ENGINEERING MATHEMATICS I

Dr. Alex Samuel BAMUNOBA

(bamunoba@gmail.com )

September 1, 2015
Contents

1 Matrices 4

1.1 Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.1.1 Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.1.2 Operations of matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.1.3 Linear equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.2 Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

1.2.1 Permutations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

1.2.2 Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

1.2.3 Minors and Cofactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

References 22

i
Introduction

Engineering Mathematics also called Mathematical Engineering is a branch of applied mathemat-

ics that concerns mathematical methods and techniques that are typically used in engineering and

industry. Along with elds like engineering physics and engineering geology (both of which may

belong in the wider category engineering science), engineering mathematics is an interdisciplinary

subject motivated by engineers' needs both for practical, theoretical and other considerations out-

with their specialization, and to deal with constraints to be eective in their work. It is an art of

applying mathematics to complex real-world problems. It combines mathematical theory, practical

engineering and scientic computing to address today's technological challenges. It comprises of

topics from pure mathematics, mathematical physics, applied mathematics, computational math-

ematics as well as statistics. With this in mind, engineering mathematics is therefore a creative

and exciting discipline, spanning traditional boundaries. Engineering mathematicians can be found

in an extraordinarily wide range of careers, from designing next generation Formula One cars to

working at the cutting edge of robotics, from running their own business creating new autonomous

vehicles to developing innovative indices for leading global nancial institutions.

Historically, engineering mathematics consisted mostly of applied analysis, most notably: dier-

ential equations; real and complex analysis (including vector and tensor analysis); approximation

theory (broadly construed, to include asymptotic, variational, and perturbative methods, represen-

tations, numerical analysis); Fourier analysis; potential theory; as well as linear algebra and applied

probability, outside of analysis. These areas of mathematics were intimately tied to the develop-

ment of Newtonian physics, and the mathematical physics of that period. This history also left a

legacy: until the early 20th century subjects such as classical mechanics were often taught in applied

mathematics departments at some American and European universities, and uid mechanics may

still be taught in (applied) mathematics as well as engineering departments.

The success of modern numerical computer methods and software has led to the emergence of com-

putational mathematics, computational science, and computational engineering (the last two are

sometimes lumped together and abbreviated as CS&E), which occasionally use high-performance

computing for the simulation of phenomena and the solution of problems in the sciences and engi-

neering. These are often considered interdisciplinary elds, but are also of interest to engineering

mathematics. Specialized branches include engineering optimization and engineering statistics.

ˆ Course Description

1
 Page 2

According to the College of Engineering, Design, Art and Technology (CEDAT) handbook,

Engineering Mathematics is fundamental to the study of Engineering. It provides the necessary

analytical skills for the study of more advanced subjects. To achieve this, the subject has been

divided into 4 sometimes overlapping courses labelled Engineering Mathematics I, II, III and

IV. In particular, Engineering Mathematics I, (EMT 1101) comprises of elementary dierential

and integral calculus, with some elements of matrix theory. Below we spell out its objectives.

ˆ Objectives

 To provide introductory treatment of mathematical concepts fundamental to Engineering.

 To consolidate and advance the material covered in pre-University Mathematics. This

course also provides the mathematical tools needed in other semesters' course units.

 To develop the analytical and critical thinking abilities fundamental to problem solving.

ˆ Course Content

 Concept of a Function [10CH]

Denition, properties, range, domain of the elementary real valued functions, concept of

a limit of a real valued function, continuity, indeterminate forms and l`Hopital's rules.

 Complex Variable Algebra [6CH]

Cartesian and polar algebra representations, absolute values; algebra of complex num-

bers i.e., products, powers and quotients; extraction of roots, De Moivre's Theorem,

exponential and hyperbolic functions of the complex variable.

 Dierential Calculus [12CH]

The Derivative: denitions, notation, properties and Theorems, dierentiation of ele-

mentary functions of a real variable. Applications: optimization, curve sketching, ap-

proximations. Multivariable dierentiation: partial derivatives, further applications.

 Integral Calculus [12CH]

The Integral: denition and properties, Fundamental Theorem of Calculus, Techniques of

integration. Denite integral: its interpretation as area under a curve and its applications

such as: length of a curve, area bound between curves, volume of revolution, moments.

Improper integrals and their evaluation using limits, integration of a continuous function;

inequalities. The denite integral as a function of its upper limit. Dierentiation of an

integral containing a parameter; double integrals and their applications.

 Page 3

 Linear Transformations and Matrices [12CH]

Denitions and types of matrices, algebraic operations on matrices: sums, products,

transposition of matrices, equality of matrices. Determinants: denition and properties;

minors and cofactors; evaluation of determinants by cofactors; rank of a matrix; inverse

matrices. Solution of systems of linear algebraic equations; consistent and inconsis-

tent equations; systems of homogeneous equations; Crammer's Rule; The Gauss-Jordan

method, Gaussian elimination.

 Vector Algebra [8CH]

Denitions: scalars, vectors, unit vector, and dimensionality. Operations on vectors: ad-

dition, subtraction, multiplication, dot and cross products, position and distance vectors.

ˆ Learning Outcomes
On completing the course the student should be able to:

 Consolidate fundamental mathematical principles learned in high school.

 Relate mathematics to the physical world, providing a sound basis for later specialization.

 Acquire persistence and manipulative skills required by engineers.

 Analyze a variety of complex relationships present in modern engineering systems.

ˆ Mode of teaching/delivery and assessment

The course shall be conducted through lectures and tutorials. Assessment will be done through

continuous interim assessments (assignments and tests) and a nal examination. Interim

assessment will carry a total of 40% and nal examination will carry 60% of the nal mark.

These notes contain more than 200 carefully selected problems and are intended to help the reader

in better understanding them, developing skills and intuition in engineering mathematics. Some

problems are very simple, to encourage the beginner. The development of the material does not

depend on the problems and omission of some or all of them does not destroy the continuity. We

strongly warn the students following the course, these notes are not a substitute for lectures.
1. Matrices

1.1 Matrices

In working with a system of linear equations or physical, economic and biological models that can

be transformed into systems of linear equations, only the coecients and their respective positions

are important. Also reducing to echelon form, it is essential to keep equations carefully aligned.

Therefore, these coecients can be eciently arranged in a rectangular array of numbers called

a matrix. Moreover, in Linear Algebra, certain abstract objects are introduced such as change of

basis, linear operator and bilinear form which can also be represented by these rectangular arrays.

In this chapter, we shall study these rectangular arrays, i.e., matrices and certain algebraic oper-

ations dened on them. The material introduced here is mainly computational. However, as with

linear equations, the abstract treatment presented in the subjects of Linear Algebra, Functional

Analysis, Abstract Algebra and many others gives more insight into the structure of these matrices.

1.1.1 Matrices

Let K be an arbitrary eld (for this course K will be R or C). A rectangular array of the form

 
a11 a12 · · · a1n
 
a21 a22 · · · a2n 
 
 .. . . . ,
 
 . . . . 
. . .
 
a11 a12 · · · a1n

where the aij are scalars, is called a matrix over K, or simply a matrix if K is implicit. It is also

denoted by (aij ), i = 1, . . . , m and j = 1, . . . , n or simply (aij ). The m horizontal n-tuples are called
the rows of the matrix and the n-vertical m-tuples are its columns. The element aij is called the

ij -entry or ij -component; it appears in the ith row and j th column. A matrix with m rows and n
columns is called an m by n matrix or m×n matrix; and the pair (m, n) is called its size or shape.
Matrices will be denoted by upper case letters A, B, . . . and the elements of the eld by lower case

letters a, b, . . .. For example, the above matrix is denoted and dened as A = (aij ), where aij ∈ K .

Two matrices A and B are equal, written A = B if they have the same shape and if the corresponding
elements are equal, i.e., aij = bij , for all i = 1, . . . , m and j = 1, . . . , n. Therefore, equality of two

4
 Section 1.1. Matrices Page 5

m×n matrices is equivalent to a system of mn equalities, one for each pair of elements.

A matrix with one row is also called a row vector and one with one column is called a column vector.
In particular, an element in the eld K can be viewed as a 1×1 matrix. A matrix with all its

entries or components as 0's is called a zero matrix, and shall be denoted by 0. There are several

types of matrices, e.g., square and non-square matrices, but we shall discuss them as we go along.

1.1.2 Operations of matrices

Let A and B be two matrices with the same size; say m × n: The sum of A and B is the matrix

A+B obtained by summing corresponding entries, A + B = (aij + bij ). The product of the scalar
k by the matrix A written k·A or simply kA is the matrix obtained by multiplying each entry of

A by k : kA = (kaij ). We dene −A = −1 · A and A − B = A + (−B). Observe that;

1. A+B and kA are also m×n matrices.

2. the sum of matrices with dierent sizes is not dened.

Basic properties of matrices under the operations of matrix addition and scalar multiplication follow.

1.1.1 Theorem. Let V be a set of all m × n matrices over a eld K . Then for any matrices
A, B, C ∈ V and any scalars k, k1 , k2 ∈ K ,

1. A + (B + C) = (A + B) + C

2. A + 0 = A

3. A + (−A) = 0

4. A + B = B + A

5. k(A + B) = kA + kB

6. (k1 + k2 )A = k1 A + k2 A

7. (k1 k2 A) = k1 (k2 A)

8. 1 · A = A and 0 · A = 0.

The above results show that V has a vector space structure, (or is a vector space) over K.
 Section 1.1. Matrices Page 6

The product of two matrices A and B, written AB is somewhat complicated. Formally, suppose

A = (aij ) and B = (bij ) are matrices such that the number of columns of A is equal to the number

of rows in B; say A is an m×p matrix and B is a p×n matrix. Then the product AB is the m×n
matrix whose ij -entry is obtained by multiplying the ith row Ai of A by the j th column Bj of B,
Pp Pp
i.e., AB = (Ai · B j ) as follows: Ai · B j = k=1 aik bkj , i.e., if C = AB , then cij = k=1 aik bkj . We
emphasize that AB is not dened if A is an m × p matrix and B is a q × n matrix, where p 6= q .
   
1 2 −1 2
Consider A=  and B =  . It is easy to see that both AB and BA are dened.
3 2 0 2

      
1 2 −1 2 1(−1) + 2(0) 1(2) + 2(2) −1 6
AB =   = = ,
3 2 0 2 3(−1) + 2(0) 3(2) + 2(2) −3 10
      
−1 2 1 2 (−1)(1) + 2(3) (−1)(2) + 2(2) .5 2
BA =   = = 
0 2 3 2 0(2) + 2(3) 0(2) + 2(2) 6 4

It is clear that AB 6= BA. (We are even lucky, it may not even be denable) In general, matrix

multiplication is not commutative, i.e., AB 6= BA. It does however satisfy the following properties.

1.1.2 Theorem. Let A, B and C be conformable matrices over a eld K and k be a scalar. Then

1. A(BC) = (AB)C .

2. A(B + C) = AB + AC .

3. (B + C)A = BA + CA.

4. k(AB) = (kA)B = A(kB)

Assume the sums and products in the above theorem are dened. Moreover, 0 · A = 0 and B · 0 = 0.

The transpose of a matrix A, written as AT is the matrix obtained by writing the rows of A, in

order, as columns: if A= (aij ), then AT = (aji ). Observe that if A is an m×n T

matrix, then A is

an n×m matrix. The transpose operation on matrices satises the following properties.

1.1.3 Theorem. .

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

2. (AT )T = A

3. (kA)T = kAT
 Section 1.1. Matrices Page 7

4. (AB)T = B T AT .

Here are some few exercises.

1. Let

   
    2 −3 0 1 2
1 −1 2 4 0 −3    
A= , B =   , C =  5 −1 −4 2 , D = −1 .
0 3 4 −1 −2 3
   
−1 0 0 3 3

Find

(a) i. A+B
ii. A+C
iii. 3A − 4B

(b) i. AB
ii. AC
iii. AD
iv. BC
v. BD
vi. CD

(c) i. AT
ii. AT C
iii. D T AT
iv. BT A
v. DT D
vi. DDT

2. Construct suitable examples of matrices and verify the results in Theorems 1.1.1 and 1.1.2.

1.1.3 Linear equations

By a linear equation over eld K, we mean an expression of the form a1 x1 + a2 x2 + · · · + an xn = b,

where ai , b ∈ K and the xi 's are indeterminants (or unknowns or variables ). The scalars ai are

called the coecients of the xi respectively and b is called the constant term or simply the constant
 Section 1.1. Matrices Page 8

of the equation. A set of values of the unknowns x1 = k1 , x2 = k2 , . . . , xn = kn is a solution of

a1 x1 + a2 x2 + · · · + an xn = b if the statement a1 k1 + a2 k2 + · · · + an kn = b is true. This set of values

is then said to satisfy the equation. If there is no ambiguity about the position of the unknowns in

the equation, then we denote this solution by an n-tuple u = (k1 , k2 , . . . , kn ) ∈ K n . For example,

if we consider the equation x + 2y − 4z + w = 3, then the 4-tuple (3, 2, 1, 0) is a solution to the it

(why?), while or however, the 4-tuple (1, 2, 4, 5) is not a solution to the same equation (why?).

The following system of linear equations


a11 x1 + a12 x2 + · · · + a1n xn = b1







a21 x1 + a22 x2 + · · · + a2n xn

= b2
(1.1)
. .
.
. = ..







am1 x1 + am2 x2 + · · · + amn xn

= bm

is equivalent to the matrix equation

   
a11 a12 · · · a1n x1 b
     1
a21 a22 · · · a2n   x2   b2 
    
 .. . . .  .  =  . 
    
 . . . .  .   .. 
. . . .
    
a11 a12 · · · a1n xn bn

or simply AX = B , where A = (aij ), X = (xi ) and B = (bi ). That is every solution of the system

is a solution of the matrix equation and vice versa. A solution (p articular solution ) to AX = B is

an n-tuple u = (k1 , . . . , kn ) of scalars in the eld K that satises each of the equations in Equation

(1.1); the set of all such solutions is called the solution set or the general solution.
Observe that the associated homogeneous system of is then equivalent to AX = 0. This system

always has a solution, namely the n-tuple 0 = (0, . . . , 0) called the zero or trivial solution. Any

other solution if it exists is called a nonzero or nontrivial solution. The above matrix A is called
thecoecient matrix of the system, X is the unknown matrix (also indeterminants or variables
matrix ), B , the constant matrix and (A|B) is the augmented matrix. Observe that the system is

completely determined by its augmented matrix. In studying linear equations it is usually simpler

to use the language and theory of matrices as indicated by the following theorem.

1.1.4 Theorem. Suppose u1 , u2 , . . . , un are solutions of a homogeneous system of linear equations

AX = 0. Then every linear combination of the ui 's is also a solution to AX = 0.
 Section 1.1. Matrices Page 9

Proof. A( ni=1 ki ui ) =
P Pn
Given that Aui = 0. Hence i=1 ki Aui = 0. Accordingly, the linear

combination k1 u1 + k2 u2 + · · · + kn un is a solution of the homogeneous system AX = 0

1.1.5 Theorem. Suppose the eld K is innite (if K is the real eld R or the complex eld C).
Then the system AX = B has no solution, a unique or an innite number of solutions.

Proof. It suces to show that if AX = B has more than one solution, the it has innitely many.

Suppose that u and v are distinct solutions of AX = B , i.e., Au = B and Av = B . For any k ∈ K,
A(u + k(u − v)) = Au + k(Au − Av) = B + k(B − B) = 0. For each k ∈ K , u + k(u − v) is a solution
to AX = B . Since all such solutions are distinct, AX = B has an innitely many solutions.

In the following problem we deduce the conditions that must be placed on a, b, and c so that the

following system of equations has no solutions, a unique solution, innitely many solutions,

x + 2y − 3z = a

2x + 6y − 11z = b

x − 2y + 7z = c.

Proof. We postpone the solution of this result till after row reductions. It is easier there.

A matrix A = (aij ) is an echelon matrix matrix, or an echelon form, if the number of zeros preceding
the rst nonzero entry of a row increases row by row until only the zero rows remain, i.e., if there

exist nonzero entries a1j1 , a2j2 , . . . , arjr , where j1 < j2 < · · · < jr such that aij = 0 for i ≤ r, j < ji ,
and for i > r. We call distinguished elements of the echelon matrix A.
a1j1 , a2j2 , . . . , arjr the In

particular, an echelon matrix is a row reduced echelon matrix if the distinguished elements are:

1. the only nonzero entries in their respective columns;

2. each equal to 1.

     
2 3 2 0 4 5 −6 0
1 2 3 1 3 0 0 4 0
     
0 0 7 1 −3 2 0 0 0 1 0 −3 0
 0 0 4 0
  
,  ,  .
 

0 0 0 0 0 6 2 0 0 0 0 0 0 0 1 2 0
     
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
 Section 1.1. Matrices Page 10

The third matrix above is an example of a row reduced echelon matrix, the other two are not. The

zero matrix for any numbers of rows or columns is also a row reduced echelon matrix.

A matrix A is said to be row equivalent to a matrix B if B can be obtained from A by a nite

sequence of the following operations called elementary operations. A matrix obtained from an

identity matrix by a single elementary row operation is called an elementary matrix.

E1 . Interchange the ith row and the j th row: Ri ↔ Rj .

E2 . Multiply the ith row by a nonzero scalar: Ri → kRi , k 6= 0.

E3 . Replace the ith row by k times the j th row plus the ith row: Ri → kRj + Ri . In actual

practice, we apply E2 and then E3 in one step, i.e., the step E: Replace the ith row by k1
times the j th row plus k2 (nonzero scalar) times the ith row: Ri → k1 Rj + k2 Ri , k2 6= 0.

The 3-square elementary matrices corresponding to the operations: R1 ↔ R2 , R3 → −7R3 and

R2 → −3R1 + R2 . Applying the above operations to the identity matrix I3 yields

     
0 1 0 1 0 0 1 0 0
     
E1 = 
1 0 0 , E2 = 0 1 0  , E3 = −3 1 0 .
    

0 0 1 0 0 −7 0 0 1

The reader no doubt recognises the similarity of the above operations and those used in solving

systems of linear equations. The following algorithm is also similar to the one used for linear

equations. The algorithm below describes steps needed to row reduce a matrix to its echelon form :

Step 1. Suppose the j1 column is the rst column with a nonzero entry. Interchange the rows

so that this nonzero entry appears in the rst row, that is so that a1j1 6= 0.

Step 2. For each i > 1, apply the operation Ri → −aij1 R1 + a1j1 Ri .

Step 3. Repeat Steps 1 and 2 with the submatrix formed by all the rows excluding the rst.

Continue the process until the matrix is in echelon form.

The term row reduce shall mean to transform by elementary row operations. Suppose A = (aij )
is a matrix in echelon form with distinguished elements a1j1 , a2j2 , . . . , arjr . Apply the operations

Rk → −akji Ri + aiji Rk , k = 1, . . . , i − 1 for i = 2, then i = 3, . . . , i = r. Thus, A is replaced

by an echelon matrix whose distinguished elements are the only nonzero entries in their respective
 Section 1.1. Matrices Page 11

columns. Next, multiply Ri by a−1

iji , i ≤ r . Thus, in addition, the distinguished elements are each

1. In other words, the process row reduces an echelon matrix to one in row reduced echelon form.
Any arbitrary matrix is row equivalent to atleast one row reduced echelon matrix. In fact, it can be

shown that A is row equivalent to only one such a matrix, we call it the row canonical form of A.

Consider the matrix

 
1 −2 3 −1
 
A=
2 −1 2 2 .

3 1 2 3

We demonstrate how to reduce A to echelon form and then its row canonical form.

     
1 −2 3 −1 1 −2 3 −1 1 −2 3 −1
     
A=
2 −1 2 2
0 3 −4 4 
 
0
 3 −4 .
4 
3 1 2 3 0 7 −7 6 0 0 7 −10

The last matrix is in echelon form. We further row reduce it to obtain its row canonical form.

       
15
1 −2 3 −1 3 0 1 5 21 0 0 45 1 0 0 7
       
0
 3 −4 4 

0 3 −4
 4 

0
 21 0 −12

0 1 0
 − 74 
.
0 0 7 −10 0 0 7 −10 0 0 7 −10 0 0 1 − 10
7

Suggest another method one would use to obtain this row canonical form (and give its shortcomings).

Try out the following problems.

1. Consider the linear system ax + by = e, and cx + dy = f in x and y over R: Show that:

a b de−bf af −ce
(a) if
c 6= d , then the system has a unique solution x= ad−bc and y= ad−bc .
a b e
(b) if
c 6= d 6= f , then the system has no solution.
a b e
(c) if
c 6= d = f , then the system has more than one solution.

(d) give the geometric interpretation of each of the above cases.

2. Determine which of the following system of linear equations is inconsistent, consistent and
 Section 1.1. Matrices Page 12

unique, consistent with innitely many solutions.





2x + y − 2z + 3w =1

3x + 2y − z + 3w =4, (1.2)


3x + 3y + 3z − 3w = 5


x + 2y − 3z =4







x + 3y + z

= 11
, (1.3)
2x + 5y − 4z = 13







2x + 6y + 2z = 22





x + 2y − 2z + 3w =2

2x + 4y − 3z + 4w =5 . (1.4)


5x + 10y − 8z + 11w

= 12

3. Do the same for

  



2x − 3y + 6z − 5w =3 


x + 2y − 3z = −1 


2x + y − 2z = 10
  
y − 4z + v =1 , 3x − y + 2z =7 ,
3x + 2y + 2z =1 .

 
 

v − w

= 2. 5x + 3y − 4z

=2

5x + 4y + 3z =1

4. Determine the values of k such that the system in unknowns x, y and z has

(a) unique solution

(b) no solution

(c) more than one solution

 


kx + y + z =1  

x + y + kz =2

 z + 2y + kz

=1


x + ky + z =1, , 3x + 4y + 2z =k.


 2x + ky + 8z

=3




x + y + kz =1 2x + 3y − z

=1

Let {u1 , u2 , . . . , ur } be a set of nonzero vectors in K n, (or an arbitrary vector space V over K ).
These vectors are linearly independent over K if the linear combination α1 u1 +α2 u2 +· · ·+αn un = 0
implies α1 = α2 = · · · = αn = 0. If there exist some non-trivial solution α1 , α2 , . . . , αn such that
 Section 1.1. Matrices Page 13

α1 u1 + α2 u2 + · · · + αn un = 0, then the vectors {u1 , u2 , . . . , ur } are linearly dependant over K.

For example, the vectors {(1, 0), (0, 1)} are linearly independent over R, since 0 = (0, 0) = x(1, 0) +
y(1, 0) = (x, y) implies x = y = 0. However, the set {(1, 0), (2, 0)} is linearly dependent over K.

1. Determine whether the following vectors are linearly dependent or independent over R.

(a) u = (1, 1, −1), v = (2, −3, 1) and w = (8, −7, 1).

(b) u = (1, −2, −3), v = (2, 3, −1) and w = (3, 2, 1).

(c) u = (a1 , a2 ), v = (b1 , b2 ) and w = (c1 , c2 ).

2. Trial problems. Do the same as above.

(a) u = (1, 3, −1), v = (2, 0, 1) and w = (1, −1, 1).

(b) u = (1, 1, −1), v = (2, 1, 0) and w = (−1, 1, 2).

(c) u = (1, −2, 3, 1), v = (3, 2, 1, −2) and w = (1, 6, −5, −4).

Let A be an arbitrary m × n matrix over K . The row space of A is the subspace of K n generated by

its rows and the column space of A is the subspace of K m generated by its columns. The dimensions

of the row and the column space of A A respectively.

are called the row rank and column rank of
 
1 2 3
 
Compute the null space, dimension of the null space, row space and the row rank of −2 −3 −4.
 
7 12 17

Proof. 1

1.1.6 Theorem. The row rank and the column rank of a matrix are equal.

The rank of the matrix A, written Rank(A), is the common value of its row rank and column

matrix. Therefore, the rank of a matrix gives the maximum number of independent rows and also

the maximum number of independent columns. To obtain, the row rank of a matrix, reduce it to

echelon form. Since row equivalent matrices have the same row space, the non zero rows of the

echelon matrix (are independent) form the basis of the row space. For column rank, do the same

use column operations or equivalently, transpose the matrix and then proceed as above. The rank

of the null space is given by obtaining the solution space, and the number of independent variables

generating the solution space of the matrix. This satises the Rank-Nullity Theorem.
 Section 1.1. Matrices Page 14

1. Determine whether the following matrices

(a)

 
    1 −1 −1
1 1 5 1 −1 −2  
P = , Q= , and R= 4 −3 −1,
2 3 13 3 −2 −3
 
3 −1 3

have the same row space.

(b) have the same column space.

2. Row reduce the following matrices to

(a) echelon form

(b) row canonical form

     
1 2 −3 0 2 3 4 5 6 0 1 3 −2
     
A=
 2 4 −2 2 
 and B=
0 0 3 2 5 
 and C=
2 1 −4 3 .

3 6 −4 3 0 0 0 0 2 2 3 3 −1

3. Given

 
1 −1
2 0
 
A=
2 6 −3 −3 .


3 10 −6 −5

Compute

(a) the row space and row rank of A

(b) the null space and the rank of the null space of A

(c) the column space and row rank of A

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 order n and is called an n-square matrix. The diagonal (or main

diagonal) of the main square matrix A = (aij ) consists of the elements a11 , a22 , . . . , ann . An upper
triangular matrix or simply a triangular matrix is a square matrix whose entries below the main

diagonal are all zero. Similarly, a lower triangular matrix is a square matrix whose entries above
 Section 1.1. Matrices Page 15

the main diagonal are all zero. A diagonal matrix is a square matrix whose non-diagonal entries are

all zeroes. In particular, the n-square matrix with 1's on the diagonal and 0's else where denoted

by In or simply I is the called the unit or identity matrix. This matrix is similar to the scalar 1 in

that for any n-square matrix A, AI = IA = A. The matrix kI for a scalar k∈K is called a scalar

matrix, it is a diagonal matrix whose all diagonal entries are each k.

Recall that not every two matrices can be added or multiplied. However, if we consider square

matrices of some given order n, then this inconvenience disappears. Specically, the operations of

addition, multiplication, scalar multiplication and transpose can be performed on any n×n matrices
and the result is again an n×n matrix. In particular, if A is any n-square matrix, we can form

powers of A: A2 = AA, A3 = A2 A, . . . and A0 = I . We can also form polynomials in the matrix A:

for any polynomial, f (x) = a0 + a1 x + · · · + an xn , where the ai 's are scalars in K; we dene f (A)
to be the matrix f (A) = a0 I + a1 A + · · · + an An . In the case that f (A) = 0, then A is called a zero

or root of the polynomial f (x). So the set of n-square matrices over K forms a matrix algebra.

 
1 2
1. Let A= . Show that A is a zero f (x) = x2 + 3x − 10 but not g(x) = 2x2 − 3x + 5.
3 −4

A square matrix A is said to be invertible if there exists a matrix B with the property that AB =
I = BA, where I is the identity matrix. Such a matrix B is unique; for AB1 = I and AB2 = I
implies that B1 = B1 I = B1 (AB2 ) = (B1 A)B2 = IB2 = B2 . We call such a matrix B, the inverse

of A and denote it by A−1 . Observe that the above relation is symmetric, that is if B is the inverse

of A, then A is the inverse of B. For square matrices, AB = I if and only if BA = I , hence it is

necessary to test only one product to determine whether two given matrices are inverses.

1. Suppose A is invertible and say, it is row reducible to the identity matrix I by the sequence

of elementary operations e1 , . . . , en .

(a) Show that this sequence of operations applied to I yields A−1 .

(b) Use this result to show obtain the inverse of

 
1 0 2
 
A=
2 −1 3 .


4 1 8

In case the nal block of (A, I) is not of the form (I, B), then the given matrix is not row

equivalent to I and so is not invertible.

 Section 1.2. Determinants Page 16

1.2 Determinants

To every square matrix A over a eld K, there is a specic scalar assigned called the determinant

of A; it is usually denoted by det(A) or |A|. This determinant function (a functional) was rst

discovered in the investigation of system of linear equations. Today, this is an indispensable tool

in investigating and obtaining properties of a linear operator. We shall begin this section with a

discussion of permutations, which is necessary for the denition of the determinant.

1.2.1 Permutations

Let X = {1, 2, . . . , n}. A permutation

 on X is any bijection on
 X. We denote permutations by

1 2 ··· n
lower case greek letters, e.g., σ =  or σ = j1 j2 . . . jn where σ(i) = ji . Since σ
j1 j2 · · · jn
is a bijection. The sequence j1 j2 . . . jn is simply a re-arrangement of the numbers 1, 2, . . . , n. The
number of such permutations is n! and that the set of them is denoted by Sn . We remark that if

σ∈ Sn , the inverse mapping σ −1 ∈ Sn ; and if σ, τ ∈ Sn , then the composition mapping σ ◦ τ ∈ Sn .

In particular, the identity mapping ι = σ ◦ σ −1 ∈ Sn . In fact ι = 12 . . . n. In a fancy way, we say

Sn forms a group under composition, and in fact it is called the Symmetric group on n elements.

Consider an arbitrary permutation σ ∈ Sn : σ = j1 j2 . . . jn . We say σ is even or odd according

to whether there is an even or odd number of pairs (i, k) for which i>k but i precedes k in σ.
For example: the permutation σ = 35142 ∈ S5 . The set {(3, 1), (3, 2), (5, 1), (5, 4), (5, 2), (4, 2)} has

even cardinality and so the permutation σ is even. The identity permutation ι = 12 . . . n is even.

In S3 , the permutations 123, 231, 312 are even whereas 132, 213, 321 are odd. The sign or parity of

σ written sgn(σ) is given by sgn(σ) =1 if σ is even and sgn(σ) = −1 if σ is odd.

A transposition τ is a permutation that interchanges two numbers i and j and xes the others:

τ (i) = j , τ (j) = i and τ (k) = k for all k 6= i, j . It is denoted by τ = (ij). For example, in S4 ,
τ = 21 is a transposition. It is not hard to show that a transposition is an odd permutation.

1.2.2 Determinants

Let A = (aij ) be an n-square matrix over a eld K. The determinant of the n-square matrix
P
A = (aij ), denoted by det(A) or |A| is the sum det(A) = |A| = σ sgn(σ)a1j1 a2j2 . . . anjn over all
P
permutations σ = j1 j2 . . . jn in Sn , i.e., |A| = σ∈Sn sgn(σ)a1σ(1) a2σ(2) . . . anσ(n) . The determinant
of the n-square matrix A is said to be of order n and is frequently denoted by |(aij )|. We emphasize
 Section 1.2. Determinants Page 17

that a square array of scalars enclosed by straight lines is not a matrix but rather the scalar that

the determinant assigns to the matrix formed by the arrays of scalars. Examples include:

1. Let A = (a11 ), then |A| = a11 . (since the one permutation in S1 is even)
 
 
a11 a12 
2. In S2 , the permutation 12 is even while 21 is odd. Hence   = a11 a22 − a12 a21 .
a21 a22 
 

3. Recall in S3 , the permutations 123, 231, 312 are even whereas 132, 213, 321 are odd. So

 
a a a 
 11 12 13 
 
a21 a22 a23  = a11 a22 a33 + a12 a23 a31 + a13 a21 a32 − a13 a22 a31 − a12 a21 a33 − a11 a23 a32 .
 
 
a31 a32 a33 

We will obtain this formula using other methods like cofactor expansion or Laplace expansions.

As n increases the number of terms in the determinant becomes astronomical. Accordingly, we use

indirect methods to evaluate determinants rather than the denition. In fact we prove a number

of properties about determinants which will permit us to shorten computation considerably. In

particular, we show that a determinant of order n is equal to a linear combination of determinants

of order n−1 as in case n=3 above. We now list some basic properties of the determinant.

1.2.1 Theorem. |A| = |AT |.

By this theorem, any theorem about the determinant of a matrix which concerns the rows of A will

have an analogous theorem concerns columns of A. This is like some sort of duality in the results.

The next theorem gives certain cases for which the determinant can be obtained immediately.

1.2.2 Theorem. Let A be a square matrix.

1. If A has a row (column) of zeros, then |A| = 0.

2. If A has two identical rows (columns), then |A| = 0.

3. If A is triangular, i.e., A has zeros above or below the diagonal, then the determinant of A
product of diagonal elements. Thus in particular, |I| = 1, where I is the identity matrix.

The next theorem shows how the determinant of a matrix is aected by the elementary operations".

1.2.3 Theorem. Let B be the matrix obtained from a matrix A by

 Section 1.2. Determinants Page 18

1. multiplying a row (column) of A by a scalar k; then |B| = k|A|.

2. interchanging two rows (columns) of |A|; then |B| = −|A|.
3. adding a multiple of a row (column) of A to another; then |B| = |A|.

We now state two of the most important and useful theorems on determinants.

1.2.4 Theorem. Let A be an n-square matrix. Then the following are equivalent.

1. A is invertible, i.e., A has an inverse A−1 .

2. A is nonsingular, i.e., AX = 0 has only the zero solution, or rank(A) = n, or the rows
(columns) of A are linearly independent.

3. |A| =
6 0.

1.2.5 Theorem. The determinant function is a completely multiplicative function.

It is almost trivial that if P is invertible, then |P −1 | = |P |−1 .

1. Compute the determinant(s) of

   
5 4 2 1 2−3 −2 5
   
2 3 1 −2  −2 −3 2 −5
A= , and B= ,
   
−5 −7 −3 9  1 3 −2 2 
   
1 −2 −1 4 −1 −6 4 3

by

(a) row reduction to echelon form (Gaussian elimination)

(b) row reduction to row reduced echelon form (Gauss-Jordan elimination)

(c) denition of determinant using permutations

1.2.3 Minors and Cofactors

Consider an n-square matrix A = (aij ). Let Mij denote the (n − 1)-square submatrix of A obtained
by deleting its ith row and j th column. The determinant |Mij | is called the minor of the element

aij of A, and dene the cofactor of aij , denoted by Aij to be the signed minor: Aij = (−1)i+j |Mij |.
Note that the signs" (−1)i+j accompanying the minors form a chessboard pattern with +'s on the

main diagonal. We emphasize that Mij denoted a matrix whereas Aij denoted a scalar.
 Section 1.2. Determinants Page 19

1.2.6 Theorem. The determinant of the matrix A = (aij ) is equal to the sum of the products
obtained by multiplying the elements of any row (column) by their respective cofactors.
n
X n
X
|A| = aij Aij = aij Aij = |A|.
i=1 j=1

The above formulas, called the Laplace expansions of the determinant of A by the ith row and

the j th column respectively, oer a method of simplifying the computation of |A|, i.e., by adding a

multiple of a row (column) to another row (column) we can reduce A to a matrix containing a row

or column with one entry 1 and the others are 0. Expanding by this row or column reduces the

computation of |A| to the computation of a determinant of order one less than that of |A|.

 
2 1 1
 
1. For the matrix 0 5 −2, nd the cofactor of
 
1 −3 4

(a) the entries 5 and −2 and deduce its determinant.

(b) Verify this answer using any other techniques.

2. Compute the determinant of of each matrix:

     
1 2 3 2 0 1 2 0 1
     
4 −2 3  , 2 −3, .
2 −3
3 3
   
2 5 −1 −1 −3 5 −1 −3 5

3. Evaluate the determinant of A by all the methods you know of !

     
2 5 −3 −2 −2 −5 −4
3 1 2 −2 3
     
−2 −3 2 −5 −5 2 8 −5 3 −1 5 0 
, , .
     
  
1
 3 −2 2 

−2 4
 7 −3

4 0
 2 1 

−1 −6 4 3 2 −3 −5 8 1 7 2 −3

Consider an n-square matrix A = (aij ). The transpose of cofactors of the elements aij of A, denoted
by adj(A) is called the classical adjoint of A. We say classical adjoint" instead of simply adjoint".

1.2.7 Theorem. For any square matrix A, A · adj(A) = adj(A) · A = |A|I . Therefore, if |A| =
6 0,

1
A−1 = adj(A).
|A|
 Section 1.2. Determinants Page 20

 
1 2 3
 
Consider the matrix A=
2 3 4.


1 5 7

1. Compute |A|.

2. Find adj(A).

3. Verify that A · adj(A) = |A|I .

4. Find A−1 .

The above theorem gives an important method for obtaining the inverse of a given matrix.

1.2.8 Theorem. A linear system AX = B has a unique solution if and only if |A| =
6 0.

The above theorem gives us Crammer's rule" for solving systems of linear equations. We emphasize

that the theorem only refers to a system with the same number of equations and unknowns and

that it only gives a solution when ∆ 6= 0. In fact, if ∆ = 0, the theorem does not tell us whether or

not the system has a solution. However, in the case of homogeneous system we get Theorem 1.2.9.

Proof. We know that AX = B has a unique solution if and only if A is invertible, and A is

invertible if and only if |A| =

6 0. Now suppose A is invertible, multiplying AX = B by A−1 , we

obtain X = A−1 AX = 1
|A| adj(A)B . It follows that xi = 1
|A| (b1 A1i + b2 A2i + · · · + bn Ani ) = |A|i , the

determinant of the matrix obtained by replacing the ith column of A by the column vector B.

1.2.9 Theorem. The homogeneous system AX = 0 has a nonzero solution if and only if |A| = 0.

We remark that the preceding theorem is of interest more for theoretical and historical reasons than

for practical reasons. The previous method of solving systems of linear system of equations, i.e., by

reducing a system to echelon form, is usually much more ecient than by using determinants.

 
1 2 3
 
1. (a) Consider the matrix A=
 2 3 4 .

1 5 7
i. Compute |A|.
ii. Find adj(A).

iii. Verify that A · adj(A) = |A|I .

Section 1.2. Determinants Page 21

iv. Find A−1 .

(b) Solve the following system of linear equations using determinants.

 



x + 2y + 3z =a 


x + 2y + 3z =0
 
2x + 3y + 4z =b , 2x + 3y + 4z =0

 


x + 5y + 7z 
=c x + 5y + 7z =1
References

[1] S. Lipschutz. Theory and Problems of Linear Algebra. Schaum's Outline Series - McGraw-Hill

Book Company, 1968.

[2] J. Stewart. Calculus - Early Transcendentals. Thomson Brooks / Cole, 2008.

[3] K. Stroud. Engineering Mathematics. Palgrave Macmillan, 2005.

22