 y  z  1
5 x  9 y  2 z  12
Step 2: Multiply the new first equation by -3 and add it to the second equation so that the
system is replaced by
1
x  3y  z2
2
1
 11y    z  5
2
5 x  9 y  z  12
Step 3: Multiply the first equation by-5 and add it to the third equation so that the system is
replaced by the following which is in the desired form:
1
x  3y  z2
2
1
|  11y  z  5
2
1
 6y  z2
2
20
U
Step 4: Subtract the third equation from the second equation and solving for y as follows:
 5 y  7
 y  1.4
Step 5: Replace the value of y in the second equation and solving for z, we obtain:
1
 11y  z  5
2
7 1
 11   z  5
5 2
1  77 
z   5
2  5 
52
z 2
5
104
z
5
z  20.8
Step 6: Finally replace the values of y and z in equation one and solve for x, as follows:
1
x  3y  z2
2
1
 x  31.4   20.8  2
2
 x  4.2  10.4  2
 x  12.4  4.2
 x  8.2
Therefore the solution to the given system of linear equation is   (8.2, 1.4, 20.8)
This process of finding the solution to the system is known as the Gauss Elimination (Reduction)
Method.
The Gauss Elimination method can be summarized by using the following steps:
i) Interchange equations and the position of the unknowns so that aij  0
21
U
ii) Multiply the first equation by the appropriate non-zero constant so that a11  1
iii) For each i  1 , multiply the first equation by  a11 and add it to the i th equation so that the first
unknown is eliminated. Then the system (1) is replaced by an equivalent system of the form:
¶ x1  a12 x 2  a13 x3    a1n x n  b1
a 22 x 2  a 23 x3    a 2 n x n  b2
-------------------------------------------
a m 2 x 2  a m 3 x3    a mn x n  bm
This is exactly what you saw in the preceding activity.
Activity 1.2
Solve the following systems of linear equations using the Gauss elimination method.
i) 2 x  y  3z  5 ii) 2x  3 y  2z  5
3x  2 y  2 z  5 x  2 y  3z  2
5 x  3 y  z  16 4x  y  4z  1
iii) x  2 y  z  3w  3
2 x  4 y  4 z  3w  9
3x  6 y  z  8w  10
iv) x1  x2  2 x3  3 v) x1  2 x 2  4 x3  7
3x1  4 x 2  5 x3  9 x1  2 x3  2
x1  x 2  x3  6 2 x1  3 x 2  7 x3  9
22
U
Theorem 1.1
Every system of m equations in n unknowns can be reduced to an equivalent system of the following
form;
x1  c12 x 2  c13 x3    c1r x r    c1n x n  d1
x 2  c 23 x3    c 2 r x r    c 2 n x n  d 2
x3    c3 r x r    c3 n x n  d 3
x r    c rn x n  d r
0  d r 1
-----------------------------------------------------------
0  dm
 rm
 A system of linear equations in the above form, where the leading coefficient is not zero,
is said to be in echelon form. The solutions of a system of equations in the echelon form
can easily be described and found.
1.4.1 Consistent and inconsistent equations
1.4.1.1 Consistent Equations
If the numbers d r 1 ,, d m are all zero, then the system is said to be consistent and there does
exist a solution. There are two cases:
Case I
23
U
r  n , that is, there are as many non-zero equations as unknowns. Then we can successfully
solve uniquely for the unknowns x1 , x r 1 ,  , x n and so there exists a unique solution for the
system
Case II
r  n , that is, there are more unknowns than are non-zero equations. Then we can arbitrary
assign values to the unknowns x r 1 ,  , x n and then solve uniquely for the unknowns x r , , x1 .
Accordingly there exist an infinite number of solutions.
1.4.1.2 Inconsistent equations
If the numbers d r 1 ,, d m are not all zero that is there is an equation of the form
0 x1  0 x 2    0 xn  d k , where d k  0 ,then the system is said to be inconsistent, and there

are no solutions.
Example 1.5
Find the solution of each of the following system of linear equations:
a) x  2 y  3z  4 b) x  2 y  3z  w  4
0x  y  2z  5 0 x  y  2 z  3w  5
0x  0 y  z  1 0 x  0 y  0 z  0w  0
0x  0 y  0z  0
c) x  2 y  3 z  4
0 x  y  2 z  5
0x  0 y  0z  3
Solutions
24
U
a) Since there is no equation of the form 0  c , that is c  0 , the system is consistent.

Furthermore, since there are three unknowns and three non-zero equations, the system has a
unique solution.
Substituting z  1 in the second equation, we obtain
y  2(1)  5
y  3.
Now substituting z  1 and y  3 in the first equation, we obtain
x  2(3)  3(1)  4
x63 4
x 1
Thus, x  1 , y  3 and z  1 or in other words, the ordered triple (1, 3,1) is the unique
solution of the system.
b) Since the system has an equation of the form 0  c with c  0 , the system is inconsistent and
has no solution
c) The system is consistent, and since there are more unknowns than non –zero equations the
system has an infinite number of solutions. In fact we can arbitrary assign values to z and w
and solve for x and y.
To obtain a particular solution, substitute, say, z  1 and w  2 in the second equation to

obtain
y  2(1)  3(2)  5
y26 5
y  3
And substituting y  3 , z  1 and w  2 in the first equation, we obtain
x  2(3)  3(1)  2  4
x 63 2  4
25
U
x  11
Thus, x  11 , y  3 , z  1 and w  2 or, in other words the row vector (11,  3, 1, 2) is a

specific or particular solution to the system.
To obtain the general solution to the above system, substitute, say, z  a and w  b in the
second equation to obtain
y  2a  3b  5
y  5  2a  3b
Now substituting y  5  2a  3b , z  a and w  b into the first equation, we obtain
x  2(5  2a  3b)  3a  b  4
x  7a  5b  6
Hence, the general solution of the system is (7a  5b  6, 5  2a  3b, a, b) , where a and b
are real numbers.
Frequently, the general solution is left in terms of z and w (instead of a and b) as follows:
(7 z  5w  6, 5  2 z  3w, z , w)
Activity 1.3
Show that the following sets equations have infinitely many solutions and express the
solutions in terms of parameters.
(i) x1  3x 2  2 ii) x1  x 2  x3  5
2 x1  6 x 2  4 3 x 2  x3  x 4  8
3x 2  x 4  7
26
U
Theorem 1.2
A system of linear equations belongs to exactly one of the following cases:
(i) The system has no solution
(ii) The system has a unique solution
(iii) The system has an infinite number of solutions
Recall that, in the special case of two equations in two unknowns, the above three cases
correspond to the following cases described geometrically:
i) The two lines are parallel
ii) The two lines intersect in exactly one point
iii) The two lines are coincident.
1.5 Homogeneous systems of equations
1.5.1 Definition
A system of linear equations is called homogeneous if it is of the form:
a11 x1  a12 x 2    a1n x n  0
a 21 x1  a 22 x 2    a 2 n x n  0
------------------------------------
a m1 x1  a m 2 x2    a mn x n  0
That is if all the constants on the right are zero. By theorem 1.1, the above system is equivalent
to a system of the form:
x1  c12 x 2    c1r x r   c1n x n  0
x2    c2r xr   c2n xn  0
27
U
-------------------------------------------
x r   crn x n  0
00
---------------------------------------------
00
Clearly the above system is consistent since the constants are all zero; in fact, the system of
linear equations above, always has the zero solution (0, 0, 0, 0, , 0) which is called the trivial
solution. It has a non – trivial solution if
r  n . Hence if we originally begin with fewer equations than unknowns, then r  n and so the
system always has a non-trivial solution.
Theorem 1.3
A homogeneous system of linear equations with more unknowns than equations has a non- zero
solution.
Example 1.6
Determine the solution for the homogeneous system of equations below:
a)
x  2 y  3z  w  0
x  3 y  z  2w  0
2 x  y  3 z  5w  0
b)
x yz 0
2x  3 y  z  0
28
U
x  4 y  2z  0
c)
x yz 0
2x  4 y  z  0
3x  2 y  2 z  0
Solutions
a) The homogeneous system has a non – trivial solution since there are more unknowns than
equations.
b) Here the number of non-zero equations is equal to the number of unknowns. Thus, we try
to reduce the given equations in echelon form as follows:
x yz 0
2x  3 y  z  0
x  4 y  2z  0
Multiplying the first equation by -2 and adding it to the second equation we obtain;
x yz 0
 5 y  3z  0
x  4 y  2z  0
Multiplying the first equation by -1 and adding it to the third equation, we obtain
x yz 0
 5 y  3z  0
 5 y  3z  0
The last system of equations further reduces to the form
x yz 0
29
U
3
y z 0
5
00
From here we conclude that the system has a non – trivial solution since we finally have
only two non –zero equations in three unknowns. For example, choose z  10 , then
solving for y in the second equation. We find
3
y z 0
5
3
 y  (10)  0
5
y 6
Now replacing y  6 and z  10 in the first equation and solving for x in the first
equation, we obtain
x yz 0
x  6  10  0
x  4
So (4, 6,10) is a particular solution.
c) Here the number of non-zero equations is equal to the number of unknowns. Thus, we try
to reduce the given equations in echelon form as follows:
x yz 0
2x  4 y  z  0
30
U
3x  2 y  2 z  0
We multiply the first equation by -2 and add it to the second equation. We obtain
x yz 0
2y  z  0
3x  2 y  2 z  0
Next we multiply the first equation by -3 and add it to the third equation.
We obtain
x yz 0
2y  z  0
 y  5z  0
Finally we have
x yz 0
1
y z0
2
 y  5z  0
Since in echelon form there are three non-zero equations in three unknowns, the system
has a unique solution, the zero solution (0, 0, 0) .
1.6 Equivalent systems of equations
1.6.1 Definition
Two systems of equations involving the same variables are said to be equivalent if and only if
they have the same solution set.
31
U
Example 1.7
x1  2 x 2  4
3x1  x 2  2
4 x1  x 2  6
and
4 x1  x 2  6
3x1  x 2  2
x1  2 x 2  4
both involve the same three equations and, consequently, they have the same solution set.
If one equation of a system is multiplied through by a non zero real number, this will have no
effect on the solution set, and the new system will be equivalent to the original system.
Example 1.8
x1  x 2  x3  3
 2 x1  x2  4 x3  1
and
2 x1  2 x 2  2 x3  6
 2 x1  x2  4 x3  1
If a multiple of one equation is added to another equation, the system will be equivalent to the
original system. This follows since the n-tuple  x1 , , x n  will satisfy the two equations.
32
U
ai1 x1    ain x n  bi
a j1 x1    a jn x n  b j
If and only if it satisfies the equations
ai1 x1    ain xn  bi
(a j1   ai1 ) x1    (a jn   ain ) xn  b j   bi
1.7 Systems in Triangular and Echelon form
1.7.1 Definition
A system is said to be in strict triangular form if, in the k th equation, the coefficients of the first
k  1 variables are all zero and the coefficient of x k is non-zero (k  1, , n) . In other words, a
system of linear equations is said to be in triangular form if the number of equations is equal to
the number of unknowns and x k is the leading unknown of the k th equation. That is
a11 x1  a12 x 2    a1( n1) x n1  a1 n x n  b1
a 22 x 2    a 2( n 1) x n 1  a 2 n xn  b2

a ( n 1) ( n 1) x n 1  a ( n 1) n x n  bn 1
a n n x n  bn
Example 1.9
The system
3x1  2 x 2  x3  1
33
U
x 2  x3  2
2 x3  4
is in strict triangular form, since in the second equation the coefficients are 0, 1, -1, respectively,
and in the third equation the coefficients are 0, 0, 2, respectively. Because of the strict triangular
form, the system is easy to solve. It follows from the third equation that x3  2 . Using this value
in the second equation, we obtain
x2  2  2
x2  4
Using x 2  4 , x3  2 in the first equation, we end up with
3x1  2  4  2  1
x1  3
Thus, the solution of the system is (-3, 4, 2).
34
U
Unit summary
In this unit we have looked at different types of systems of linear equations. We have exposed you to
various ways of solving systems of linear equations and above all how to determine the solutions of systems
of linear equations.
35
U
Unit 2
Matrix Algebra
2.0 Introduction
Welcome to unit 2. In this unit you are going to learn about different types of matrices. You will
also be taken through matrix operations which you were introduced to in MAT 110. You will
also use matrices to solve systems of linear equations using the Gauss Jordan Elimination
method and elementary row operation. Furthermore you will also be exposed to various proofs of
theorems on matrix algebra. I ‘ m sure you are ready to embark on this ‘journey’
36
U
 Define different types of matrices
 Add, subtract and multiply matrices
Outcomes  Find solutions of systems of equations using matrices
2.1 Definition
A matrix is a rectangular array of numbers called the entries or elements.
Let K be an arbitrary field. A rectangular array of the form
 a11 a12  a1n 

 
 a 21 a 22  a2n 
   
 
a am2  a mn 
 m1
where the aij are scalars (numbers) in , is called a matrix over .
A matrix with m rows and n columns is called an m by n matrix, or m  n matrix; the pair of
numbers m, n  is called its size or shape.
In the above matrix, the m horizontal n- tuples a11 , a12 , a13 ,  , a1n  , a 21 , a 22 , a 23 , , a 2 n  ,  ,
a m1 , a m 2 , a m3 , , a mn  are the rows of the matrix, and the
 a11   a12   a1n 

     
 a 21   a 22   a2n 
n-vertical m –tuples  a31  ,  a32  ,  ,  a  are its columns.
     3n 
     
     
 a m1   a m 2   a mn 
2.2 Matrix notation
37
U
Matrices will usually be denoted by capital letters, say A, B … and the elements of the field K by
lower case letters, say a, b …
We also use double subscript notation to refer to the entries of a matrix A.
For instance the entry of matrix A in rows i and columns j is denoted by aij .
Thus, if
0 5 4 
A    , then
 3 9  1
a 23  1 and a12  5
The notation Aij is sometimes used interchangeably with aij . We can therefore completely
   
denote a matrix A by aij or a ij mn
if it is important to specify the size of A.
 a11 a12  a1n 

 
a a 22  a2n 
Thus, in the matrix  21 , i  1, 2, 3, , m and j  1, 2, 3,  , n .
   
 
a am2  a mn 
 m1
The element aij , called the i j- entry or i j component, appears in the i th row and j th column.
Theorem 2.1 Equality of matrices
   
If A  aij is an m  n matrix and B  bij is a p  q matrix, then A  B if and only if (iff)
m  p and n  q and aij  bij for all i and j.
In other words two matrices are equal if they have the same size and if there are corresponding
entries are equal.
Example 2.1
Let A and B be two matrices with the same size, that is the same number of rows and of
columns, say,
38
U
 a11 a12 a13  a1n   b11 b12 b13  b1n 

   
a a 22 a 23  a2n  b b22 b23  b2 n 
A   21  and B   21 .
         
   
a am2 am3  a mn  b bm 2 bm 3  bmn 
 m1  m1
Then A  B if and only iff
a11  b11 ; a12  b12 ; a13  b13 ; a1n  b1n and so on.
2.3 Special type of matrices
2.3.1 Zero matrix
 
The matrix a ij  0 such that aij  0 (i  1, 2, 3,  , m) , ( j  1, 2, 3, , n) is called
a zero matrix.
Example 2.2
 0 0 0
 
A   0 0 0
 0 0 0
 
2.3.2 Square matrix
A matrix with the same number of rows as columns is called a square matrix.
A square matrix with n rows and n columns is said to be of order n, and is called a
an n- square matrix.
Example 2.3
 1 0 0
 
B   0 1 0
0 0 1
 
39
U
2.3.3 Triangular matrix
2.3.3.1 Upper triangular matrix
An upper triangular matrix or simply a triangular matrix is a square matrix whose entries below
 
the main diagonal are all zeros; that is A  aij where aij  0 whenever i  j .
Example 2.4
 a11 a12 a13  a1n   a11 a12 a13  a1n 

   
 0 a 22 a 23  a 2 n   a 22 a 23  a 2 n 
 0 0 a33  a3n  or  a33  a3n 
   
         
   
 0 0 0 0 a nn   a nn 
2.3.3.2 Lower triangular matrix
A lower triangular matrix is a square matrix whose entries above the main diagonal are all zero;
 
that is A  aij where aij  0 whenever i  j .
Example 2.5
 a11 0 0 0 0   a11 
   
 a 21 a 22 0 0 0   a 21 a 22 
a a 32 a33 0 0  or  a a 32 a33 
 31   31 
         
   
 a n1 an2 a n3  a nn   a n1 an2 a n3  a nn 
2.3.4 Diagonal matrix
A diagonal matrix is square matrix whose non- diagonal entries are all zero.
40
U
Example 2.6
 a1 0 0  0   a1 
   
0 a 22 0  0   a2 
0 0 a33  0 or 
 a3 
 1   
0 0 0 a4 0   a4 
   
0 0 0  a nn   an 
2.3.5 The unit or identity matrix
The n- square matrix with 1’s on the main diagonal and 0’s elsewhere is called the unit or
identity matrix. It is denoted by I n or simply I.
Example 2.7
 1 0 0
 
I3   0 1 0
0 0 1
 
 
I  aij where aii  1 , aij  0 , (i  j ) .
The identity matrix I n has the property
AI n  I n A  A for all n  n matrices.
2.3.6 Inverse matrix
 
If A  aij is an n  n matrix then an n  n matrix B such that AB  BA  I n is called an inverse
of the matrix A and is denoted by A 1 .
Theorem 2.2
The inverse of a matrix is unique.
Proof:
Let Q be an inverse of P, then
41
U
PQ  QP  I n and Q  QI n  Q(PR )
 (QP) R
 InR
R
QR
Thus, the inverse of P is unique.
2.3.7 Invertible matrix
A matrix which has an inverse is called an invertible or non- singular matrix while a matrix
which has no inverse is called a singular matrix.
Theorem 2.3
(i) If A is an invertible matrix, then A 1 is invertible and ( A 1 ) 1  A .
(ii) If A is an invertible matrix and c is a non zero scalar, then cA is an invertible matrix and
1 1
(cA) 1  A
c
(iii)If P and Q are invertible matrices of the same size or n  n matrices, then PQ is an
invertible and ( PQ) 1  Q 1 P 1 .
(iv) If A is an invertible matrix, then ( A t ) 1  ( A 1 ) t
(v) If A is an invertible matrix, then A n is invertible for all non- negative integers n and
( A n ) 1  ( A 1 ) n
Proof of theorem (iii)
If P and Q are invertible then
PP 1  I n  P 1 P and QQ 1  I n  Q 1Q
42
U
Now 
( PQ ) (Q 1 P 1 )  ( PQ )Q 1 ( P 1 ) 

 P(QQ 1 ) P 1 
 ( PI n )  P 1
 PP 1
 In
Q 1 P 1  ( PQ) 1 or simply ( PQ) 1  Q 1 P 1 .
2.3.8 Orthogonal matrix
An n  n matrix Q is said to be orthogonal matrix if the column vectors form an orthogonal set in
n .
Theorem 2.4
An n  n matrix Q is orthogonal if and if Q T Q  I
Proof
It follows from the definition that an n  n matrix Q is orthogonal if and only if its column
vectors satisfy
T
qi q j   ij
T
However, qi q j is the (i , j ) entry of the matrix Q T Q. Thus Q is orthogonal if and only if
QT Q  I .
It follows from the theorem that if Q is an orthogonal matrix, then Q is invertible and Q 1  Q T
Example 2.8
Consider the matrix Q below
43
U
 cos   sin  
Q   
 sin  cos  
 cos  sin  
Now Q T    and
  sin  cos  
cos   sin 
det Q 
sin  cos 
det Q  cos 2   sin 2   1
 Q 1   
  sin  cos  
You can see that
 cos   sin  
Q 1  Q T    . Therefore the matrix Q is orthogonal.
 sin  cos  
2.3.9 Idempotent matrix
An n  n matrix A is said to be idempotent if and only if A 2  A .
2.4.0 Nilpotent matrix
An n  n matrix A is said to be nilpotent of index p if and only if A p  Z but
A p 1  Z for some integer p  1 . Any strictly triangular matrix is nilpotent.
An n  n matrix A is called
a) symmetric matrix if A t  A . That is if and only if Aij  A ji for all i and j
b) A skew symmetric matrix if A t   A
c) An orthogonal matrix if A t A  AAt  I n
44
U
Activity 2.1
1. Which of the following matrices are symmetric, and which are skew- symmetrical.
0 1  2  2 3 1 
 1 2  0 2 1 0     
  ;   ;   ;   1 0 3  ;  3 0  1
 2 3    2 0   0 1  2  3 0   1  1 2 
   
2. Let A be any square matrix. Show that A  AT is a symmetric matrix
3. Show that the products AT A and AAT are symmetric matrices.
4. Show that the following matrices are orthogonal.
 2 1 
 0 
 1 2  6 3  3 4
   
5 5 ;  1 1 1 

 ;  5 5
 2 1 
 2 6 3  4 3
   
 5 5  1 1 1   5 5
 
 2 6 3
5. Show that a product of two orthogonal matrices of the same size is an orthogonal matrix.
6. Find which of the matrices are invertible and which are singular. Find the inverses of
those which are singular
2 3 2 3 
1 2 1  1 1 2  0 1 1  
1 2        1 0 2 1 
  ;  0 1 2  ;   1 2  1 ;  1 0 1  ;  ;
1 3   0 0 1   1  3 1   1 1 0    1 1 4  2
      3 
 0 0 4 
 1 1 1 1 
 
 2 1 0 3 
 3 0 2 5 
 
 1  1  1  3
 
45
U
 a 0 0
 
7. Find the inverse of the diagonal matrix  0 b 0  , where a, b, and c are non-zero.
0 0 c
 
8. Let A, B and C be invertible matrices of the same size. Show that the inverse of the
product ABC is C 1 B 1 A 1
9. Show that, for any invertible matrix A, ( A 1 ) T AT  I and AT ( A 1 ) T  I
2.4 Matrix addition
2.4.1Definition
 
If A  aij is an m  n matrix and B  bij  
is a p  q matrix then, A  B is defined if and only if m  p , and n  q and then
A  B  (aij  bij ) is the m  n matrix obtained by adding the corresponding elements in A and
B.
Example 2.9
columns, say, m  n matrices.
 a11 a12  a1n   b11 b12  b1n 

   
a a 22  a2n   b21 b22  b2 n 
A   21 and B 
       
   
a am2  a mn  b bm 2  bmn 
 m1  m1
The sum of A and B, written A+B, is the matrix obtained by adding corresponding entries
 a11 a12  a1n   b11 b22  1n

   
a a 22  a 2 n   b21 b22  2n 
A  B   21 
       
   
a am2  a mn   bm1 bm 2  bmn 
 m1
46
U
 a11  b11 a12  b12  a1n  b1n 

 
 a  b21 a 22  b22  a 2 n  b2 n 
A  B   21 
   
 
a  b a m 2  b1m  a mn  bmn 
 m1 m1
2.5 Scalar Multiplication
   
If A  aij is an m  n matrix and   K , then A   aij is the m  n obtained by multiplying
each element of A by  . A is called the scalar multiple of A by  .
Theorem 2.5
Let V be a set of all m  n matrices over a field K. Then for any matrices A, B, C  V and any
scalars h1 , h2  K ,
i) ( A  B )  C  A  ( B  C )  Associative
ii) A  0  A
iii) A  ( A)  0
iv) A  B  B  A  commutative
v) h1 ( A  B)  h1 A  h1 B
vi) (h1  h2 ) A  h1 A  h2 A
vii) (h1 h2 ) A  h1 (h2 A)
viii) 1 A  A
ix) 0 A  0
x) A  A  2 A
47
U
2.6 Matrix Multiplication
2.6.1 Definition:
If A  aik  is an m  n matrix and B  bkj  is an n  r matrix, then the product AB is the

m  r matrix C  cij .  
The i, j  entry of the product is computed as follows; multiplying elements in the i th row of A
with the corresponding elements in the j th column of B and summing up.
That is
C ij  ai1b j1  a i 2 b j 2    ain b jn .
n
Cij can be represented in short form as C ij   aik bkj , for each i  1, , m and each j  1, , n .
k 1
The number of columns of A must be the same as the number of rows of B.
Theorem 2.6
Properties of matrix multiplication

Let A, B, and C be matrices (whose sizes are such that the indicated operations can be
performed) and let k be a scalar. Then
(i) ( AB)C  A( BC )  Associative law
(ii) A( B  C )  AB  AC  distributive law
(iii) ( B  C ) A  BA  CA
(iv) k ( AB )  kA( B)  A(kB) , where k is a scalar
(v) 0 A  0 , where 0 is the zero matrix.
(vi) I m A  A  AI n if A is m  n  multiplicative identity
48
U
1. AB  0 does not necessarily mean that A  0 or B  0 .
2. AB  AC does not imply that B  C .
That is AB  AC  0
 A( B  C )  0 can be true with A  0 and B  C  0 .
2.7 The transpose of a matrix
2.7.1 Definition
 
If A  aij is an m  n matrix, the transpose of A is the n  m matrix denoted by A t
obtained by interchanging the rows and columns of A.
 
That is A t  Bij where
Bij  a ji ; ( (i  1, 2, 3, , n; j  1, 2, 3, , m)
Theorem 2.7
Properties of the transpose
If A and B are m  n and p  q matrices respectively, and k be a scalar.
Then:
i) ( A  B ) t  At  B t where both sides are defined if and only if m  p and n  q
ii) ( At ) t  A
iii) ( KA) t  K ( A t ) , for k a scalar
iv) ( AB) t  B t A t both sides are defined if and only if n  q
v) ( A r ) t  ( A t ) r
49
U
If A is an n  n matrix then A t is invertible if and only if A is invertible.
Further ( A t ) 1  ( A 1 ) t
Proof:
If A is invertible then AA 1  I n  A 1 A
Transposing gives
( A 1 ) t A t  I n
 A t ( A 1 ) t  I n
And so A t is invertible.
Conversely, if A t is invertible, then A t ( A t ) 1  I n  ( A t ) 1 A t
Transposing gives
A 1 A  I n  AA 1
Clearly from the above also implies by the previous theorem, that is ( A t ) 1  ( A 1 ) t .
Activity 2.2
1. Given that
2
1 1 2 4 0  3  
A    ; B    ; C    1 ;
 0 3 4  1  2 3  3
 
Find
i) A + B ii) 3A – 4B iii) AT iv) C T AT v) A 3
2. Let D x be the m  square scalar matrix with diagonal elements k. Show that:
i) For any m  n matrix A, D x A  kA
50
U
ii) Show that the sum, product and scalar multiple of:
a) Upper triangular matrices is upper triangular
b) Scalar matrices are scalar.
2.8 Matrices and echelon form
Recall that a system of m linear equations in n variables over a field F, that is
a11 x1  a12 x 2    a1n x n  b1
a 21 x1  a 22 x 2    a 2 n x n  b2

a m1 x1  a m 2 x 2    a mn x n  bm can be written in matrix form as:
 a11 a12  a1n  x1   b1 

    
 a 21 a 22  a 2 n  x 2   b2 
a a32  a 3n  x3    b3 
 31    
          
    
 a m1 am2  a mn  x n   bm 
However, the above matrix can be written in short or abbreviated form as
AX  B , where A is the matrix of coefficients, X is the matrix of variables and B is the matrix
of constants.
Most often we will dispense (do away) with the matrix of variables X and will simply write the
argument matrix of the system as
a11 a12  a1n b1

a 21 a 22  a 2 n b2
A/ B  a31 a32  a3n b3
    
a m1 am2  a mn bm
51
U
Example 2.9.1
Write the argument matrix of the systems given
a) 3x  y  2 z  9
x  5y  z  0
 2 x  10 y  3z  4
b) 3w  2 x  y  2 z  12
2x  y  2z  1
w x  3
Answers
a) The argument matrix is
3 1 2 9
A/ B  1 5 1 0
 2 10 3 4
b) The argument matrix is
3 2  1  2 12
A / B  0 2 1  2 1
1 1 0 0 3
2.8.1 The pivot
52
U
If a row of a matrix is non-zero, we call the first non-zero entry of this row a pivot or leading
entry for this row. For instance, in example (a) above, 3, 1 and -2 are leading entries for the first
row, second row and third row respectively.
2.8.2 The row echelon matrix
A matrix M  M mn (F ) is a row echelon matrix if;
i) All the zero rows of M, if any, are at the bottom of M
ii) For any two consecutive rows Rt and Rt 1 , either Rt 1 is all zeros or the pivot of Rt 1 is
immediately to the right of the pivot of Rt . In other words in each non-zero row, the first
non-zero entry (pivot) is in a column to the left of any leading entries below it.
The variables accompanying the pivots are called leading variables. Those variables which
are not leading variables are called free parameters.
Example 2.9.2
State whether the given matrices are in echelon form or not.
1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
       
0 0 1 2 0 0 2 2 0 0 0 0 0 0 0 1
A  , B  , C   , D
0 0 1 2 0 0 0 3 0 0 0 1 0 0 0 0
       
0 0 0 0      0 
 0 0 0 0 0 0 0 0 0 0 0
Answers
Matrices A and C are not in echelon form because they dot not satisfy the two conditions above.
For instance matrix A does not meet condition (ii) while matrix C does not satisfy condition (i).
Matrices B and D on the other hand, are in row echelon form because they both meet the two
conditions.
2.8.3 Reduced row Echelon form
53
U
A matrix is in reduced row echelon form if it satisfies the following properties:
i) Any rows consisting entirely of zeros are at the bottom.
ii) The leading entry in each non-zero row is a 1 (called a leading 1).
iii) Each column containing a leading 1 has zeros everywhere else.
The matrix below is an example of a reduced row echelon matrix
1 3 0 2 0 10
 
0 1 0 4 5 1 0
0 0 0 1 3  2 0
 
0 0 0 0 0 0 1
0 0 0 0 0 0 0 

2.9 The Gauss - Jordan Elimination method.

The Gauss – Jordan elimination is a method employed in further reducing the matrix in echelon
form. It follows the following steps:
i) Write the argument matrix of the system of linear equations
ii) Use elementary row operations to reduce the argument matrix to reduced row echelon
form
iii) If the resulting system is consistent, solve for the leading variables in terms of any
remaining free variables.
Example 2.9.3
Solve the system
x 2  x3  5
2 x1  x 2  5 x3  1
x1  x 2  3 x3  3
Solution
We begin by writing the argument matrix as follows:
54
U
0 1 15
2 1 51
1 1 33
Now interchanging row 1 ( R1 ) and row 3 ( R3 ) we obtain
1 1 33
2 1 51
0 1 15
Replacing R2 with R2  2R1 , we obtain
1 1 1 3
0 1 1  5
0 1 1 5
Replacing R3 with R3  R2 , we obtain
1 1 1 3
0 1 1  5 which is a matrix in echelon form.
0 0 0 0
The system has now been reduced to
x1  x 2  x3  3
 x 2  x 3  5
Since we have more unknowns (variables) than equations, the system has an infinite number of
solutions. That is any value of x1 , x 2 and x3 satisfy the equation represented by the third row .
Example
Solve the system of linear equations
1 1 1 1  x    3 
    
0 2 1 0  y    1 
0 
0 1  1 z   9 
    
0 0 0 2  w    6 

55
U
Solution
We observe that the matrix of coefficients is already in row-echelon form. Clearly every variable
is a leading variable, and by back substitution
2w  6
 w  3
Then
zw4
z  43
 z 1
Also
2 y  z  1
2 y  1  1
y  1
And
x  y  z  w  3
x  1  1  3  3
x0
x  0 
   
 y    1
Therefore the unique solution is     
z 1
   
 w    3
   
Given a matrix A  M mn (F ) , by following Gauss Jordan reduction, we can find a matrix
P  Glm ( f ) such that PA  B is in row-echelon form
56
U
Activity 2.3
1. State which of the following matrices is in Gauss-reduced form, and for those that are not,
explain why not.
0 1 2
2 1 3 2  1 3 1 2  
    0 0 0
a)  0 0 1  2 b) 0 0 0 0 c) 0
0 0 0 0   0 1 1 3 0 0
     
0 0 0 

2. Find a Gauss- reduced form for each of the following matrices.
1 2 3 4 7
 1 1  8  14   
   1  2 1 4 0 5 7 6 8
a) 3  4  3 0  b)   c) 0
 2  1  7  10   2  3 1 2 5 2 4 7
   
0 0 2 1 0 

 4 1 2 6 
 
d)   1 5  1  3
3 4 1 3 

2.8 Elementary matrices
2.8.1 Definition
An n  n matrix is called an elementary matrix if it can be obtained by applying an elementary

row operation on the identity matrix ( I n ).
An Elementary row operation or simply row operation on a matrix A is any one of the
following three types of operations:
57
U
(i) Interchange of two rows of A. That is ri  r j
(ii) Replacement of a row ri of A by ri for some number   0 . That is ri  ri
(iii)Replacement of a row ri of A by the sum ri  r j of that row and a multiple of another
row r j of A. That is ri  ri  r j
Example 2.9.5
i) ri  ri
1 0 0 0 0 1
   
 0 1 0  r1  r3  0 1 0 
0 0 1 1 0 0
   
ii) ri  ri
1 0 0 1 0 0 
   
 0 1 0  r3  2r3  0 1 0 
0 0 1  0 0  2
   
iii) ri  ri  r j
1 0 0 1 0 0 
   
 0 1 0  r2  2r3  r2  0 1 2 
0 0 1  0 0  2
   
Theorem 2.7
If an m  n matrix B can be obtained from an m  n matrix A by applying an elementary row

operation, then B is equal to the product of the corresponding m  n elementary matrix A. If e is
the elementary row operation, then
58
U
B  eA  EA
Proof:
We shall prove the theorem by considering each separate type.
i) If ri  ri , then
  11  12   1n 
 
   i 2   in 
eA   ii
   
 
  m2   mn 
 m1
1    11  12   1n 
  
     ii  i 2   in 
 EA  eA  B
 1      
  
 1    m2   mn 
   m1
  12  12   1n 
 
  11  12   1n 
M i ( ) A  
   
 
    
 m1 m2 mn 
ii) If ri  r j , then
  11   1n 
 
i   j1  in 
H ij A      B
 
j   j1   1n 
 
  mi   mn 
iii) If ri  ri   r j , then
  11   1n 
 
  i1   j1   in   in 
Aij ( ) A   B
 
 
    mn 
 mi
59
U
Example 2.9.6
1 0 1 1 0 1
  r2  2 r1  r2  
Let A   2  1 0     0  1 2   B
1 0 0 1 0 0
   
Next we apply the same elementary row operation on the identity matrix I 3 . We obtain
1 0 0  1 0 0
  r2 2 r  r2  
 0 1 0     2 1 0   E
0 0 1  0 0 1
   
 1 0 01 0 1
  
Now note that EA    2 1 0   2  1 0 
 0 0 1  1 0 0
  
1 0 1 
 
 0 1  2  B
1 0 0 

Theorem 2.8
Every elementary matrix is invertible and the inverse of each elementary is elementary.
Proof:
We shall consider each separate type by verifying that
1
(i) ( M i ( ) 1  M i  
 
(ii) ( H ij ) 1  H ij
(iii) ( Aij ( ) 1  Aij ( )
Proving using (i)
60
U
1 
 
 1 
1  1 
M i    row i 
   
  
 1 

1 1  1 
    
       
1  1    
M i    M i ( )      1
     
       
 

 1   1   1 
1
= M i ( ) M i   .
 
Proving (iii)
( H ij ) 1 H ij  H ij H ij ri  r j
1  1  1 
    
 1 1   1    
 0   0    1 
    
 1   1     
 1   1   1 

= H ij ( H ij ) 1
Proof of (iii)
( Aij ( )) 1  Aij ( )  Aij ( ) Aij ( )
61
U
1  1 
   
 1    1  
     
    ri  r j  ri
 1   1 
     
  
 1   1 
 
1 
 
 1 0 
  
=  
 1 
  

 1 

= Aij ( )  Aij ( )
Example 2.9.7
1 0 0 1 0 0
  r3 2 r1  r3  
 0 1 0     0 1 0 
0 0 1  2 0 1
   
 1 0 0
1
 
E  A31 (2)   0 1 0 
 2 0 1
 
 1 0 0  1 0 0 
1
  
E E   0 1 0  0 1 0 
  2 0 1  2 0 1 
  
1 0 0
 
=  0 1 0   EE 1
0 0 1
 
2.9 Row equivalent matrices
62
U
2.9.1 Definition
A matrix B is row equivalent to matrix A if there exists a finite sequence E1 , E 2 , , E k of
elementary matrices such that B  E k E k 1  E1 A .
In other words, B is row equivalent to A if B can be obtained from A by a finite number of row
operations.
Theorem 2.9
If an m  n matrix A is row equivalent to an m  n matrix B, then there exists (  ) a finite number

of elementary matrices 1 ,  2 ,  3 , , E k A
Proof:
B is row equivalent to A  B can be obtained from A by finite, say k sequences of elementary

row operations. We shall prove the theorem by induction.
If k  1, then B  1 A
If K  1, we assume that if a matrix B' can be obtained from A by k  1 elementary row

operations, then  elementary matrices  2 ,  3 , ,  k such that B'   2  3 , ,  k A
Now B is obtained from B' by applying one further elementary row operation that is B  1 B '
for some elementary matrix 1 thus B  1 ( 2  3   k A)  1 2  E k A) as required.
Cor 2.1
If R is the reduced echelon matrix of A, then there exists a finite number of elementary matrices
 '1 ,  k such that :
A  1  2   k R
Proof:
By theorem 2.9, there exist elementary matrices  '1 ,  ' 2 , ,  ' k such that
63
U
R   '1 2   ' k A .
' ' '

But 1 ,  2 , ,  k is non-singular and its inverse is also an elementary matrix. Thus
' 1 ' 1
A   k  '1 k 1  1 R, as required
A  1  2   k R .
Cor 2.2
The following matrices are equivalent.
i) The n  n matrix A is invertible
ii) The homogeneous system AX  0 of n linear equations in n variable has no trivial

solution
iii) A is a product of elementary matrices.\
Proof : (i )  (ii )
If A is invertible, the system of n equations in n variables given by AX  0 has non-trivial

solution since X  ( A 1 A) X  A 1 ( AX )
 A 1 (0)
0
(ii )  (iii ) the reduced echelon matrix is I n and thus A  1 2   k I n
That is
A  1  2   k
This implies that A is a product of elementary matrices.
(iii )  (i) If A  1 2  E k , where each  i is an elementary matrix then since elementary
matrices are invertible, then
64
U
1 1
A 1  E k   1 k 1  1
( AB) 1  B 1 A 1
Thus A is invertible.
To determine whether a matrix is invertible or not we note that if the reduced echelon matrix of
A is R and
i) If R  I n , the matrix A is invertible
ii) If R contains a row of zeros, then A is not invertible.
Corollary 2.2 is very important because it gives us an explicit method for finding the inverse of
an invertible matrix. If A is non-singular, then the reduced echelon matrix of A is the identity
matrix.
If 1 ,  2 , ,  k are the elementary matrices which correspond to the elementary row operations
which must be performed on A to give I n . That is I n   k  k 1  1 A , then
1 1
A  1 ,  2   k I n and A 1   k  k 11 I n . That is A 1 is determined by applying the same
elementary row operations to I n as were used to obtain I n from A.
Example 2.9.8
2 4 6 
 
Let A   4 5 6  , calculate A 1
 3 1  2
 
Solution
We first put I next to A in an augmented matrix form as follows
65
U
2 4 6 1 0 0
4 5 6 0 1 0
3 1 2 0 0 0
And perform the elementary row operations to obtain a reduced row echelon matrix
2 4 6 1 0 0
4 5 6 0 1 0
3 1 2 0 0 0
8 7
 1
1 0 0 3 3
13 11
 0 1 0  2
3 3
0 0 1 11 5
 1
6 3
 8 7 
  1
 3 3 
 13 11
1
Since A has now been reduced A   2
 3 3 
 11 5 
  1
 6 3 
  16 14  6 
11

A   26  22 12 
6 
  11 10  6 
Example 2.9.9
 1  3 4
 
Is the matrix A given by A   2  5 7  invertible?
0 1 1
 
Solution
1 3 4 1 0 0
2  5 7 0 1 0 r
1  2 r1  r3

0 1 1 0 0 1
66
U
1 3 4 1 0 0 r
1 3 r2  r1

0 1  1  2 1 0 r
3  r2  r3
 
0 1 1 0 0 1
1 0 1 5 3 0
0 1 1  2 1 0
0 0 0 2 1 1
This is as far as we can go. The matrix A cannot be reduced to the identity matrix and we
conclude that A is not invertible.
If in the row reduction of A we end up with a row of zeros, then A is not invertible.
2.9.1 Elementary Column Operations and Equivalence of matrices

The definition of elementary row operations and row equivalent matrices may be adapted to
define elementary column operations on a matrix and column equivalent matrices.
We know that if R is the reduced echelon matrix of an m  n matrix A, then there exist an
invertible matrix P such that
PA  R
By theorem 2.9 R  1 2   k A where E k ,  k 1 , 1 are the elementary matrices which

correspond to the sequence of elementary row operations which are carried out successively in
order to obtain r from matrix A.
Thus P  1 2   k
67
U
P  1 ( 2 (  k 1 ( k I m )) )
And the matrix P is obtained by applying to I m the same elementary row operations as were used
to obtain R from A.
Example 2.1.0
 2 1 0 1 
 
A   1 2 1  1
 2 9 4  5
 
2  1 0 1 1 0 0 r
1  2 r2  r1

1 2 1 1 0 1 0
2 9 4  5 0 0 1 r
3  2 r2  r3
 
1
r1   r1
5
0  5  2 3 1  2 0   
1 2 1 1 0 1 0
r3  r1  r3
0 5 2  3 0  2 1   
2 3 1 2
0 1   0
5 5 5 5
0 2 1 1 0 1 0 r
1  r2

0 0 0 0 1 4 1
1 2 1 1 0 1 0
2 3 1 2
0 1   0 r
1  2 r1  r1

5 5 5 5
0 0 0 0 1 4 1
1 1 2 1
1 0 0
5 5 5 5
2 3 1 2
0 1   0
5 5 5 5
0 0 0 0 1 4 1
68
U
 2 1   1 1 
 0 1 0 
 5 5 2 1 0 1   5 5 
 1 2   2 3
0   1 2 1  1   0 1   ; PA= R
 5 5   5 5
 
 1  4 1   2 9 4  5  0 0 0 0 
   
   
Theorem 2.9 and the fact that every m  n matrix is row equivalent to an m  n reduced echelon
matrix have their equivalent statements in terms of elementary column operations. This leads us
to the following theorem;
Theorem 2.9.1
If an m  n matrix A is column equivalent to an m  n matrix B, then there exist a finite number

of elementary matrices 1 ,  2 , ,  l such that
B  A1 2   l
B  1  2   k A
From theorem 2.9.1 we see that there exists an invertible m  n matrix Q such that
B = AQ
Let us now apply this to the reduced echelon matrix R of an m  n matrix A.
Suppose that the leading 1s appear in columns j1 j 2 , j r then by subtracting suitable multiples
of these columns from succeeding columns and then columns interchanges
ci  c j , c 2  cj 2 , , cr  cjr , we have that R is column equivalent to the matrix
 1r 0r  ( n  r ) 
N    where 1r is the r  r identity matrix and 0 p  q is the zero
 0 ( mn ) r 0 ( mr )( n r ) 
p  q matrix.
69
U
Theorem 2.9.2
If A is an m  n then there exist an invertible m  m matrix P and an invertible m  n matrix Q

such that PAQ  N where N is the matrix given above.
The matrices P and Q are calculated as explained above and an illustration is given in the
example below
Example 2.1.1
 2 1 0 1 
 
If A   1 2 1  1  from the example above if
 2 9 4  5
 
 2 1 
  0
 5 5 
 1 2
P  0  then
 5 5 
 1  4 1
 
 
 1 1 
1 0 
 5 5 
2 3
PA   0 1  
 5 5
0 0 0 0 
 
 
Now we apply the elementary column operations on the argument matrix.
 1 1 
1 0 
 5 5 
0 1 2 3
 
 5 5
0 0 0 0 
 
 
70
U
PAQ  3  3 3  4  3  4
1 1 1
c3   c1  c3
1 0 1 0 0 0 5
  

5 5  
2 3 0 1 0 0
0 1  0
5 5 0 1 0 1
c 4   c1  c4
0 0 0 0   5
  

0 0 0 1 

2
 1 1 c3   c2  c3
5
1 0 0 0 1 0      

 5 5
2 3
0 1  0 1 0 0 
3
5 5 0 0 1 0  5
c 4  c 2  c4
0 0 0 0     
0 0 0 1 
 1 1
1 0   
1 0 0 0 5 5
2 3 
0 1 0 0 0 1 
 5 5 
0 0 0 0 0 0 1 0 
 
0 0 0 1 
 1 1
1 0   
 5 5
2 3 
 Q  0 1 
 5 5 
0 0 1 0 
 
0 0 0 1 
 1 0 0 0
 
 PAQ  N   0 1 0 0 
 0 0 0 0
 
2.9.2 Equivalent Matrices
2.9.2.1 Definition
Two m  n matrices A and B are equivalent if  invertible matrices P and Q such that PAQ  B
If PAQ  B, then P 1 BQ 1  A
71
U
PAQ  B and P BQ   C then ( P P) A(QQ )  C
What has been proved above is that every m  n A is equivalent to a matrix with simple form N,
where r denotes the number of non zero rows in the reduced echelon form of A
N may be regarded as a canonical form of A under this equivalence relation. That is a

representative of the equivalence class which contains A under this equivalence relation. Later it
will be shown that r is the rank of the matrix A and N is called the normal form of the matrix A.
Example 2.1.3
Find invertible matrices P and Q such that PAQ is normal form for the matrix
 2 1 1
 
 1 1 1
A
0 1 3
 
 1 2 4 

Solution
Since A is 4  3 matrix we set an argument matrix using the associated 4  4 identity matrix and
reduce it to reduced row echelon form.
2 1 1 1 0 0 0  r
1  r4

 
1 1 1 0 1 0 0
0 1 3 0 0 1 0
 
1  2 4  0 0 0 1 
1 2 4 0 0 0 1 r2  r1  r2
 
1 1 1 0 1 0 0
0 1 3 0 0 1 0 r
4   r1  r4

2 1 1 1 0 0 0
1 2 4 0 0 0 1 r
2  r3

0 1  3 0 1 0 1 r 3  r2  r3
 
r4 3r2  r4 2
0 1 3 0 0 1 0  
0 3 9 1 0 0 2
72
U
1 0 20 0 2 1 0 0 2 1 
 
0 1 30 0 1 0 0 0 1 0 
P
0 0 00 1 1 1 0 1 1 1 
 
0 0 0 1 0 3 2 1 0  3  2 

0 0 2
1  2 1 1  1 0 2
    
0 0 1 0 1 1 1  0 1 3
Thus PA   
0 1 1 1  0 1 3  0 0 0
    
1 0  3  2   1  2  4   0 0 0 

We then set an argument matrix using the associated 3 3 identity matrix and then apply the
elementary column operations.
1 0 2
1 0 0 C3  2 c1  c3
 
0 1 3 c3  3c 2  c3
0 1 0    
0 0 0
0 0 1
0 0 0
1 0 0
1 0 2 1 0  2
0 1 0  
0 1 3  Q   0 1  3
0 0 0 0 0 1 
0 0 1  
0 0 0
Activity 2.4
1. Use elementary row operations to reduce the given matrix to:
a) Row echelon form
b) Reduced echelon form
 0 0 1 3 5   3  2  1
   4 3   2  4  2 6  
 0 1 1 ;   ;  5  2  ;   ;  2  1  1
 1 1 1  2 1  2 4   3 1 6 6  
     4  3  1
2. Show that the given matrices are row equivalent and find a sequence of elementary row
operations that will convert A into B
73
U
1 2  3  1
i) A    and B    ,
3 4 1 0 
 2 0  1  3 1  1
   
ii) C   1 1 0  and  3 5 1  .
1 1 1  2 2 0 
   
Unit Summary
In this unit we exposed you to some of the special matrices, operations on matrices such as
addition, and multiplication. We have also proved some theorems and furthermore extended the
Gauss Jordan method of solving systems of linear equations to solving augmented matrices.
Finally, the use of elementary row operations was adequately discussed.
74
U
Assessment
1. Determine by inspection whether a linear system with the given augmented matrix has :
a) Unique solution
b) Infinitely many solutions
c) No solution. Justify your answers
0 0 12 3 2 0 1 1 1 2 3 4 0 1 2 3 4 56
0 1 31 ; 1 2  3 1 1 ; 5 6 7 8 0 ; 6 5 4 3 21
1 0 11 2 4 6 2 0 9 10 11 12 0 7 7 7 7 77
2. For what value(s) of k, if any, will the system have
a) No solution
b) A unique solution
c) Infinitely many solutions?
i) kx  2 y  3 ii) x  ky  1 iii) x  2 y  3 z  2
2 x  4 y  6 kx  y  1 x yz k
2x  y  4z  k 2
iv) x   y  kz  1
x  ky  z  1
kx  y  z  2
3. Give examples of homogeneous systems of m linear equations in n unknowns or

variables with m = n and with m > n that have
75
U
a) Infinitely many solutions
b) A unique solution
4. Use gauss elimination for augmented matrices to solve
a)  x  y  2 z  5 b)  3x  6 y  9 z  0
 3x  y  7 z  22 x  4y  z  6
x  3 y  z  10 2 x  8 y  3z  13
5. Prove that if A is invertible and AB = 0, then B = 0.
Selected answers to the activities
Activity 1.1
2 a) inconsistent b) consistent c) consistent
Activity 1.2
4 104 73
i) x ,y ,z  
3 15 15
Activity 2.1
1. Symmetrical matrices
2 3 1 
 1 2 1 0   
  ;   ;  3 0  1
 2 3   0 1 1 1 2 
 
Skew-symmetrical
0 1  2
 0 2  
  ;   1 0 3 
 2 0  2  3 0 
 
76
U
2. Activity 2.2
1.
 1 0
 5  1  1   13  3 18   
i)   ii)   iii)   1 3  iv)  3 9 
1 1 7   4 17 0    2 4
 
77
U
REFERENCES
Poole D; (200) Linear Algebra: A Modern Introduction
Pacific Groove: Thomson Learning.
Finkbeiner II D.T; (1972) Elements of Linear Algebra
San Franscisco: W.H Freeman & Co.
Hamilton A.G; (1989) Linear Algebra: An introduction with concurrent examples.
New York: Cambridge University Press
James W.D &Noble B; (1988) Applied Linear Algebra
New Jersey: Prentice-Hall
Leon S.J ; (2010) Linear Algebra with applications
New jersey: Pearson Education inc.
78
U
Unit 2
Matrix algebra
2.0 Introduction
Welcome to unit 2. In this unit you are going to learn about different types of matrices. You
will also be taken through matrix operations which you were introduced to in MAT 110.
You will also use matrices to solve systems of linear equations using the Gauss Jordan
Elimination method and elementary row operation. Furthermore you will also be exposed
to various proofs of theorems on matrix algebra. I ‘ m sure you are ready to embark on this
‘journey’
 Define different types of matrices

 Add, subtract and multiply matrices
 Find solutions of systems of equations using matrices
Outcomes
2.1 Definition
A matrix is a rectangular array of numbers called the entries or elements.
Let K be an arbitrary field. A rectangular array of the form
 a11 a12  a1n 
 
 a 21 a 22  a2n 
   
 
a am2  a mn 
 m1
where the aij are scalars (numbers) in , is called a matrix over .
A matrix with m rows and n columns is called an m by n matrix, or m  n matrix; the pair of
numbers m, n  is called its size or shape.
79
U
In the above matrix, the m horizontal n- tuples a11 , a12 , a13 ,  , a1n  , a 21 , a 22 , a 23 , , a 2 n  ,
 , a m1 , a m 2 , a m 3 ,  , a mn  are the rows of the matrix, and the
 a11   a12   a1n 

     
 a 21   a 22   a2n 
n-vertical m –tuples  a31  ,  a32  ,  ,  a  are its columns.
     3n 
     
     
 a m1   a m 2   a mn 
2.2 Matrix notation
Matrices will usually be denoted by capital letters, say A, B … and the elements of the field
K by lower case letters, say a, b …
We also use double subscript notation to refer to the entries of a matrix A.
For instance the entry of matrix A in rows i and columns j is denoted by aij .
Thus, if
0 5 4 
A    , then
 3 9  1
a 23  1 and a12  5
The notation Aij is sometimes used interchangeably with aij . We can therefore completely
   
denote a matrix A by a ij or aij mn
if it is important to specify the size of A.
 a11 a12  a1n 

 
a a 22  a2n 
Thus, in the matrix  21 , i  1, 2, 3, , m and j  1, 2, 3,  , n .
   
 
a am2  a mn 
 m1
The element aij , called the i j- entry or i j component, appears in the i th row and j th column.
Theorem 2.1 Equality of matrices

   
If A  aij is an m  n matrix and B  bij is a p  q matrix, then A  B if and only if (iff)
m  p and n  q and aij  bij for all i and j.
In other words two matrices are equal if they have the same size and if there are
corresponding entries are equal.
Example 2.1
columns, say,
80
U
 a11 a12 a13  a1n   b11 b12 b13  b1n 

   
a a 22 a 23  a2n  b b22 b23  b2 n 
A   21  and B   21 . Then
         
   
a am2 am3  a mn  b bm 2 bm 3  bmn 
 m1  m1
A  B if and only iff
a11  b11 ; a12  b12 ; a13  b13 ; a1n  b1n and so on.
2.3 Special type of matrices

2.3.1 Zero matrix
 
The matrix a ij  0 such that aij  0 (i  1, 2, 3, , m) , ( j  1, 2, 3,, n) is called
a zero matrix.
Example 2.2
 0 0 0
 
A   0 0 0
 0 0 0
 
2.3.2 Square matrix
A matrix with the same number of rows as columns is called a square matrix.
A square matrix with n rows and n columns is said to be of order n, and is called a
an n- square matrix.
Example 2.3
 1 0 0
 
B   0 1 0
0 0 1
 
2.3.3 Triangular matrix
2.3.3.1 Upper triangular matrix
An upper triangular matrix or simply a triangular matrix is a square matrix whose entries
 
below the main diagonal are all zeros; that is A  aij where aij  0 whenever i  j .
81
U
Example 2.4
 a11 a12 a13  a1n   a11 a12 a13  a1n 
   
 0 a 22 a 23  a 2 n   a 22 a 23  a 2 n 
 0 0 a33  a3n  or  a33  a3n 
   
         
   
 0 0 0 0 a nn   a nn 
2.3.3.2 Lower triangular matrix

A lower triangular matrix is a square matrix whose entries above the main diagonal are all
 
zero; that is A  aij where aij  0 whenever i  j .
Example 2.5
 a11 0 0 0 0  a11 
   
 a 21 a 22 0 0 0  a 21 a 22 
a a 32 a33 0 0 or  a a32 a33 
 31   31 
         
   
 a n1 an2 a n3  a nn   a n1 an 2 an3  a nn 
2.3.4 Diagonal matrix
A diagonal matrix is square matrix whose non- diagonal entries are all zero.
Example 2.6
 a1 0 0  0  a1 
   
0 a 22 0  0  a2 
0 0 a33  0 or  a3 
 1   
0 0 0 a4 0   a4 
   
0 0 0  a nn   an 
2.3.5 The unit or identity matrix
The n- square matrix with 1’s on the main diagonal and 0’s elsewhere is called the unit or
identity matrix. It is denoted by I n or simply I.
82
U
Example 2.7
 1 0 0
 
I3   0 1 0
0 0 1
 
 
I  aij where aii  1 , aij  0 , (i  j ) .
The identity matrix I n has the property
AI n  I n A  A for all n  n matrices.

2.3.6 Inverse matrix
 
If A  aij is an n  n matrix then an n  n matrix B such that AB  BA  I n is called an
inverse of the matrix A and is denoted by A 1 .
Theorem 2.2
The inverse of a matrix is unique.
Proof:
Let Q be an inverse of P, then
PQ  QP  I n and Q  QI n  Q(PR )
 (QP) R
 InR
R
QR
Thus, the inverse of P is unique.
2.3.7 Invertible matrix
A matrix which has an inverse is called an invertible or non- singular matrix while a matrix
which has no inverse is called a singular matrix.
83
U
Theorem 2.3
(vi) If A is an invertible matrix, then A 1 is invertible and ( A 1 ) 1  A .
(vii) If A is an invertible matrix and c is a non zero scalar, then cA is an invertible
1
matrix and (cA) 1  A 1
c
(viii) If P and Q are invertible matrices of the same size or n  n matrices, then PQ
is an invertible and ( PQ) 1  Q 1 P 1 .
(ix) If A is an invertible matrix, then ( A t ) 1  ( A 1 ) t

(x) If A is an invertible matrix, then A n is invertible for all non- negative integers n and
( A n ) 1  ( A 1 ) n
Proof of theorem (iii)

If P and Q are invertible then
PP 1  I n  P 1 P and QQ 1  I n  Q 1Q
Now 
( PQ ) (Q 1 P 1 )  ( PQ )Q 1 ( P 1 ) 
 P(QQ 1
) P 1 
 ( PI n )  P 1
 PP 1
 In
Q 1 P 1  ( PQ) 1 or simply ( PQ ) 1  Q 1 P 1 .
2.3.8 Orthogonal matrix
An n  n matrix Q is said to be orthogonal matrix if the column vectors form an orthogonal
set in  n .
Theorem 2.4
An n  n matrix Q is orthogonal if and if Q T Q  I
Proof
It follows from the definition that an n  n matrix Q is orthogonal if and only if its column
vectors satisfy
T
qi q j   ij
84
U
T
However, qi q j is the (i , j ) entry of the matrix Q T Q. Thus Q is orthogonal if and only if
QT Q  I .
It follows from the theorem that if Q is an orthogonal matrix, then Q is invertible and
Q 1  Q T
Example 2.8
Consider the matrix Q below
 cos   sin  
Q   
 sin  cos  
Now Q T    and
  sin  cos  
cos   sin 
det Q 
sin  cos 
det Q  cos 2   sin 2   1
 Q 1   
  sin  cos  
You can see that
 cos   sin  
Q 1  Q T    . Therefore the matrix Q is orthogonal.
 sin  cos  
2.3.9 Idempotent matrix
An n  n matrix A is said to be idempotent if and only if A 2  A .
2.4.0 Nilpotent matrix

An n  n matrix A is said to be nilpotent of index p if and only if A p  Z but
A p 1  Z for some integer p  1 . Any strictly triangular matrix is nilpotent.
An n  n matrix A is called
a) A symmetric matrix if A t  A . That is if and only if Aij  A ji for all i and j
b) A skew symmetric matrix if A t   A

c) An orthogonal matrix if A t A  AA t  I n
85
U
Activity 2.1
2. Which of the following matrices are symmetric, and which are skew- symmetrical.
0 1  2  2 3 1 
 1 2  0 2 1 0     
  ;   ;   ;   1 0 3  ;  3 0  1
 2 3    2 0   0 1  2  3 0   1  1 2 
   
2. Let A be any square matrix. Show that A  AT is a symmetric matrix
3. Show that the products AT A and AAT are symmetric matrices.
4. Show that the following matrices are orthogonal.
 2 1 
 0 
 1 2  6 3  3 4
   
5 5 ;  1 1 1 

 ;  5 5
 2 1 
 2 6 3  4 3
   
 5 5  1 1 1   5 5
 
 2 6 3
10. Show that a product of two orthogonal matrices of the same size is an orthogonal
matrix.
11. Find which of the matrices are invertible and which are singular. Find the inverses
of those which are singular
2 3 2 3 
1 2 1  1 1 2  0 1 1  
1 2        1 0 2 1 
  ;  0 1 2  ;   1 2  1 ;  1 0 1  ;  ;
1 3   0 0 1   1  3 1   1 1 0    1 1 4  2
       
3 0 0 4 
 1 1 1 1 
 
 2 1 0 3 
 3 0 2 5 
 
 1  1  1  3
 
 a 0 0
 
12. Find the inverse of the diagonal matrix  0 b 0  , where a, b, and c are non-zero.
0 0 c
 
13. Let A, B and C be invertible matrices of the same size. Show that the inverse of the
product ABC is C 1 B 1 A 1
86
U
14. Show that, for any invertible matrix A, ( A 1 ) T AT  I and AT ( A 1 ) T  I
2.4 Matrix addition

2.4.1Definition
 
If A  aij is an m  n matrix and B  bij  
is a p  q matrix then, A  B is defined if and only if m  p , and n  q and then
A  B  (aij  bij ) is the m  n matrix obtained by adding the corresponding elements in A
and B.
Example 2.9
columns, say, m  n matrices.
 a11 a12  a1n   b11 b12  b1n 
   
a a 22  a2n  b b22  b2 n 
A   21  and B   21
       
   
a am2 
 a mn    bmn 
 m1  bm1 bm 2
The sum of A and B, written A+B, is the matrix obtained by adding corresponding entries
 a11 a12  a1n   b11 b22  1n
   
a a 22  a 2 n   b21 b22  2n 
A  B   21 
       
   
a am2  a mn   bm1 bm 2  bmn 
 m1
 a11  b11 a12  b12  a1n  b1n 
 
 a  b21 a 22  b22  a 2 n  b2 n 
A  B   21
    
 
a  b a m 2  b1m  a mn  bmn 
 m1 m1
2.5 Scalar Multiplication

   
If A  aij is an m  n matrix and   K , then A   aij is the m  n obtained by
multiplying each element of A by  . A is called the scalar multiple of A by  .
Theorem 2.5
Let V be a set of all m  n matrices over a field K. Then for any matrices A, B, C  V and any
scalars h1 , h2  K ,
xi) ( A  B )  C  A  ( B  C )  Associative
87
U
xii) A  0  A
xiii) A  ( A)  0
xiv) A  B  B  A  commutative
xv) h1 ( A  B)  h1 A  h1 B
xvi) (h1  h2 ) A  h1 A  h2 A
xvii) (h1 h2 ) A  h1 (h2 A)
xviii) 1  A  A
xix) 0A  0
xx) A  A  2 A
2.6 Matrix Multiplication
2.6.1 Definition:
If A  aik  is an m  n matrix and B  bkj  is an n  r matrix, then the product AB is the
m  r matrix C  cij .  
The i, j  entry of the product is computed as follows; multiplying elements in the i th row
of A with the corresponding elements in the j th column of B and summing up. That is
C ij  ai1b j1  a i 2 b j 2    ain b jn .
n
Cij can be represented in short form as C ij   aik bkj , for each i  1, , m and
k 1
each j  1, , n .
The number of columns of A must be the same as the number of rows of B.

Theorem 2.6
Properties of matrix multiplication
Let A, B, and C be matrices (whose sizes are such that the indicated operations can be
performed) and let k be a scalar. Then
(vii) ( AB)C  A( BC )  Associative law
(viii) A( B  C )  AB  AC  distributive law
(ix) ( B  C ) A  BA  CA
(x) k ( AB )  kA( B)  A(kB) , where k is a scalar
(xi) 0 A  0 , where 0 is the zero matrix.
88
U
(xii) I m A  A  AI n if A is m  n  multiplicative identity
3. AB  0 does not necessarily mean that A  0 or B  0 .

4. AB  AC does not imply that B  C .
That is AB  AC  0
 A( B  C )  0 can be true with A  0 and B  C  0 .
2.7 The transpose of a matrix

2.7.1 Definition
 
If A  aij is an m  n matrix, the transpose of A is the n  m matrix denoted by A t
obtained by interchanging the rows and columns of A.
 
That is A t  Bij where
Bij  a ji ; ( (i  1, 2, 3, , n; j  1, 2, 3, , m)
Theorem 2.7
Properties of the transpose
If A and B are m  n and p  q matrices respectively, and k be a scalar.
Then:
vi) ( A  B ) t  At  B t where both sides are defined if and only if m  p and
nq
vii) ( At ) t  A
viii) ( KA) t  K ( A t ) , for k a scalar

ix) ( AB) t  B t A t both sides are defined if and only if n  q
x) ( A r ) t  ( A t ) r
If A is an n  n matrix then A t is invertible if and only if A is invertible.

Further ( A t ) 1  ( A 1 ) t
Proof:
If A is invertible then AA 1  I n  A 1 A
89
U
Transposing gives
( A 1 ) t A t  I n
 A t ( A 1 ) t  I n
And so A t is invertible.
Conversely, if A t is invertible, then A t ( A t ) 1  I n  ( A t ) 1 A t
Transposing gives
A 1 A  I n  AA 1
Clearly from the above also implies by the previous theorem, that is ( A t ) 1  ( A 1 ) t .
Activity 2.2
3. Given that
2
1 1 2 4 0  3  
A    ; B    ; C    1 ;
 0 3 4  1  2 3  3
 
Find
ii) A + B ii) 3A – 4B iii) AT iv) C T AT v) A 3
4. Let D x be the m  square scalar matrix with diagonal elements k. Show that:
iii) For any m  n matrix A, D x A  kA

iv)Show that the sum, product and scalar multiple of:
c) Upper triangular matrices is upper triangular
d) Scalar matrices are scalar.
2.8 Matrices and echelon form

Recall that a system of m linear equations in n variables over a field F, that is
a11 x1  a12 x 2    a1n x n  b1
a 21 x1  a 22 x 2    a 2 n x n  b2

a m1 x1  a m 2 x 2    a mn x n  bm can be written in matrix form as:
90
U
 a11 a12  a1n  x1   b1 

    
 a 21 a 22  a 2 n  x 2   b2 
a a32  a 3n  x3    b3 
 31    
          
    
 a m1 am2  a mn  x n   bm 
However, the above matrix can be written in short or abbreviated form as
AX  B , where A is the matrix of coefficients, X is the matrix of variables and B is the
matrix of constants.
Most often we will dispense (do away) with the matrix of variables X and will simply write
the argument matrix of the system as
a11 a12  a1n b1
a 21 a 22  a 2 n b2
A/ B  a31 a32  a3n b3
    
a m1 am2  a mn bm
Example 2.9.1
Write the argument matrix of the systems given
a) 3x  y  2 z  9
x  5y  z  0
 2 x  10 y  3z  4
b) 3w  2 x  y  2 z  12
2x  y  2z  1
w x  3
Answers
c) The argument matrix is
3 1 2 9
A/ B  1 5 1 0
 2 10 3 4
91
U
d) The argument matrix is

3 2  1  2 12
A / B  0 2 1  2 1
1 1 0 0 3
2.8.1 The pivot
If a row of a matrix is non-zero, we call the first non-zero entry of this row a pivot or
leading entry for this row. For instance, in example (a) above, 3, 1 and -2 are leading
entries for the first row, second row and third row respectively.
2.8.2 The row echelon matrix
A matrix M  M mn (F ) is a row echelon matrix if;
ii) All the zero rows of M, if any, are at the bottom of M

ii) For any two consecutive rows Rt and Rt 1 , either Rt 1 is all zeros or the pivot of
Rt 1 is immediately to the right of the pivot of Rt . In other words in each non-zero
row, the first non-zero entry (pivot) is in a column to the left of any leading entries
below it.
The variables accompanying the pivots are called leading variables. Those variables
which are not leading variables are called free parameters.
Example 2.9.2
State whether the given matrices are in echelon form or not.
1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
       
0 0 1 2 0 0 2 2 0 0 0 0 0 0 0 1
A  , B  , C   , D
0 0 1 2 0 0 0 3 0 0 0 1 0 0 0 0
       
0 0 0 0  0 0 0 0  0 0 0 0  0 0 0 0 
   
Answers
Matrices A and C are not in echelon form because they dot not satisfy the two conditions
above. For instance matrix A does not meet condition (ii) while matrix C does not satisfy
condition (i).
Matrices B and D on the other hand, are in row echelon form because they both meet the
two conditions.
92
U
2.8.3 Reduced row Echelon form

A matrix is in reduced row echelon form if it satisfies the following properties:
iv) Any rows consisting entirely of zeros are at the bottom.
v) The leading entry in each non-zero row is a 1 (called a leading 1).
vi) Each column containing a leading 1 has zeros everywhere else.
The matrix below is an example of a reduced row echelon matrix
1 3 0 2 0 1 0
 
0 1 0 4 5 1 0
0 0 0 1 3 2 0
 
0 0 0 0 0 0 1
0 0 0 0 0 0 0 

2.9 The Gauss - Jordan Elimination method.
The Gauss – Jordan elimination is a method employed in further reducing the matrix in
echelon form. It follows the following steps:
iv) Write the argument matrix of the system of linear equations
v) Use elementary row operations to reduce the argument matrix to reduced row
echelon form
vi) If the resulting system is consistent, solve for the leading variables in terms of any
remaining free variables.
Example
Solve the system
x 2  x3  5
2 x1  x 2  5 x3  1
x1  x 2  3 x3  3
Solution
We begin by writing the argument matrix as follows:
93
U
0 1 15
2 1 51
1 1 33
Now interchanging row 1 ( R1 ) and row 3 ( R3 ) we obtain
1 1 33
2 1 51
0 1 15
Replacing R2 with R2  2R1 , we obtain
1 1 1 3
0 1 1  5
0 1 1 5
Replacing R3 with R3  R2 , we obtain
1 1 1 3
0 1 1  5 which is a matrix in echelon form.
0 0 0 0
The system has now been reduced to

x1  x 2  x3  3
 x 2  x 3  5
Since we have more unknowns (variables) than equations, the system has an infinite
number of solutions. That is any value of x1 , x2 and x3 satisfy the equation represented by
the third row .
Example
Solve the system of linear equations
1 1 1 1  x    3 
    
0 2 1 0  y    1 
0 
0 1  1 z   9 
    
0 0 0 2  w    6 

Solution
We observe that the matrix of coefficients is already in row-echelon form. Clearly every
variable is a leading variable, and by back substitution
2w  6
94
U
 w  3
Then
zw4
z  43
 z 1
Also
2 y  z  1
2 y  1  1
y  1
And
x  y  z  w  3
x  1  1  3  3
x0
x  0 
   
 y    1
Therefore the unique solution is     
z 1
   
 w    3
   
Given a matrix A  M mn (F ) , by following Gauss Jordan reduction, we can find a matrix
P  Glm ( f ) such that PA  B is in row-echelon form
Activity 2.3
1. State which of the following matrices is in Gauss-reduced form, and for those that are
not, explain why not.
0 1 2
2 1 3 2  1 3 1 2  
    0 0 0
a)  0 0 1  2 b) 0 0 0 0 c) 0
0 0 0 0   0 1 1 3 0 0
     
0 0 0 

2. Find a Gauss- reduced form for each of the following matrices.
95
U
1 2 3 4 7
 1 1  8  14   
   1  2 1 4 0 5 7 6 8
a) 3  4  3 0  b)   c) 0
 2  1  7  10   2  3 1 2 5 2 4 7
   
0 0 2 1 0 

 4 1 2 6 
 
d)   1 5  1  3
3 4 1 3 

2.8 Elementary matrices

2.8.1 Definition
An n  n matrix is called an elementary matrix if it can be obtained by applying an
elementary row operation on the identity matrix ( I n ).
An Elementary row operation or simply row operation on a matrix A is any one of the
following three types of operations:
(iv) Interchange of two rows of A. That is ri  r j
(v) Replacement of a row ri of A by ri for some number   0 . That is ri  ri
(vi) Replacement of a row ri of A by the sum ri  r j of that row and a multiple of

another row r j of A. That is ri  ri  r j
Example
iv) ri  ri
1 0 0 0 0 1
   
 0 1 0  r1  r3  0 1 0 
0 0 1 1 0 0
   
v) ri  ri
1 0 0 1 0 0 
   
 0 1 0  r3  2r3  0 1 0 
0 0 1  0 0  2
   
vi) ri  ri  r j
96
U
1 0 0 1 0 0 
   
 0 1 0  r2  2r3  r2  0 1 2 
0 0 1  0 0  2
   
Theorem 2.7
If an m  n matrix B can be obtained from an m  n matrix A by applying an elementary
row operation, then B is equal to the product of the corresponding m  n elementary matrix
A. If e is the elementary row operation, then
B  eA  EA
Proof:
We shall prove the theorem by considering each separate type.
iv) If ri  ri , then
  11  12   1n 
 
   i 2   in 
eA   ii
   
 
  m2   mn 
 m1
1    11  12   1n 
  
     ii  i 2   in 
 EA  eA  B
 1      
  
 1    m1  m2   mn 

  12  12   1n 
 
  11  12   1n 
M i ( ) A  
   
 
    
 m1 m2 mn 
v) If ri  r j , then
  11   1n 
 
i   j1  in 
H ij A      B
 
j   j1   1n 
 
  mi   mn 
97
U
vi) If ri  ri   r j , then
  11   1n 
 
  i1   j1   in   in 
Aij ( ) A   B
 
 
    mn 
 mi
Example
1 0 1 1 0 1
  r2  2 r1  r2  
Let A   2  1 0     0  1 2   B
1 0 0 1 0 0
   
Next we apply the same elementary row operation on the identity matrix I 3 . We obtain
1 0 0  1 0 0
  r2  2 r  r2  
 0 1 0     2 1 0   E
0 0 1  0 0 1
   
 1 0 01 0 1
  
Now note that EA    2 1 0   2  1 0 
 0 0 1  1 0 0
  
1 0 1 
 
 0 1  2  B
1 0 0 

Theorem 2.8
Every elementary matrix is invertible and the inverse of each elementary is elementary.
Proof:
We shall consider each separate type by verifying that
1
(iv) ( M i ( ) 1  M i  
 
(v) ( H ij ) 1  H ij
(vi) ( Aij ( ) 1  Aij ( )
98
U
Proving using (i)

1 
 
 1 
1  1 
M i    row i 
   
 
 
 1 

1 1  1 
    
       
1  1    
M i    M i ( )      1
     
       
 

 1   1   1 
1
= M i ( ) M i   .
 
Proving (iii)
( H ij ) 1 H ij  H ij H ij ri  r j
1  1  1 
    
 1 1   1    
 0   0    1 
    
 1   1     
 1   1   1 

= H ij ( H ij ) 1
Proof of (iii)
( Aij ( )) 1  Aij ( )  Aij ( ) Aij ( )
99
U
1  1 
   
 1    1  
     
    ri  r j  ri
 1   1 
     
  
 1   1 
 
1 
 
 1 0 
  
= 
 1 
  

 1 

= Aij ( )  Aij ( )
Example
1 0 0 1 0 0
  r3 2 r1  r3  
 0 1 0     0 1 0 
0 0 1  2 0 1
   
 1 0 0
1
 
E  A31 (2)   0 1 0 
 2 0 1
 
 1 0 0  1 0 0 
1
  
E E   0 1 0  0 1 0 
  2 0 1  2 0 1 
  
1 0 0
 
=  0 1 0   EE 1
0 0 1
 
2.9 Row equivalent matrices
2.9.1 Definition
A matrix B is row equivalent to matrix A if there exists a finite sequence E1 , E 2 , , E k of
elementary matrices such that B  E k E k 1  E1 A .
100
U
In other words, B is row equivalent to A if B can be obtained from A by a finite number of

row operations.
Theorem 2.9
If an m  n matrix A is row equivalent to an m  n matrix B, then there exists (  ) a finite
number of elementary matrices 1 ,  2 ,  3 , , E k A
Proof:
B is row equivalent to A  B can be obtained from A by finite, say k sequences of
elementary row operations. We shall prove the theorem by induction.
If k  1, then B  1 A
If K  1, we assume that if a matrix B' can be obtained from A by k  1 elementary row

operations, then  elementary matrices  2 ,  3 , ,  k such that B'   2  3 , ,  k A
Now B is obtained from B' by applying one further elementary row operation that is
B  1 B'
for some elementary matrix 1 thus B  1 ( 2  3   k A)  1 2  E k A) as required.
Cor 2.1
If R is the reduced echelon matrix of A, then there exists a finite number of elementary
matrices  '1 ,   k such that :
A  1  2   k R
Proof:
By theorem 2.9, there exist elementary matrices  '1 ,  ' 2 , ,  ' k such that
R   '1 2   ' k A .
' ' '
But 1 ,  2 , ,  k is non-singular and its inverse is also an elementary matrix. Thus
' 1 ' 1
A   k  '1 k 1  1 R, as required
A  1  2   k R .
Cor 2.2
The following matrices are equivalent.
iv) The n  n matrix A is invertible
v) The homogeneous system AX  0 of n linear equations in n variable has no trivial
solution
vi) A is a product of elementary matrices.\
101
U
Proof : (i )  (ii )
If A is invertible, the system of n equations in n variables given by AX  0 has non-trivial
solution since X  ( A 1 A) X  A 1 ( AX )
 A 1 (0)
0
(ii )  (iii ) the reduced echelon matrix is I n and thus A  1 2   k I n

That is
A  1  2   k
This implies that A is a product of elementary matrices.
(iii )  (i) If A  1 2  E k , where each  i is an elementary matrix then since elementary
matrices are invertible, then
1 1
A 1  E k   1 k 1  1
( AB) 1  B 1 A 1
Thus A is invertible.
To determine whether a matrix is invertible or not we note that if the reduced echelon
matrix of A is R and
iii) If R  I n , the matrix A is invertible
iv) If R contains a row of zeros, then A is not invertible.
Corollary 2.2 is very important because it gives us an explicit method for finding the
inverse of an invertible matrix. If A is non-singular, then the reduced echelon matrix of A
is the identity matrix.
If 1 ,  2 , ,  k are the elementary matrices which correspond to the elementary row
operations which must be performed on A to give I n . That is I n   k  k 1  1 A , then
1 1
A  1 ,  2   k I n and A 1   k  k 11 I n . That is A 1 is determined by applying the
same elementary row operations to I n as were used to obtain I n from A.
Example
102
U
2 4 6 
 
Let A   4 5 6  , calculate A 1
 3 1  2
 
Solution
We first put I next to A in an augmented matrix form as follows
2 4 6 1 0 0
4 5 6 0 1 0
3 1 2 0 0 0
And perform the elementary row operations to obtain a reduced row echelon matrix
2 4 6 1 0 0
4 5 6 0 1 0
3 1 2 0 0 0
8 7
 1
1 0 0 3 3
13 11
 0 1 0  2
3 3
0 0 1 11 5
 1
6 3
 8 7 
  1
 3 3 
 13 11
1
Since A has now been reduced A   2
 3 3 
 11 5 
  1
 6 3 
  16 14  6 
1
1

A   26  22 12 
6 
  11 10  6 
Example
 1  3 4
 
Is the matrix A given by A   2  5 7  invertible?
0 1 1
 
Solution
103
U
1 3 4 1 0 0
2  5 7 0 1 0 r
1  2 r1  r3

0 1 1 0 0 1
1  3 4 1 0 0 r1 3 r2  r1

r3  r2  r3
0 1  1  2 1 0   
0 1 1 0 0 1
1 0 1 5 3 0
0 1 1  2 1 0
0 0 0 2 1 1
This is as far as we can go. The matrix A cannot be reduced to the identity matrix and we
conclude that A is not invertible.
If in the row reduction of A we end up with a row of zeros, then A is not invertible.
2.9.1 Elementary Column Operations and Equivalence of matrices

The definition of elementary row operations and row equivalent matrices may be adapted
to define elementary column operations on a matrix and column equivalent matrices.
We know that if R is the reduced echelon matrix of an m  n matrix A, then there exist an
invertible matrix P such that
PA  R
By theorem 2.9 R  1 2   k A where E k ,  k 1 , 1 are the elementary matrices which

correspond to the sequence of elementary row operations which are carried out
successively in order to obtain r from matrix A.
Thus P  1 2   k
P  1 ( 2 (  k 1 ( k I m )) )
And the matrix P is obtained by applying to I m the same elementary row operations as
were used to obtain R from A.
Example
104
U
 2 1 0 1 
 
A   1 2 1  1
 2 9 4  5
 
2 1 0 1 1 0 0 r
1  2 r2  r1

1 2 1 1 0 1 0
2 9 4  5 0 0 1 r
3  2 r2  r3
 
1
r1   r1
5
0  5  2 3 1  2 0    
1 2 1 1 0 1 0
r3  r1  r3
0 5 2  3 0  2 1   
2 3 1 2
0 1   0
5 5 5 5
0 2 1 1 0 1 0 r
1  r2

0 0 0 0 1 4 1
1 2 1 1 0 1 0
2 3 1 2
0 1   0 r
1  2 r1  r1

5 5 5 5
0 0 0 0 1 4 1
1 1 2 1
1 0 0
5 5 5 5
2 3 1 2
0 1   0
5 5 5 5
0 0 0 0 1 4 1
 2 1   1 1 
 0 1 0 
 5 5   2  1 0 1   5 5 
 1 2   2 3
0   1 2 1  1   0 1   ; PA= R
 5 5   5 5
 
 1  4 1   2 9 4  5  0 0 0 0 
   
   
105
U
Theorem 2.9 and the fact that every m  n matrix is row equivalent to an m  n reduced
echelon matrix have their equivalent statements in terms of elementary column operations.
This leads us to the following theorem;
Theorem 2.9.1
If an m  n matrix A is column equivalent to an m  n matrix B, then there exist a finite
number of elementary matrices 1 ,  2 , ,  l such that
B  A1 2   l
B  1  2   k A
From theorem 2.9.1 we see that there exists an invertible m  n matrix Q such that
B = AQ
Let us now apply this to the reduced echelon matrix R of an m  n matrix A.
Suppose that the leading 1s appear in columns j1 j 2 , j r then by subtracting suitable
multiples of these columns from succeeding columns and then columns interchanges
ci  c j , c 2  cj 2 ,, cr  cjr , we have that R is column equivalent to the matrix
 1r 0r  ( n  r ) 
N    where 1r is the r  r identity matrix and 0 p  q is the zero
 0 ( mn ) r 0 ( mr )( n r ) 
p  q matrix.
Theorem 2.9.2
If A is an m  n then there exist an invertible m  m matrix P and an invertible m  n matrix
Q such that PAQ  N where N is the matrix given above.
The matrices P and Q are calculated as explained above and an illustration is given in the
example below
Example
 2 1 0 1 
 
If A   1 2 1  1  from the example above if
 2 9 4  5
 
 2 1 
  0
 5 5 
 1 2
P  0  then
 5 5 
 1  4 1
 
 
106
U
 1 1 
1 0 
 5 5 
2 3
PA   0 1  
 5 5
0 0 0 0 
 
 
Now we apply the elementary column operations on the argument matrix.
 1 1 
1 0 
 5 5 
0 1 2 3
 
 5 5
0 0 0 0 
 
 
PAQ  3  3 3  4  3  4
1 1 1
c3   c1  c3
1 0 1 0 0 0   
 5
5 5  
2 3 0 1 0 0
0 1  0
5 5 0 1 0 1
c 4   c1  c4
0 0 0 0   5
  

0 0 0 1 

2
 1 1 5
c3   c2  c3
1 0 0 0 1 0      

 5 5
2 3
0 1  0 1 0 0 
3
5 5 0 0 1 0  5
c 4  c 2  c4
0 0 0 0     
0 0 0 1 
 1 1
1 0   
1 0 0 0 5 5
2 3 
0 1 0 0 0 1 
 5 5 
0 0 0 0 0 0 1 0 
 
0 0 0 1 
 1 1
1 0   
 5 5
2 3 
 Q  0 1 
 5 5 
0 0 1 0 
 
0 0 0 1 
107
U
 1 0 0 0
 
 PAQ  N   0 1 0 0 
 0 0 0 0
 
2.9.2 Equivalent Matrices
2.9.2.1 Definition
Two m  n matrices A and B are equivalent if  invertible matrices P and Q such that
PAQ  B
If PAQ  B, then P 1 BQ 1  A
PAQ  B and P BQ   C then ( P P ) A(QQ )  C
What has been proved above is that every m  n A is equivalent to a matrix with simple
form N, where r denotes the number of non zero rows in the reduced echelon form of A
N may be regarded as a canonical form of A under this equivalence relation. That is a
representative of the equivalence class which contains A under this equivalence relation.
Later it will be shown that r is the rank of the matrix A and N is called the normal form of
the matrix A.
Example
Find invertible matrices P and Q such that PAQ is normal form for the matrix
 2 1 1
 
 1 1 1
A
0 1 3
 
 1 2 4 

Solution
Since A is 4  3 matrix we set an argument matrix using the associated 4  4 identity matrix
and reduce it to reduced row echelon form.
2 1 1 1 0 0 0  r
1  r4

 
1 1 1 0 1 0 0
0 1 3 0 0 1 0
 
1  2 4  0 0 0 1 
1  2  4 0 0 0 1 r
2  r1  r2
 
1 1 1 0 1 0 0
0 1 3 0 0 1 0 r
4   r1  r4

2 1 1 1 0 0 0
108
U
1 2 4 0 0 0 1 r 2  r3

r3  r2  r3
0 1  3 0 1 0 1   
r4 3r2  r4 2
0 1 3 0 0 1 0  
0 3 9 1 0 0 2
1 0 2 0 0 2 1 0 0 2 1 
 
0 1 3 0 0 1 0 0 0 1 0 
P
0 0 0 0 1 1 1 0 1 1 1 
 
0 0 0 1 0 3 2 1 0  3  2 

0 0 2 1  2 1 1  1 0 2
    
0 0 1 0  1 1 1  0 1 3
Thus PA   
0 1 1 1  0 1 3  0 0 0
    
1 0  3  2   1  2  4   0 0 0 

We then set an argument matrix using the associated 3 3 identity matrix and then apply
the elementary column operations.
1 0 2
1 0 0 C3  2 c1  c3
 
0 1 3 c3  3c 2  c3
0 1 0    
0 0 0
0 0 1
0 0 0
1 0 0
1 0 2 1 0  2
0 1 0  
0 1 3  Q   0 1  3
0 0 0 0 0 1 
0 0 1  
0 0 0
Activity 2.4
3. Use elementary row operations to reduce the given matrix to:
c) Row echelon form
d) Reduced echelon form
 0 0 1 3 5   3  2  1
   4 3   2  4  2 6  
 0 1 1 ;   ;  5  2  ;   ;  2  1  1
 1 1 1  2 1  2 4  3 1 6 6  
     4  3  1
4. Show that the given matrices are row equivalent and find a sequence of elementary
row operations that will convert A into B
109
U
1 2  3  1
iii) A    and B    ,
3 4 1 0 
 2 0  1  3 1  1
   
iv) C   1 1 0  and  3 5 1  .
1 1 1  2 2 0 
   
Unit Summary
In this unit we exposed you to some of the special matrices, operations on matrices such as
addition, and multiplication. We have also proved some theorems and furthermore
extended the Gauss Jordan method of solving systems of linear equations to solving
augmented matrices. Finally, the use of elementary row operations was adequately
discussed.
110
U
Assessment
6. Determine by inspection whether a linear system with the given augmented matrix
has :
d) Unique solution
e) Infinitely many solutions
f) No solution. Justify your answers
0 0 12 3 2 0 1 1 1 2 3 4 0 1 2 3 4 56
0 1 31 ; 1 2  3 1 1 ; 5 6 7 8 0 ; 6 5 4 3 21
1 0 11 2 4 6 2 0 9 10 11 12 0 7 7 7 7 77
7. For what value(s) of k, if any, will the system have

d) No solution
e) A unique solution
f) Infinitely many solutions?
ii) kx  2 y  3 ii) x  ky  1 iii) x  2 y  3 z  2
2 x  4 y  6 kx  y  1 x yz k
2x  y  4z  k 2
iv) x   y  kz  1
x  ky  z  1
kx  y  z  2
8. Give examples of homogeneous systems of m linear equations in n unknowns or
variables with m = n and with m > n that have
c) Infinitely many solutions
d) A unique solution
9. Use gauss elimination for augmented matrices to solve
b)  x  y  2 z  5 b)  3x  6 y  9 z  0
 3x  y  7 z  22 x  4y  z  6
x  3 y  z  10 2 x  8 y  3z  13
10. Prove that if A is invertible and AB = 0, then B = 0.
111
U
Selected answers to the activities

Activity 1.1
2 a) inconsistent b) consistent c) consistent
Activity 1.2
4 104 73
ii) x   , y   ,z  
3 15 15
Activity 2.1
3. Symmetrical matrices
2 3 1 
 1 2 1 0   
  ;   ;  3 0  1
 2 3   0 1 1 1 2 
 
Skew-symmetrical
0 1  2
 0 2  
  ;   1 0 3 
 2 0  2  3 0 
 
4. Activity 2.2
1.
 1 0
 5  1  1   13  3 18   
ii)   ii)   iii)   1 3  iv)  3 9 
1 1 7   4 17 0    2 4
 
112
U
REFERENCES
Poole D; (200) Linear Algebra: A Modern Introduction
Pacific Groove: Thomson Learning.
Finkbeiner II D.T; (1972) Elements of Linear Algebra
San Franscisco: W.H Freeman & Co.
Hamilton A.G; (1989) Linear Algebra: An introduction with concurrent examples.
New York: Cambridge University Press
James W.D &Noble B; (1988) Applied Linear Algebra
New Jersey: Prentice-Hall
Leon S.J ; (2010) Linear Algebra with applications
New jersey: Pearson Education inc.
113

MAT 260 Module 1 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MAT 260 Module 1 PDF

Uploaded by

Copyright:

Available Formats

MODULE

LINEAR ALGEBRA- MAT 260

COPPERBELT UNIVERSITY COLLEGE

COPPERBELT UNIVERSITY COLLEGE

How this module is structured ....................................................................................... 1

Welcome to module 1 MAT 260 .................................................................................... 3

Getting around this Module 7

Margin icons .................................................................................................................... 7

Systems of linear equations ...............................................Error! Bookmark not defined.

summary .............................................................................Error! Bookmark not defined.

Matrix Algebra .............................................................................................................. 79

Unit summary .....................................................................Error! Bookmark not defined.

About this module

How this module is structured

The module overview

 If the module is suitable for you.

 What you will already need to know.

 What you can expect from the module.

The overview also provides guidance on:

 Where to get help.

 Module assignments and assessments.

We strongly recommend that you read the overview carefully

The module content

 An introduction to the unit content.

 Core content of the unit with a variety of learning activities.

 Assignments and/or assessments, as applicable.

 Module content and structure.

 Module reading materials and resources.

 Module support (assigned tutors, technical help, etc.)

Your constructive feedback will help us to improve and enhance

Welcome to module 1 MAT 260

Module 1 MAT 260

Is this module for you?

i) Solve systems of equations

The expected study time is 100 hours

As an adult learner your approach to learning will be different to that from

Essentially you will be taking control of your learning environment. As a

We recommend that you take time now—before starting your self-study—

For any help contact the University College through:

Help 1. University College Secretary - Phone +260 212293003

3. The Mathematics HOD -

4. District Resource centre in your district -

All contacts during Office hours.

Three, tutor marked assignments will be considered for assessment.

To be submitted to the university college through District Resource

Getting around this module

Activity Assessment Assignment Case study

Discussion Group activity Help Note it!

Outcomes Reading Reflection Study skills

Upon completion of this unit you will be able to:

 State the type of solution the system of linear equation has.

 Solve systems of linear equations using the Jordan Gaussian

1.1 Definition of the general linear equation

Let us consider a linear equation in any arbitrary number of unknowns, say

A linear equation in n unknowns x1 , x 2 , x3 , x n over a real field  is an equation of the form

a1 x1  a 2 x 2    a n x n  b , where a1 , a 2 , a n and b are given real numbers

and the xi are in determinants or variables. The numbers ai are coefficients of xi

Now   (3, 2,1, 0) is one of the solutions of the equation since

1.2 Solving linear equations

These points give the graph below

Then every n – tuple of real numbers is a solution of the equation.

1.3 System of linear equations

a11 x1  a12 x 2    a1n x n  b1