Debre Markos University

Department of mathematics

Applied mathematics for health informatics

Chapter One

1. Matrices and Determinants

1.1. Definition of Matrix

The trace/idea of matrices started back in the 4th century BC by the Babylonians.
Officially, matrices began in the 2nd century BC with the Chinese. However, it was not until
near the end of the 17th Century that the ideas reappeared and development really got
underway.

The Babylonians studied problems which lead to simultaneous linear equations and some
of these are preserved in clay tablets which survive. For example, a tablet dating from around
300 BC contains the following problem: - There are two fields whose total area is 1800 square
yards. One produces grain at the rate of 2/3 of a bushel per square yard while the other
produces grain at the rate of 1/2 a bushel per square yard. If the total yield is 1100 bushels,
what is the size of each field.

The Chinese, between 200 BC and 100 BC, came much closer to matrices than the
Babylonians. Indeed, it is fair to say that the text Nine Chapters on the Mathematical
Art written during the Han Dynasty gives the first known example of matrix methods. It is
one of the ten most important arithmetic books of ancient China. First a problem is set up
which is similar to the Babylonian example given above: - There are three types of corn, of
which three bundles of the first, two of the second, and one of the third make 39 measures. Two
of the first, three of the second and one of the third make 34 measures. And one of the first, two
of the second and three of the third make 26 measures. How many measures of corn are
contained of one bundle of each type?

1
Now the author does something quite remarkable. He sets up the coefficients of the system
of three linear equations in three unknowns as a table on a 'counting board'. See(a)

(a) (b) (c)

Most remarkably the author, writing in 200 BC, instructs the reader to multiply the middle
column by 3 and subtract the right column as many times as possible, the same is then done
subtracting the right column as many times as possible from 3 times the first column, which
gives result (b). Next the left most column is multiplied by 5 and then the middle column is
subtracted as many times as possible (C) from which the solution can be found for the third
type of corn, then for the second, then the first by back substitution. This method, now known
as Gaussian elimination, would not become well known until the early 19th Century.

Definition: A rectangular arrangement of mn numbers (real or complex) in to m horizontal

rows and n vertical columns enclosed by a pair of brackets [ ] or () such as
𝑎11 𝑎12 . . . 𝑎1𝑛 𝑎11 𝑎12 . . . 𝑎1𝑛
𝑎21 𝑎22 . . . 𝑎2𝑛 𝑎21 𝑎22 . . . 𝑎2𝑛
. . . . . .
or
. . ⋱ . . . ⋱ .
. . . . . .
[𝑎𝑚1 𝑎𝑚1 . . . 𝑎𝑚𝑛 ] (𝑎𝑚1 𝑎𝑚1 . . . 𝑎𝑚𝑛 )
is called an 𝑚 × 𝑛 (read “m by n”) matrix or a matrix of order 𝑚 × 𝑛.
matrices are denoted by single capital letters such as A, B, C and so on. The numbers
(𝑎𝑖𝑗 ) in a matrix are called the entries or the elements of the matrix.
For the entry 𝑎𝑖𝑗 , the first subscript i specify the row and the second subscript j the
column
in which the entry appears.

2
That is, 𝑎𝑖𝑗 is an element of matrix A which is located in the 𝑖 𝑡ℎ row and 𝑗 𝑡ℎ column of
the
matrix A. It is customary to abbreviate the matrix
𝑎11 𝑎12 . . . 𝑎1𝑛
𝑎21 𝑎22 . . . 𝑎2𝑛
. . .
𝐴=
. . ⋱ .
. . .
[𝑎𝑚1 𝑎𝑚2 . . . 𝑎𝑚𝑛 ]
by the symbol (𝑎𝑖𝑗 )𝑚×𝑛 or more simply (𝑎𝑖𝑗 ). This notation merely indicates what
type of symbols we are using to denote the general entry.

The order of a matrix is the number of rows and columns it has. When we say a
matrix is a 3 by 4 matrix, we are saying that it has 3 rows and 4 columns. The rows
are always mentioned first and the columns second. This means that a 3  4 matrix
does not have the same order as a 4  3 matrix. It must be noted that even though an
𝑚 × 𝑛 matrix contains mn elements, the entire matrix should be considered as a
single entity.

Remark: The size of a matrix or the dimension of a matrix has equivalent meaning
with the order of the matrix.

1 5 2
Example: Let 𝐴 = [ ].
0 3 6
Solution: Since A has 2 rows and 3 columns, we say A has order 2 × 3, where the
number of rows is specified first. The element 6 is in the position a23 (read ‘a two
three’) because it is in row 2 and column 3.
−1 4 7
Example: What is the value of 𝑎23 and 𝑎32 in 𝐴 = [ 2 3 1] ?
5 7 8
Solution: 𝑎23 , the element in the second row and third column, is 1 and 𝑎32 , the
element in the third row and second column, is 7. What is the size of
this matrix?
Activity: 1. Suppose A is a 5x7 matrix, then

3
 a. A has 7 rows. (True/False)
b. 𝑎𝑖𝑗 is an element of A for i = 6 and j = 4.(True/False)
c. For what values of i and j, 𝑎𝑖𝑗 is an element of A?
4 8 −1
4 −7 5
2. Suppose 𝐴 = [ ]and 𝐵 = [−7 1 −5]
8 1 6
5 6 7
a. What is the order of A and B?
b. Find 𝑎22 , 𝑎13 , 𝑏13 and 𝑏31 .

Example: Form a 4 by 5 matrix, B, such that bij = i + j.

Solution: Since the number of rows is specified first, this matrix has four rows and
five columns.
𝑏11 𝑏12 𝑏13 𝑏14 𝑏15 2 3 4 5 6
𝑏21 𝑏22 𝑏23 𝑏24 𝑏25 3 4 5 6 7
𝐵=[ ]=[ ].
𝑏31 𝑏32 𝑏33 𝑏34 𝑏35 4 5 6 7 8
𝑏41 𝑏42 𝑏43 𝑏44 𝑏45 5 6 7 8 9
Activity:1. Form a 4 by 3 matrix, B, such that
a) 𝑏𝑖𝑗 = 𝑖 × 𝑗 b) 𝑏𝑖𝑗 = (−1)𝑖+𝑗
𝑒 𝑖+𝑗 , 𝑖 < 𝑗
2. form a 4 by 4 matrix, B, such that 𝑏𝑖𝑗 = { ln(2𝑖 + 𝑗) , 𝑖 = 𝑗
cos(𝑖 + 𝑗)𝜋, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

Definition: Two matrices A and B are said to be equal, written A = B, if they are of
the same order and if all corresponding entries are equal.
5 1 0 2+3 1 0 9
For example, [ ]=[ ] but[9 2] ≠ [ ]. Why?
2 3 4 2 3 2×2 2
𝑥+𝑦 6 1 𝑧
Example: Given the matrix equation[ ]=[ ]. Find the values of x, y
𝑥 − 𝑦 12 3 2𝑧
and z.

Solution: By the definition of equality of matrices,

𝑥 + 𝑦 = 1, 𝑥 − 𝑦 = 3, z=6 and 2z=8.
solving the above gives x = 2 ,y = -1 and z=6.
Activity: Find the values of x, y, z and w which satisfy the matrix equation

4
 𝑥−𝑦 2𝑥 + 𝑧 −1 5
a. [ ]=[ ]
2𝑥 − 𝑦 3𝑧 + 𝑤 0 13
𝑥 + 3 2𝑦 + 𝑥 0 −7
b. [ ]=[ ]
𝑧 − 1 4𝑤 − 6 3 2𝑤
1.2. Types of Matrices
Classification of matrices do have some unique characteristics; it may be based on
dimension (square, row, column) and elements (zero, diagonal, identity, lower
triangular, upper triangular….).
Row Matrix:
A matrix that has exactly one row is called a row matrix. Denoted by 𝐴 = (𝑎1𝑗 ), 𝑗 =
1,2, … … … , 𝑛.
For example, the matrix 𝐴 = [5 2 −1 4] is a row matrix of order1 × 4.
Column Matrix: A matrix consisting of a single column is called a column matrix.
Denoted by 𝐴 = (𝑎𝑖1 ), 𝑖 = 1,2, … … … , 𝑛.
3
For example, the matrix 𝐵 = [1]is a 3 × 1 column matrix.
4
Zero or Null Matrix: A matrix whose entries are all 0 is called a zero or null matrix.
0 0 0 0
It is usually denoted by 0𝑚×𝑛 or more simply by 0. For example, 𝟎 = [ ]
0 0 0 0
is a 2 × 4 zero matrix.
Square Matrix: An 𝑚 × 𝑛 matrix is said to be a square matrix of order n if m = n.
That is, if it has the same number of columns as rows.
−3 4 6
2 −1
For example, [ 2 1 3 ]and [ ] are square matrices of order 3 and 2
5 6
5 2 −1
respectively.
In a square matrix 𝐴 = (𝑎𝑖𝑗 ) of order n, the entries 𝑎11 , 𝑎22 , . . . , 𝑎𝑛𝑛 which lie on the
diagonal extending from the left upper corner to the lower right corner are called the
3 2 4
Principal diagonal entries. Thus, in the matrix C =[1 6 0] the entries 𝑐11 =
5 1 8
3, 𝑐22 = 6 and 𝑐33 = 8 constitute the principal diagonal.
Note:

5
 a. The sum of the entries on the main diagonal of a square matrix A of order n is
called the trace of A, denoted by 𝒕𝒓(𝑨)and define by 𝒕𝒓(𝑨) = ∑𝑛𝑖=1 𝑎𝑖𝑖 .
b. Suppose 𝑘, 𝑘1 , 𝑘2 , 𝑘3 … . 𝑘𝑚 ∈ ℝ and 𝐴, 𝐵, 𝐴1 , 𝐴2 , 𝐴3 , … 𝐴𝑚 are each square
matrix. Then the following statements are true.
i. 𝑡𝑟𝑎(𝐴) = 𝑡𝑟(𝐴𝑇 );
ii. 𝑡𝑟(𝑘𝐴) = 𝑘𝑡𝑟(𝐴);
iii. 𝑡𝑟(𝐴1 + 𝐴2 + 𝐴3 + ⋯ + 𝐴𝑚 ) = 𝑡𝑟(𝐴1 ) + (𝐴2 ) + (𝐴3 ) + ⋯ +(𝐴𝑚 )
iv. 𝑡𝑟(∑𝑚 𝑚
𝑖=1 𝑘𝑖 𝐴𝑖 ) = ∑𝑖=1 𝑘𝑖 𝑡𝑟(𝐴𝑖 )

v. 𝑡𝑟(𝐴𝐵) = 𝑡𝑟(𝐵𝐴) called cyclic property of trace.

vi. If 𝐴𝑚𝑥𝑛 = (𝑎𝑖𝑗 ), then 𝑡𝑟(𝐴𝑇 𝐴) = ∑𝑚 𝑛 2
𝑖=1 ∑𝑗=1 𝑎 𝑖𝑗 .

Activity: Find the trace of C in the above example.

Triangular Matrix: A square matrix is said to be an upper (lower) triangular matrix
if all entries below (above) the main diagonal are zeros.
5 0 0 0
2 4 8
1 3 0 0
For example, [0 1 2 ] and [ ]are upper and lower triangular
6 1 2 0
0 0 −3
−2 −4 8 6
matrices, respectively.
Diagonal Matrix: A square matrix is said to be diagonal if each of the entries not
falling on the main diagonal is zero. Thus, a square matrix 𝐴 = (𝑎𝑖𝑗 ) is diagonal if
𝑎𝑖𝑗 = 0 for 𝑖 ≠ 𝑗.
Activity: What about for i = j?
5 0 0
For example, [0 0 0] is a diagonal matrix.
0 0 7
Scalar matrix: A diagonal matrix with all the diagonal elements are equal is called a
scalar matrix.
2 0 0
For example, [0 2 0] is a scalar matrix.
0 0 2

6
Note: Let 𝐴 = (𝑎𝑖𝑗 ) be a square matrix. A is a scalar matrix if and only if 𝑎𝑖𝑗 =
𝑘, 𝑖𝑓 𝑖 = 𝑗
{ .
0, 𝑖𝑓 𝑖 ≠ 𝑗
Identity Matrix or Unit Matrix: A square matrix is said to be identity matrix or unit
matrix if all its main diagonal entries are 1’s and all other entries are 0’s. In other
words, a diagonal matrix whose all main diagonal elements are equal to 1 is called an
identity or unit matrix. An identity matrix of order n is denoted by In or more simply
by I.
1 0 0
1 0
Example, 𝐼3 = [0 1 0]is identity matrix of order 3. 𝐼2 = ( ) is identity matrix
0 1
0 0 1
of order 2.
Note: Let 𝐴 = (𝑎𝑖𝑗 ) be a square matrix. A is an identity matrix if and only if 𝑎𝑖𝑗 =
1, 𝑖𝑓 𝑖 = 𝑗
{
0, 𝑖𝑓 𝑖 ≠ 𝑗
1.3. Algebra of Matrices
1.3.1. Addition of matrices
Let A and B be two matrices of the same order. Then the addition of A and B, denoted
by A + B, is the matrix obtained by adding corresponding entries of A and B. Thus, if
𝐴 = (𝑎𝑖𝑗 )𝑚×𝑛 and 𝐵 = (𝑏𝑖𝑗 )𝑚×𝑛 , then 𝐴 + 𝐵 = (𝑎𝑖𝑗 + 𝑏𝑖𝑗 )𝑚×𝑛 .
Or

Remark: Notice that we can add two matrices if and only if they are of the same order.
If they are, we say they are conformable for addition. Also, the order of the sum of
two matrices is same as that of the two original matrices.
Activity: Given the matrices A, B, and C below

7
  1 2 4   2 -1 3   4 
A =  2 3 1 
 B =  2 4 2  C =  2 
 5 0 3   3 6 1   3 
Find, if possible. a) A + B b) B + C
If A is any matrix, the negative of A, denoted by –A, is the matrix obtained by replacing
each entry in A by its negative. For example, if
2 −1 −2 1
𝐴=[5 4 ], then −𝐴 = [ −5 −4]
−6 0 6 0
Properties of Addition of Matrices
1. Matrix addition is commutative. That is, if A and B are two matrices of the
same order, then A + B = B + A.
2. Matrix addition is associative. That is, if A, B and C are three matrices of the
same order, then (A + B) + C = A + (B + C).
3. Existence of additive identity. That is, if 0 is the zero matrix of the same order
as that of the matrix A, then A + 0 = A = 0 + A.
4. Existence of additive inverse. That is, if A is any matrix, then
A + (-A) = 0 = (-A) + A
Note: The zero matrix plays the same role in matrix addition as the number zero does
in addition of numbers.
Subtraction of Matrices: Let A and B be two matrices of the same order. Then by
A – B, we mean A + (-B). In other words, to find A – B we subtract each entry
of B from the corresponding entry of A.
4 −1 0 2
Example: Let 𝐴 = [2 3 ] and 𝐵 = [5 −2]
5 −7 6 1
4 − 0 −1 − 2 4 −3
Then 𝐴 − 𝐵 = [2 − 5 3 − (−2)] = [−3 5 ]
5 − 6 −7 − 1 −1 −8
1.3.2. Scalar Multiplication of a Matrix
Let A be an 𝑚 × 𝑛 matrix and k be a real number (called a scalar). Then the
multiplication of A by k, denoted by k A, is the 𝑚 × 𝑛 matrix obtained by multiplying
each entry of A by k. This operation is called scalar multiplication.
If 𝐴 = (𝑎𝑖𝑗 ) 𝑚𝑥𝑛 and 𝑘𝜖ℜ, then 𝑘𝐴 = (𝑘𝑎𝑖𝑗 ) 𝑚𝑥𝑛 .

8
 0 2 3 7 6 3
Example: If 𝐴 = [ ] 𝑎𝑛𝑑𝐵 = [ ]. Find 2A + 3B.
2 1 4 1 4 5
0 2 3 0 4 6 7 6 3 21 18 9
Solution: 2𝐴 = 2 [ ]=[ ]and 3𝐵 = 3 [ ]=[ ]
2 1 4 4 2 8 1 4 5 3 12 15
0 4 6 21 18 9 21 22 15
2𝐴 + 3𝐵 = [ ]+[ ]=[ ]
4 2 8 3 12 15 7 14 23
Properties of scalar multiplications
1. If A and B are two matrices of the same order and if k is a scalar, then
𝑘(𝐴 + 𝐵) = 𝑘𝐴 + 𝑘𝐵
2. If k1 and k2 are two scalars and if A is a matrix, then
(𝑘1 + 𝑘2 )𝐴 = 𝑘1 𝐴 + 𝑘2 𝐴
3. If k1 and k2 are two scalars and if A is a matrix, then
(𝑘1 𝑘2 )𝐴 = 𝑘1 (𝑘2 𝐴) = 𝑘2 (𝑘1 𝐴)
1.3.3. Multiplication of Matrices
While the operations of matrix addition and scalar multiplication are fairly
straightforward, the product AB of matrices A and B can be defined under the
condition that the number of columns of A must be equal to the number of rows of B. If
the number of columns in the matrix A equals the number of rows in the matrix B, we
say that the matrices are conformable for the product AB.
If matrix A is of dimension m  n and B of dimension n  p, the product AB will have
the dimension m  p. Let 𝐴 = (𝑎𝑖𝑗 ) be an 𝑚 × 𝑛 matrix and 𝐵 = (𝑏𝑗𝑘 ) be an 𝑛 × 𝑝
matrix. Then the product AB is the 𝑚 × 𝑝 matrix defined by 𝐴𝐵 = (𝑐𝑖𝑘 ), where
𝑐𝑖𝑘 = 𝑎𝑖1 𝑏1𝑘 + 𝑎𝑖2 𝑏2𝑘 +. . . +𝑎𝑖𝑛 𝑏𝑛𝑘 = ∑𝑛𝑗=1 𝑎𝑖𝑗 𝑏𝑗𝑘 , I = 1,2,…,m and k = 1,2,…,p
Thus, the product AB is the 𝑚 × 𝑝 matrix, where each entry 𝑐𝑖𝑘 of AB is obtained by
multiplying corresponding entries of the ith row of A by those of the kth column of B
and then finding the sum of the results.

 a 
Example: Given A = [ 2 3 4 ] and B =  b  , find the product AB.
 c 
Solution: The product is a 1  1 matrix whose entry is obtained by multiplying the
corresponding entries and then forming the sum.

9
  a 
AB = [ 2 3 4 ]  b  = [(2a + 3b + 4c)]
 c 
Note that AB is a 1  1 matrix, and its only entry is 2a + 3b + 4c.

 5 
Example: Given A = [ 2 3 4 ] and B =  6  , find the product AB.
 7 

 5 
Solution: AB = [ 2 3 4 ]  6  = [10 + 18 + 28] = [56]
 7 
 5 3 
Example: Given A = [ 2 3 4 ] and B =  6 4  , find the product AB.
 7 5

Solution: AB = [ 2 . 5 + 3 . 6 + 4 . 7 2.3 + 3.4+ 4.5 ]

= [56 38]
Note: it is possible to multiply two matrices with different dimensions as long as the
number of entries in the rows of the first matrix is the same as the number of entries
in columns of the second matrix.
Activity: 1. Given the matrices E, F, G and H, below
 1 2 
 2 -1   –3 
E =  4 2  F =  3 2  G = [ 4 1 ] H =  –1 
 3 1
Find, if possible. a) GH b) FH c) EF d) FE

2. Given the matrices E, F, G and H, below

 1 2 
   2 -1   –3 
E =  4 2  F =  3 2  G = [ 4 1 ] H =  –1 
 3 1
Find, if possible. a) GH b) FH c) EF d) FE
Remark: The definition refers to the product AB, in that order, A is the left factor
called pre-factor and B is the right factor called post factor.
1 −4
−2 4 1 6
Example: Find the product AB if 𝐴 = [5 3 ] and 𝐵 = [ ]
2 7 3 8
0 2

10
Solution: Since the number of columns of A is equal to the number of rows of B, the
product AB=C is defined. Since A is 3 × 2 and B is 2 × 4, the product AB will be 3 × 4
𝑐11 𝑐12 𝑐13 𝑐14
𝐴𝐵 = [ 21 𝑐22 𝑐23 𝑐24 ]
𝑐
𝑐31 𝑐32 𝑐33 𝑐34
The entry c11 is obtained by summing the products of each entry in row 1
of A by the corresponding entry in column 1 of B, that is.
𝑐11 = (1)(−2) + (−4)(2) = −10. Similarly, for C21 we use the entries in
row 2 of A and those in column 1 of B, that is C21 = (5) (-2) + (3) (2) = -4.
Also, C12 = (1) (4) + (-4) (7) = -24 C13 = (1) (1) + (-4) (3) = -11
C14 = (1) (6) + (-4) (8) = -26 C22 = (5) (4) + (3) (7) = 41
C23 = (5) (1) + (3 ) (3) = 14 C24 = (5) (6) + (3) (8) = 54
C31 = (0) (-2) + (2) (2) = 4 C32 = (0) (4) + (2) (7) = 14
C33 = (0) (1) + (2) (3) = 6 C34 = (0) (6) + (2) (8) = 16
1 0 1 2
Example: Let 𝐴 = [ ], and 𝐵 = [ ], then
0 0 1 0
1 2 1 0
𝐴𝐵 = [ ], 𝐵𝐴 = [ ]. Thus, 𝐴𝐵 ≠ 𝐵𝐴.
0 0 1 0
Activity: Which of the following are defined?
3 𝑎
i) (5 2 10) ( ) iii) ( ) (𝑐 𝑑)
−4 𝑏
𝑎
ii) (𝑎 𝑏 )(𝑐 𝑑) iv) (𝑎 𝑏 ) (𝑏 )
3 2 𝑎 𝑏
2. If 𝐴 = [ ] and 𝐵 = [ ], find a and b such that AB = BA.
4 1 3 5
Note:
2. AB = 0 does not necessarily imply A = 0 or B = 0.
3. AB = AC does not necessarily imply B = C.
4. If AB and BA are defined, then it does not imply that AB=BA.
1 −1 1 1 2 3 0 0 0
Example 1) Let 𝐴 = [−3 2 −1], 𝐵 = [2 4 6], then 𝐴𝐵 = [0 0 0].
−2 1 0 1 2 3 0 0 0

11
 1 −3 2 1 4 1 0
2) Let 𝐴 = [2 1 −3 ], 𝐵 = [ 2 1 1 1], 𝐶 =
4 −3 −1 1 −2 1 2
3 1 −1 −2 −3 −3 0 1
[3 −2 −1 −1],then 𝐴𝐵 = [ 1 15 0 −5] = 𝐴𝐶. But 𝐵 ≠ 𝐶.
2 −5 −1 0 −3 15 0 −5
Properties of Matrix multiplication
If A, B and C are any matrices, and if I is an identity matrix, then the following hold,
whenever the dimensions of the matrices are such that the products are defined.
A(BC) = (AB)C Associative Law

A.I = I.A =A (The order of I and A is the Multiplicative Identity

same) Law

A(B + C) = AB + AC Left Distributive Law

(A + B)C =AC+BC Right Distributive Law

A.0 = 0.A = 0 Multiplication by Zero

Remark:
1. For real numbers, a multiplied by itself n times can be written as an. Similarly, a
square matrix A multiplied by itself n times can be written as An. Therefore, A2
means AA, A3 means AAA and so on.
1 0
2. A matrix A is said to be idempotent if 𝐴2 = 𝐴. For example, 𝐴 = [ ] is
1 0
idempotent.
3. A square matrix A is said to be nilpotent if there exists a positive integer n such
that
𝐴𝑛 = 0. The least positive integer k for which 𝐴𝑘 = 0 is called the order of
nilpotency.
For example, if 𝐴 = [𝑎𝑖𝑗] is an n×n matrix with 𝑎𝑖𝑗 equal to 1 if i-j = 1 and 0,
otherwise
then An = 0 and 𝐴𝑙 = 0 for 1 ≤ l ≤ n - 1.

12
Exercise
1 2 3 4 5 6 −1 −2 1
1. If 𝐴 = [−1 0 2 ] , 𝐵 = [−1 0 1] and 𝐶 = [−1 2 3]
1 −3 −1 2 1 2 −1 −2 2
Find each of the following
(i) A + B ii) 2B – 3C
iii) A+B–C iv) A – 2B + 3C v) 2A – C
5 2 2 4 1 3
2. Let 𝐴 = [ ],𝐵 = [ ],𝐶 = [ ]
−1 3 6 1 7 2
Find the following: i) AB ii) BC iii) (AB)C iv) A(BC)
4 −1 −4
If 𝐴 = [4 0 −4], compute A2. Is it equal to I3, where I3 is the identity matrix of
3 −1 −3
order 3?
Activity

1. If A = and I = , find a and b so that A2 + aI = bA

2. If A = , then find the value of A2 –3A + 2I.

 − 1 − 4 1 − 2n − 4n 
3. Let A =   , prove by mathematical induction that: A n = 
1 3  n 1 + 2n 

.
1.3.4. Transpose of a matrix
Definition: Let A be an 𝑚 × 𝑛 matrix. The transpose of A, denoted by 𝐴′ or 𝐴𝑡 , is the
𝑛 × 𝑚 matrix obtained from A by interchanging the rows and columns of A. Thus, the
first row of A is the first column of At, the second row of A is the second column of At
and so on.
2 3
2 −4 6 𝑡
Example: If 𝐴 = ( ), then 𝐴 = (−4 1)
3 1 4
6 4
Activity: Find a 3x3 matrix A for which A = At.
Properties: a) If A and B have the same order, (𝐴 ± 𝐵)𝑡 = 𝐴𝑡 ± 𝐵 𝑡 . b) For a scalar
𝑘, (𝑘𝐴)𝑡 = 𝑘𝐴𝑡 .
c) If A is 𝑚 × 𝑛 and B is 𝑛 × 𝑝, then (𝐴𝐵)𝑡 = 𝐵 𝑡 𝐴𝑡 . d) (𝐴𝑡 )𝑡 = 𝐴

13
Definition: A square matrix A is said to be orthogonal if 𝐴𝐴𝑡 = 𝐴𝑡 𝐴 = 𝐼
1 1 −1
Example: 𝐴 = ( ) is orthogonal (verify)
√2 1 1
Definition 3.3.3: A square matrix 𝐴 = (𝑎𝑖𝑗 ) is said to be symmetric if 𝐴𝑡 = 𝐴, or
equivalently, if 𝑎𝑖𝑗 = 𝑎𝑗𝑖 for each i and j.
2 1 5
Example:𝐴 = (1 0 −3) is symmetric.
5 −3 6
𝑎 3 4 8
𝑏 𝑐 −3 9
Activity: 1. For 𝐴 = ( ) is to be a symmetric matrix, what numbers
𝑑 𝑒 𝑓 10
𝑔 ℎ 𝑖 𝑗
should the letters a to j represent?
2. a) Does a symmetric matrix have to be square?
b) Are all square matrices symmetric?
Definition: A square matrix 𝐴 = (𝑎𝑖𝑗 ) is said to be skew symmetric if 𝐴𝑡 = −𝐴, or
equivalently, if 𝑎𝑖𝑗 = −𝑎𝑗𝑖 for each i and j
Remark: 𝒂𝒊𝒊 = −𝒂𝒊𝒊 ⇒ 𝟐𝒂𝒊𝒊 = 𝟎 or 𝒂𝒊𝒊 = 𝟎. Hence elements of main diagonal of a
skew-symmetric matrix are all zero.
 0 5 7 0 −5 −7
 
Example: For A =  − 5 0 3 , 𝐴𝑡 = (5 0 −3) = −𝐴. So A is skew-symmetric.
 − 7 − 3 0 7 3 0
 
Properties of symmetric and skew-symmetric matrices
1. For any square matrix A, A + At is symmetric and A – At is skew- symmetric
2. If A and B are two symmetric (or skew symmetric) matrices of the same order,
then so is A + B. Proof (exercise)
3. If A is symmetric or skew symmetric, then so is kA. Proof (exercise)
4. Let A and B be symmetric matrices of the same order. The product AB is
symmetric if and only if AB = BA.
Proof: AB is symmetric ⇒ (𝐴𝐵)𝑡 = 𝐴𝐵 ⇒ 𝐵 𝑡 𝐴𝑡 = 𝐴𝐵
⇒ 𝐵𝐴 = 𝐴𝐵
Suppose AB = BA. Then (AB)t = BtAt = BA = AB

14
Exercise
1. a) Form a 4 by 5 matrix, B, such that 𝒃𝒊𝒋 = 𝒊 ∗ 𝒋, where ∗ represents
multiplication.
b) What is BT? c) Is B symmetric? Why or why not?
3 −1 0 −2 4 3
2. Given 𝐴 = [2 4 5] and B=[ 5 1 7]. Verify that
1 3 6 2 3 8
i) (𝐴 + 𝐵)𝑡 = 𝐴𝑡 + 𝐵 𝑡 , ii) (𝐴𝐵)𝑡 = 𝐵 𝑡 𝐴𝑡 iii) (2𝐴)𝑡 =
2𝐴𝑡
1 1 1
3. Let 𝐴 = [ ], is𝐴𝑡 𝐴 is symmetric
1 2 3
Definition: let 𝐴 = (𝑎𝑖𝑗 ) be an 𝑚𝑥𝑛 matrix with complex entries. Then the
conjugate
transpose of A, denoted A*, is an n × m matrix 𝐴∗ = (𝑎
̅̅̅)
𝑗𝑖 for all i, j, where for a ∈ C, a

denotes the complex-conjugate of a (i.e. ̅̅̅̅̅̅̅̅

𝛼 + 𝑖𝛽 = 𝛼 − 𝑖𝛽).
1 4+𝑖 1 0
Example: If 𝐴 = ( ), then 𝐴∗ = ( )
0 1−𝑖 4−𝑖 1+𝑖
Remark: Let A be a square matrix with complex entries. Then, A is called,
1 𝑖
(a) Normal if 𝐴∗ 𝐴 = 𝐴𝐴∗ . For example, 𝐴 = [ ] is normal matrix.
𝑖 1
1 1+𝑖
(b) Hermitian if 𝐴∗ = 𝐴. For example, 𝐴 = [ ] is a Hermitian matrix.
1−𝑖 2
0 1+𝑖
(c) Skew-Hermitian if 𝐴∗ = −𝐴. For example, 𝐴 = [ ] is a skew-
−1 + 𝑖 0
Hermitian matrix
1 1+𝑖 1
(d) Unitary if 𝐴∗ 𝐴 = 𝐴𝐴∗ = 𝐼. For example, 𝐴 = √3 [ ] is unitary
−1 1−𝑖
matrix.
1 1
Let 𝐴 be a complex square matrix. Then 𝐴 = ⏟ (𝐴 + 𝐴∗ ) + ⏟(𝐴 − 𝐴∗ ) .
2 2
𝐻𝑒𝑟𝑚𝑖𝑡𝑖𝑎𝑛 𝑠𝑘𝑒𝑤−𝐻𝑒𝑟𝑚𝑖𝑡𝑖𝑎𝑛

1.3.5. Determinants
Definition: if A is a square matrix, then the determinant function associates with A
exactly one numerical value called the determinant of A, that gives us valuable

15
information about the matrix. By denoting the determinant of A by |𝐴| or det A we
can think of the determinant function as correspondence:
𝐴 → |𝐴|
𝑠𝑞𝑢𝑎𝑟𝑒𝑚𝑎𝑡𝑟𝑖𝑥 𝑑𝑒𝑡 𝑒 𝑟 𝑚𝑖𝑛 𝑎 𝑛𝑡 𝑜𝑓 𝐴
In this case, the straight bars do NOT mean absolute value; they represent the
determinant of the matrix. We will see some of the uses of the determinant in the
subsequent sections. For now, let's find out how to compute the determinant of a
matrix so that we can use it later.
Definition: (Determinant of order 1): Let 𝐴 = [𝑎11 ] be a square matrix of order 1.
Then determinant of A is defined as the number 𝑎11 itself. That is,|𝑎11 | = 𝑎11 .
Example: |3| = 3, |−5| = −5 and |0| = 0
𝑎11 𝑎12
Definition: (Determinant of order 2): Let [𝑎 𝑎22 ] be a 2 × 2 matrix, then
21

A = a11a 22 − a12 a 21 .

That is, the determinant of a 2 × 2 matrix is obtained by taking the product of the
entries in the main diagonal and subtracting from it the product of the entries in the
other diagonal.
To define the determinant of a square matrix A of order n(n > 2), we need the concepts
of the minor and the cofactor of an element.
Let |𝐴| = |𝑎𝑖𝑗 | be a determinant of order n. The minor of aij, is the determinant that
is left by deleting the ith row and the jth column. It is denoted by Mij.
𝑎11 𝑎12 𝑎13
For example, given the 3 x 3 determinant | 21 𝑎22 𝑎23 |. The minor of 𝑎11 is
𝑎
𝑎31 𝑎32 𝑎33
𝑎22 𝑎23 𝑎21 𝑎23
𝑀11 = |𝑎 |, the minor of 𝑎 is 𝑀 = |𝑎32 𝑎33 |, and so on.
32 𝑎33
12 12

Let |𝐴| = |𝑎𝑖𝑗 |be a determinant of order n. The cofactor of aij denoted Cij or Aij, is
defined as (−1)𝑖+𝑗 𝑀𝑖𝑗 , where i + j is the sum of the row number i and column
𝑀𝑖𝑗 , 𝑖𝑓𝑖 + 𝑗𝑖𝑠𝑒𝑣𝑒𝑛
number j in which the entry lies. Thus 𝐶𝑖𝑗 = { . For example, the
−𝑀𝑖𝑗 , 𝑖𝑓𝑖 + 𝑗𝑖𝑠𝑜𝑑𝑑

16
 𝑎11 𝑎12 𝑎13
cofactor of a12 in the 3 x 3 determinant |𝑎21 𝑎22 𝑎23 | is 𝐶12 =
𝑎31 𝑎32 𝑎33
𝑎21 𝑎23 𝑎21 𝑎23
(−1)1+2 |𝑎 𝑎 | = − |𝑎 |
31 33 31 𝑎33

0 1 2
Example: Evaluate the cofactor of each of the entries of the matrix: (1 2 3)
3 1 1
Solution: C11 = -1, C21 = 1 , C31 = -1, C12 = 8, C13 = -5, C22 = -6, C32 = 2, C23 = 3, C33 = -1
Activity: Evaluate the cofactor of each of the entries of the given matrices:
2 3 4 2 0 −1
a. (3 2 1) b. (5 1 0)
1 1 −2 0 1 3
Definition: (Determinant of order n): If A is a square matrix of order n (n >2), then
its determinant may be calculated by multiplying the entries of any row (or column)
by their cofactors and summing the resulting products. That is,
𝒅𝒆𝒕 𝑨 = 𝒂𝒊𝟏 𝑪𝒊𝟏 + 𝒂𝒊𝟐 𝑪𝒊𝟐 +. . . . . +𝒂𝒊𝒏 𝑪𝒊𝒏 Or 𝒅𝒆𝒕 𝑨 = 𝒂𝟏𝒋 𝑪𝟏𝒋 +
𝒂𝟐𝒋 𝑪𝟐𝒋 +. . . . . +𝒂𝒏𝒋 𝑪𝒏𝒋
Remark: It is a fact that determinant of a matrix is unique and does not depend on
the row or column chosen for its evaluation.
1 −3 4
Example: Find the value of | 0 2 5|
−2 6 3
Solution: Choose a given row or column. Let us arbitrarily select the first row. Then
1 −3 4
2 5 0 5 0 2
|0 2 5| = (1) | | + (−3)(−1) | | + 4| |=1(6 − 30) + 3(0 +
6 3 −2 3 −2 6
−2 6 3
10) + 4(0 + 4)
= 22
If we had expanded along the first column, then
1 −3 4
2 5 −3 4
|0 2 5| = (1) | | + 0 + (−2) | | = 1(6 − 30) − 2(−15 − 8) = 22, as
6 3 2 5
−2 6 3
before

17
 1 2 0 −1
3 −1 4 1
Example: Find the value of |𝐴| = | |
−2 0 −3 3
4 3 1 2
Solution: Expanding along first row, we have
|𝐴| = 𝑎11 𝐶11 + 𝑎12 𝐶12 + 𝑎13 𝐶13 + 𝑎14 𝐶14=𝑎11 𝑀11 − 𝑎12 𝑀12 + 𝑎13 𝑀13 − 𝑎14 𝑀14
−1 4 1 3 4 1 3 −1 4
=(1) | 0 −3 3| − 2 |−2 −3 3| + 0 + (1) |−2 0 −3|= 54 – 94 + 13 = -27
3 1 2 4 1 2 4 3 1
Activity: Compute the determinant of A if:
1 3 5 1 5 0
a. 𝐴 = (2 1 1) c. 𝐴 = (2 4 −1)
3 4 2 0 −2 0
5 −7 2 2 3 0 0 0
0 3 0 −4 5 −2 0 0
b. 𝐴 = ( ) d. 𝐴 = ( )
−5 −8 0 3 −8 3 1 0
0 5 0 −6 4 −7 5 2
Note: 1. 𝒅𝒆𝒕 𝑰𝒏 = 𝟏, where 𝐼𝑛 is an identity matrix of order n.
2. 𝒅𝒆𝒕 𝑨 = ∏𝒏𝒏=𝟏 𝒅𝒊𝒊 , where 𝒅𝒊𝒊 𝒊𝒔 the diagonal element of a diagonal matrix or
lower triangular matrix or upper triangular matrix.
The following diagram called Sarrus’ diagram, enables us to write the value of the
determinant of order 3 very conveniently. This technique does not hold for
determinant of higher order.
Working Rule: Make the Sarrus’ diagram by repeating the first two columns of the
determinant as shown below. Then multiply the elements joined by arrows. Assign
the positive sign to an expression if it is formed by a downward arrow and negative
sign to an expression if it is formed by an upward arrow.
Value: 𝑎11 𝑎22 𝑎33 + 𝑎12 𝑎23 𝑎31 + 𝑎13 𝑎21 𝑎32 −
𝑎31 𝑎22 𝑎13 − 𝑎32 𝑎23 𝑎11 − 𝑎33 𝑎21 𝑎12

Example: Find the value of |𝐴| =

2 −1 3
|5 7 0| with the help of Sarrus’ diagram
4 1 6

18
Solution: The Sarrus’ diagram for the given determinant is to the right. Thus the value
of the determinant is
|𝐴| = (2)(7)(6) + (−1)(0)(4) + (3)(5)(1) − (4)(7)(3) − (1)(0)(2) −
(6)(5)(−1) = 45
Exercise:
2 1 5
3 2 −2 −𝑎
1) Evaluate the following determinants: a) | | b) | | c) |−3 4 −1|
−5 −4 −𝑎 2
0 6 −1
1 2 3
2) Let𝐴 = [4 5 4]. Determine each of the following
3 2 1
a) the minor of a21 b) the minor of a22 c) the cofactor of a22
d) the cofactor of a23 e) the cofactor of a32.
Properties of Determinants
We now state some useful properties of determinants. These properties help a good
deal in the evaluation of determinants. We use the notations Ri and Cj to denote
respectively the i-th row and the j-th column of a determinant.
Property 1: The value of a determinant remains unchanged if rows are changed into
columns and columns into rows. That is,
𝑎11 𝑎12 𝑎13 𝑎11 𝑎21 𝑎31
𝑎
| 21 𝑎 22 𝑎 23 | = |𝑎12 𝑎22 𝑎32 | or 𝒅𝒆𝒕 𝑨 =
𝑎31 𝑎32 𝑎33 𝑎13 𝑎23 𝑎33
𝒅𝒆𝒕 𝑨𝒕
Remark: In terms of matrices, if A is a square matrix, then |𝐴| = |𝐴′|.

19
Property 2: If any two rows (or columns) of a determinant are interchanged, the
value of the determinant so obtained is the negative of the value of the original
determinant. That is,
𝑎11 𝑎12 𝑎13 𝑎11 𝑎12 𝑎13
|𝑎21 𝑎22 𝑎23 | = − |𝑎31 𝑎32 𝑎33 |.
𝑎31 𝑎32 𝑎33 𝑎21 𝑎22 𝑎23
Remark: The notation 𝑅𝑖 ↔ 𝑅𝑗 (𝐶𝑖 ↔ 𝐶𝑗 ) is used to represent interchange of ith and jth
row (column).
Property 3: If any two rows (or columns) of a determinant are identical, the value of
𝑎1 𝑏1 𝑐1
the determinant is zero. That is, |𝑎1 𝑏1 𝑐1 | = 0. R1 and R2 are identical
𝑎2 𝑏2 𝑐2
Property 4: If each element of a row (or column) of a determinant is multiplied by a
constant k, the value of the determinant so obtained is k times the value of the original
𝑎11 𝑎12 𝑎13 𝑎11 𝑎12 𝑎13
𝑘𝑎
determinant. That is, | 21 𝑘𝑎 22 𝑘𝑎 23 | = 𝑘 | 21 𝑎22 𝑎23 |
𝑎
𝑎31 𝑎32 𝑎33 𝑎31 𝑎32 𝑎33
Remark: 1. the notation 𝑅𝑖 → 𝑘𝑅𝑖 (𝐶𝑖 → 𝑘𝐶𝑖 ) is used to represent multiplication of
each element of ith row (column) by the constant k.
2.𝒅𝒆𝒕( 𝒌𝑨) = 𝒌𝒏 𝒅𝒆𝒕 𝑨, where k is any real number and A is an 𝑛 × 𝑛 matrix.
Property 5: If the elements of a row (or column) of a determinant are added k times
the elements of another row (or column), the value of the determinant so obtained is
equal to the value of the original determinant. That is,
𝑎11 𝑎12 𝑎13 𝑎11 + 𝑘𝑎31 𝑎12 + 𝑘𝑎32 𝑎13 + 𝑘𝑎33
𝑎
| 21 𝑎22 𝑎23 | = | 𝑎21 𝑎22 𝑎23 |
𝑎31 𝑎32 𝑎33 𝑎31 𝑎32 𝑎33
Property 6: If each element of a row (or column) of a determinant is the sum of two
elements, the determinant can be expressed as the sum of two determinants. That is,
𝑎11 𝑎12 𝑎13 𝑎11 𝑎12 𝑎13 𝑎11 𝑎12 𝑎13
| 𝑎21 𝑎22 𝑎23 | = |𝑎21 𝑎22 𝑎23 | + |𝑎21 𝑎22 𝑎23 |
𝑎31 + 𝑏1 𝑎32 + 𝑏2 𝑎33 + 𝑏3 𝑎31 𝑎32 𝑎33 𝑏1 𝑏2 𝑏3
1 18 72
Example: Find the value of the determinant |𝐴| = |2 40 148|
3 45 150

20
Solution: By applying various properties of determinants, we make maximum
number of zeros in a row or a column. We shall make maximum number of zeros in
C1. Performing the operations 𝑅2 → 𝑅2 − 2𝑅1 and 𝑅3 → 𝑅3 − 2𝑅1 (property 5),
1 18 72
4 4
|𝐴| = |0 4 4 |=| |= 24 - 36 = -12
9 6
0 9 6
𝑦+𝑧 𝑥 𝑦
Example: Show that | 𝑧 + 𝑥 𝑧 𝑥 | = (𝑥 + 𝑦 + 𝑧)(𝑥 − 𝑧)2
𝑥+𝑦 𝑦 𝑧
𝑦+𝑧 𝑥 𝑦
Solution: Let𝛥 = | 𝑧 + 𝑥 𝑧 𝑥 |. Performing R1 → R1 + R2 , 𝑅1 → 𝑅1 + 𝑅3 , we get
𝑥+𝑦 𝑦 𝑧
2(𝑥 + 𝑦 + 𝑧) 𝑥 + 𝑦 + 𝑧 𝑥+𝑦+𝑧 2 1 1
𝛥=| 𝑧+𝑥 𝑧 𝑥 |= (𝑥 + 𝑦 + 𝑧) | 𝑧 + 𝑥 𝑧 𝑥|
𝑥+𝑦 𝑦 𝑧 𝑥+𝑦 𝑦 𝑧
Performing 𝐶1 → 𝐶1 − 2𝐶2 and𝐶2 → 𝐶2 − 𝐶3, we get
0 0 1
= 𝛥 = (𝑥 + 𝑦 + 𝑧) | 𝑥 − 𝑧 𝑧−𝑥 𝑥| (Property 5)
𝑥−𝑦 𝑦−𝑧 𝑧
= (𝑥 + 𝑦 + 𝑧)(𝑥 − 𝑧)2
Activity: Evaluate the following determinants by using the properties listed above:
1 −3 1 −2
3 1 43 2 4 6
2 −5 −1 −2
a) |2 7 35| b) |7 9 11| c) | |
0 −4 5 1
1 3 17 8 10 12
−3 10 −6 8
Product of two determinants
Theorem: The determinant of the product of two matrices of order n is the product
of their determinants. That is, |𝐴𝐵| = |𝐴||𝐵|.
3 0 2 5 3 0 2 5
Example: Let 𝐴 = [ ] and 𝐵 = [ ], then |𝐴𝐵| = |𝐴||𝐵| = | || |=
4 1 1 4 4 1 1 4
(3)(3) = 9
Example: Let A and B be 3x3 matrix with det A = 2 and det B = -3.
Find det (2ABt).
Solution: 𝒅𝒆𝒕( 𝟐𝑨𝑩𝒕 ) = 𝟐𝟑 𝒅𝒆𝒕 𝑨 𝒅𝒆𝒕 𝑩𝒕 = 𝟖(𝟐)(−𝟑) = −𝟒𝟖, since det B = det Bt.
Adjoint and Inverse of a matrix

21
Definition: Let A = (aij) be a square matrix of order n and let Cij be the cofactor of aij.
Then the adjoint of A, denoted by adj A, is defined as the transpose of the cofactor
matrix (Cij).
1 2 3
Example: Find adj A, if 𝐴 = [−1 0 1]
4 3 2
Solution: We have C11=-3, C12=6, C13= -3, C21=5, C22=-10, C23=5, C31=2,C32=-4, C33=2.
−3 5 2
Thus, 𝑎𝑑𝑗𝐴 = [ 6 −10 −4].
−3 5 2
Activity: Evaluate the following determinants by using the properties listed above:
1 −3 1 −2
3 1 43 2 4 6
2 −5 −1 −2
a) |2 7 35| b) |7 9 11| c) | |
0 −4 5 1
1 3 17 8 10 12
−3 10 −6 8
Properties of the Adjoint of a matrix
1. If A is a square matrix of order n, then
A(adj A) = |A| In = (adj A)A, where In is an identity matrix of order n.
2. If A is a square matrix of order n, then adj (𝐴′) = (𝑎𝑑𝑗𝐴)′
3. If A and B are two square matrices of the same order, then
adj(AB) = adj(B) adj (A).
2 1 3
Example: If𝐴 = [ 2 0 1], verify that A(adjA) = |A| I3 = (adjA)A
−4 5 6
Solution: We have |A| = 2(-5)-1(12+4) +3(10) = -10 – 16 + 30 = 4
Now C11=-5, C12=-16, C13=10, C21 = 9, C22=24, C23=-14, C31=1, C32=4, C33=-2.
−5 9 1
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒, 𝑎𝑑𝑗𝐴 = [−16 24 4]
10 −14 −2
2 1 3 −5 9 1 4 0 0 1 0 0
Hence 𝐴(𝑎𝑑𝑗𝐴) = [ 2 0 1] [−16 24 4 ]= [0 4 0] = 4 [0 1 0] =
−4 5 6 10 −14 −2 0 0 4 0 0 1
|𝐴|𝐼3
Similarly, it can be proved that (𝑎𝑑𝑗𝐴)𝐴 = |𝐴|𝐼3
Definition: Let A be a square matrix of order n. Then a square matrices B of order n,
if it exists, is called an inverse of A if AB = BA = In. A matrix A having an inverse is

22
called an invertible matrix. It may easily be seen that if a matrix A is invertible, its
inverse is unique. The inverse of an invertible matrix A is denoted by A-1.
Does every square matrix possess an inverse? To answer this let us consider the
matrix
0 0
𝐴=[ ]. If B is any square matrix of order 2, we find that AB = BA = 0.
0 0
We thus see that there cannot be any matrix B for which AB and BA both are equal to
I2. Therefore A is not invertible. Hence, we conclude that a square matrix may fail to
have an inverse. However, if A is a square matrix such that|𝐴| ≠ 0, then A is invertible
and
1
𝐴−1 = |𝐴| 𝑎𝑑𝑗𝐴. For, we know that 𝐴(𝑎𝑑𝑗𝐴) = (𝑎𝑑𝑖𝐴)𝐴 = |𝐴|𝐼𝑛
1 1 1
𝐴 (|𝐴| 𝑎𝑑𝑗𝐴) = (|𝐴| 𝑎𝑑𝑗𝐴) 𝐴 = 𝐼𝑛 . Thus A is invertible and 𝐴−1 = |𝐴| 𝑎𝑑𝑗𝐴.

A square matrix A is said to be singular (not invertible) if A = 0 , and it is called non-

singular (invertible) if |𝐴| ≠ 0.

6 7 −1
Example: Find 𝜆 if the matrix 𝐴 = [3 𝜆 5 ]has no inverse.
9 11 𝜆
Solution: 6(𝜆2 − 55) − 7(3𝜆 − 45) − (33 − 9𝜆) = 0
➢ 𝜆2 − 2𝜆 − 8 = 0
➢ (𝜆 − 2)(𝜆 − 4) = 0
➢ 𝜆 = 2 or 𝜆 = 4
3 1 2
Example: If 𝐴 = [2 −3 −1], then |𝐴| = 3(−3 + 2) − 1(2 + 1) + 2(4 + 3) = 8
1 2 1
Since |𝐴| ≠ 0, 𝐴 is non-singular or invertible.
Activity: Find A-1.
2 0 0
Further, if𝐵 = [3 −1 4]then|𝐵| = 0and it is singular.
5 −2 8
1
Note: If A is an invertible nxn matrix, then AA-1 = In and det A-1 = 𝑑𝑒𝑡 𝐴, where 𝑑𝑒𝑡 𝐴 ≠

0.
Properties of the inverse of a matrix

23
 1. A square matrix is invertible if and only if it is non-singular.

(
2. The inverse of the inverse is the original matrix itself, i.e. A − 1 )−1 = A )
3. The inverse of the transpose of a matrix is the transpose of its inverse, i.e.,
(𝐴𝑡 )−1 = (𝐴−1 )𝑡
4. If A and B are two invertible matrices of the same order, then AB is also
invertible and moreover, (𝐴𝐵)−1 = 𝐵 −1 𝐴−1
4 −2 1
Example: Find the inverse of the matrix A=[7 3 3].
2 0 1
Solution: |𝐴| = (4)(3) − (−2)(1) + 1(−6) = 8 ≠ 0. Thus 𝐴−1 exists and is given by
1
𝐴−1 = |𝐴| 𝑎𝑑𝑗𝐴. To find adjA, let Cij denote the cofactor of aij, the element in the ith row

and jth column of |A|. Thus C11=3, C12 = -1, C13=-6, C21 = 2, C22=2, C23=-4, C31=-9 C32=-
3 2 −9
5 and C33 =26. 𝑎𝑑𝑗𝐴 = [−1 2 −5].
−6 −4 16
3 2 −9
1 1
Hence 𝐴−1 = |𝐴| 𝑎𝑑𝑗𝐴 = 8 [−1 2 −5]
−6 −4 16
1 4 0
𝑎 𝑏
Activity: 1. Find the inverse of A, if i)𝐴 = [ ]
ii) 𝐴 = [−1 2 2]
𝑐 𝑑
0 0 2
3 4 2 8
2. Find matrix A such that 𝐴 [ ]=[ ].
6 2 9 4
3. If 𝐴𝑋 = 𝑏 then𝑋 = 𝐴−1 𝑏. (True/False)

1.4. System of Linear equations

In this section we will present certain systematic methods for solving system of linear
equations
Definition: A linear equation in the variables 𝑥1 , 𝑥2 , . . . , 𝑥𝑛 over the real field ℜis an
equation that can be written in the form 𝑎1 𝑥1 + 𝑎2 𝑥2 +. . . . +𝑎𝑛 𝑥𝑛 = 𝑏
(1)
where b and the coefficients 𝑎1 , 𝑎2 , . . . , 𝑎𝑛 are given real numbers.
Definition: A system of linear equations (or a linear system) is a collection of one
or more linear equations involving the same variables, say 𝑥1 , 𝑥2 , . . . , 𝑥𝑛 .

24
Now consider a system of m linear equations in n-unknowns𝑥1 , 𝑥2 , . . . , 𝑥𝑛 :
𝑎11 𝑥1 + 𝑎12 𝑥2 +. . . +𝑎1𝑛 𝑥𝑛 = 𝑏1
𝑎21 𝑥1 + 𝑎22 𝑥2 +. . . +𝑎2𝑛 𝑥𝑛 = 𝑏2
.
.
.
𝑎𝑚1 𝑥1 + 𝑎𝑚2 𝑥2 +. . . +𝑎𝑚𝑛 𝑥𝑛 = 𝑏𝑚 (2)
If 𝑏1 = 𝑏2 =. . . = 𝑏𝑚 = 0 then we say that the system is homogeneous. If 𝑏𝑖 ≠ 0 for
some
i{1,2,3, . . ., m} then the system is called non homogeneous. In matrix notation, the
linear system (2) can be written as 𝐴𝑋 = 𝐵where
𝑎11 𝑎12 ... 𝑎1𝑛 𝑥1 𝑏1
𝑎21 𝑎22 ... 𝑎2𝑛 𝑥2 𝑏2
. . .
𝐴= ,𝑋 = and 𝐵 =
. . .
. . .
[𝑎𝑚1 𝑎𝑚2 ... 𝑎𝑚𝑛 ] [𝑥𝑛 ] [𝑏𝑚 ]
We call A the coefficient matrix of the system (2).
Observe that entries of the k-th column of A are the coefficients of the variable𝑥𝑘 in
(2).
The 𝑚 × (𝑛 + 1) matrix whose first n columns are the columns of A (the coefficient
matrix) and whose last column is B is called the augmented matrix of the system.
We denote it by [AB]. The augmented matrix determines the system (2) completely
because it contains all the coefficients and the constants to the right side of each
equation in the system. For example for the non homogeneous linear system
𝑥1 + 3𝑥2 − 𝑥3 = 2
𝑥2 − 2𝑥3 = 4
−2𝑥1 − 3𝑥2 − 3𝑥3 = 5 (3)
1 3 −1 1 3 −1 2
The matrix 𝐴 = [ 0 1 −2] is the coefficient matrix and [ 0 1 −2 4]is the
−2 −3 −3 −2 −3 −3 5
augmented matrix.

25
Are the coefficient matrix and the augmented matrix of a homogeneous linear system
equal? Why?
A solution of a linear system in n-unknowns 𝑥1 , 𝑥2 , . . . , 𝑥𝑛 is an n-tuple (𝑠1 , 𝑠2 , . . . , 𝑠𝑛 )
of real numbers that makes each of the equations in the system a true statement when
si is substituted for xi, i = 1,2, . . ., n. The set of all possible solutions is called the
solution set of the linear system. We say that two linear systems are equivalent if
they have the same solution set.
Activity: Give the coefficient matrix and the augmented matrix of the linear system
𝑥1 + 3𝑥2 = 2𝑥3
𝑥1 = 11𝑥3
2𝑥1 − 𝑥2 + 4𝑥3 = 0
−𝑥1 + 𝑥2 = 2
2) Does the linear system have a solution?
𝑥1 − 𝑥2 = 0
𝑢+𝑣 =2
3) Find the solution set of the linear system
𝑢−𝑣 =1
How many solutions does it have?
3
2𝑥1 − 𝑥2 + 2 𝑥3 = 8 13
4) Given the system { . Is (5, 2 , 3)a solution of the linear
𝑥1 − 4𝑥3 = −7
−25
system? What about (-7, -22, 0) and (3, , 1)?
2

The activity given above illustrates the following general fact about linear systems.
A system of linear equations has either
1. no solution, or
2. exactly one solution, or
3. Infinitely many solutions.
We say that a linear system is consistent if it has either one solution or infinitely
many solutions; a system is inconsistent if it has no solution.
Activity
1. The homogeneous linear system 𝐴𝑋 = 𝑂 is consistent for any 𝑚 × 𝑛matrix
A.
Explain, why?

26
 2. Consider a linear system of two equations in two unknowns, give geometric
interpretation if the system has
i) no solution ii) exactly one solution iii) many solutions
Do the same for a linear system of three equations in three unknowns.
Solving a linear system
This is the process of finding the solutions of a linear system. We first see the
technique of elimination (Gaussian elimination method) and then we add two more
techniques, matrix inversion method and Cramer’s rule.
Cramer’s rule for solving system of linear equations (homogeneous and non-
homogeneous)
Suppose we have to solve a system of n linear equations in n unknowns Ax = b. Let
Ai(b) be the matrix obtained from A by replacing column i by the vector b and Ak be
the k-th column vector of matrix A.

Now let e1, e2, . . . en be columns of the n × n identity matrix I and Ii(x) be the matrix
obtained from I by replacing column i by x.
If Ax = b then by using matrix multiplication we have
AIi(x) = A[e1 . . . x . . . en] = [Ae1 . . . Ax . . . Aen]
= [A1 . . . b . . . An] = Ai(b)
By the multiplicative property of determinants, (detA)(detIi(x)) = det Ai(b)
The second determinant on the left is xi. (Make a cofactor expansion along the ith
𝑑𝑒𝑡 𝐴𝑖 (𝑏)
row.) Hence (det A). xi = det Ai(b). Therefore if detA  0 then we have 𝑥𝑖 = .
𝑑𝑒𝑡 𝐴

This method for finding the solutions of n linear equations in n unknowns is known
as Cramer’s Rule.
Example: Solve the following system of linear equations by Cramer’s Rule.
2𝑥1 − 𝑥2 + 𝑥3 = 6
𝑥1 + 4𝑥2 − 2𝑥3 = −4
3𝑥1 + 𝑥3 = 7

27
Solution: Matrix form of the given system is 𝐴𝑥 = 𝑏
2 −1 1 𝑥1 6
𝑥
where 𝐴 = (1 4 −2) 𝑥 = ( 2 ) 𝑎𝑛𝑑𝑏 = (−4)
3 0 1 𝑥3 7
𝑑𝑒𝑡 𝐴𝑖 (𝑏)
By Cramer’s Rule, 𝑥𝑖 = (𝑖 = 1,2,3)
𝑑𝑒𝑡 𝐴

2 −1 1
𝑑𝑒𝑡 𝐴 = |1 4 −2| = 3
3 0 1
6 −1 1 2 6 1
|−4 4 −2| |1 −4 −2|
𝑑𝑒𝑡 𝐴1 (𝑏) 7 0 1 6 𝑑𝑒𝑡 𝐴2 (𝑏) 3 7 1 −3
𝑥1 = = = 3 = 2 ,𝑥2 = = = = −1 and 𝑥3 =
𝑑𝑒𝑡 𝐴 2 𝑑𝑒𝑡 𝐴 2 3
2 −1 6
|1 4 −4|
𝑑𝑒𝑡 𝐴3 (𝑏) 3 0 7 3
= = 3 = 1.
𝑑𝑒𝑡 𝐴 2

Example: Solve the following system of linear equations by Cramer’s Rule.

2𝑥1 + 𝑥2 = 7
−3𝑥1 + 2𝑥3 = −8
𝑥2 + 2𝑥3 = −3
Solution: Matrix form of the given system is 𝐴𝑥 = 𝑏
2 1 0 𝑥1 7
where 𝐴 = (−3 0 1) 𝑥 = (𝑥2 ) 𝑎𝑛𝑑𝑏 = (−8)
0 1 2 𝑥3 −3
𝑑𝑒𝑡 𝐴𝑖 (𝑏)
By Cramer’s Rule, 𝑥𝑖 = (𝑖 = 1,2,3)
𝑑𝑒𝑡 𝐴

2 1 0
𝑑𝑒𝑡 𝐴 = |−3 0 1| = 4
0 1 2
7 1 0 2 7 0
|−8 0 1| |−3 −8 1|
𝑑𝑒𝑡 𝐴1 (𝑏) −3 1 2 6 3 𝑑𝑒𝑡 𝐴2 (𝑏) 0 −3 2 16
𝑥1 = = = 4 = 2,𝑥2 = = = = 4and 𝑥3 =
𝑑𝑒𝑡 𝐴 4 𝑑𝑒𝑡 𝐴 4 4
2 1 7
|−3 0 −8|
𝑑𝑒𝑡 𝐴3 (𝑏) 0 1 −3 −14 −7
= = = .
𝑑𝑒𝑡 𝐴 4 4 2

Activity
1. Use Cramer’s rule to solve each of the following
a) 2z + 3 = y + 3x b) –a + 3b – 2c = 7
x – 3z = 2y + 1 3a + 3c = -3
3y + z = 2-2x 2a + b + 2c = -1

28
 2. Consider the system of homogeneous n linear equations in n unknowns
Ax = 0. Discuss about the cases that we can use Cramer’s Rule.
For the non -homogeneous system𝐴𝑥 = 𝑏, if 𝑑𝑒𝑡 𝐴 = 0, then the Cramer’s rule does
not give any information whether or not the system has a solution. However, in the
case of homogeneous system we have the following useful theorem.
Theorem: A system of n homogenous linear equations in n unknowns Ax = 0, has a
non- trivial solution if𝑑𝑒𝑡 𝐴 = 0. If 𝑑𝑒𝑡 𝐴 ≠ 0, it has only the trivial
solution x1 = x2 = … = xn = 0.
6 7 −1
Example: Let 𝐴 = (3 𝜆 5 ). Find the value(s) of 𝜆if 𝐴𝑥 = 0has non-zero
9 11 𝜆
solution.
Solution: det A = 0
6 7 −1
➢ |3 𝜆 5 |=0
9 11 𝜆
➢ 6(𝜆2 − 55) − 7(3𝜆 − 45) − (33 − 9𝜆) = 0
➢ 𝜆2 − 2𝜆 − 8 = 0
➢ (𝜆 − 2)(𝜆 − 4) = 0 or 𝜆 = 2, 𝜆 = 4
REMARK:
1. If A is a 2x2 matrix, the area of the parallelogram determined by the
columns of A is det A.
2. If A is a 3x3 matrix, the volume of the parallelepiped determined by the
columns of A is det A.
1.5. Eigenvalue and Eigenvector

Definition:  is an eigenvalue of an (n  n) matrix A if, for some column vector X ≠ O,

AX = X

The non-trivial column vector X is an eigenvector of A for that eigenvalue.

Example

29
  1 1 1  6  1
AX =     =   = 6   = 6X
 −5 7   5   30  5

1
X =   is therefore an eigenvector of A for eigenvalue  = 6 (the 6-eigenvalue).
5

If a non-zero column vector X is an eigenvector of (n  n) matrix A for eigenvalue , then

AX =  X  AX =  IX   IX − AX = O  (  I − A) X = O

But this square homogeneous linear system cannot have a non-trivial solution unless
(  I − A) is singular  det (  I − A) = 0 .

The characteristic polynomial of any (n  n) matrix A is cA (  ) = det (  I − A) , which is a

polynomial of degree n in  . The eigenvalues of A are the n solutions of

cA (  ) = det (  I − A) = 0 .

The -eigenvectors of A are the non-trivial solutions to the homogeneous linear system

(  I − A) X = O .

 1 1
Example: Find all eigenvalues and their eigenvectors of A =  .
 −5 7 

  0   1 1   − 1 −1
det (  I − A) = det    −   = = (  2 − 8 + 7 ) + 5
  0    −5 7   5  −7

=  2 − 8 + 12 = (  − 2 )(  − 6 )

30
det (  I − A) = 0   = 2 or  = 6

 1 = 2:

Solving ( 2I − A) X1 = O :

 2 − 1 −1   1 −1   x  0
( 2 I − A) X =   X =  5 −5   y  =  0 
 5 2−7      

 x− y =0  y=x

1
Therefore the 2-eigenvectors of A are any non-zero multiples of X 1 =   .
1

 2 = 6 (which is the case considered earlier):

Solving ( 6I − A) X 2 = O :

 6 − 1 −1   5 −1   x  0
( 6 I − A) X =   X =  5 −1   y  =  0 
 5 6−7      

 5x − y = 0  y = 5x

1
Therefore the 6-eigenvectors of A are any non-zero multiples of X 2 =   .
5

 − 3 1 −1 − 3 −  1 −1 
   
Let A =  − 7 5 − 1  . Then A − I 3 =  − 7 5− − 1  and
 − 6 6 − 2  −6 − 2 −  
   6

5− −1 −7 −1 −7 5−
A − I 3 = (−3 −  ) − − = −3 + 12 + 16 .
6 −2− −6 −2− −6 6

31
In general, it is not easy to find the roots of a polynomial of degree 3. We try some
simple values and hope we will be lucky. For  = −2 , you get

A − I 3 = −(−2) 3 + 12(−2) + 16 = 0 ,

so that  = −2 is a root of A − I 3 .

Hence ( + 2) is a factor of A − I 3 and A − I 3 can be written in the form

A − I 3 = ( + 2)(a2 + b + c) = a3 + 2 (b + 2a) +  (c + 2b) + 2c.

We now have

A − I 3 = −3 + 12 + 16 = a3 + 2 (b + 2a) +  (c + 2b) + 2c,

so that a, b and c must satisfy

 a = −1
 b + 2a = 0

 ,
 c + 2b = 12
 2c = 16

which yields

 a = −1

b=2.
 c=8


32
We are now in position to find all the factors of A − I 3 :

A − I 3 = ( + 2)(−2 + 2 + 8) = −( + 2) 2 ( − 4).

A therefore has two eigenvalues: 0 = −2 and 1 = 4 .

To get the eigenvectors associated with 0 = −2 , we solve ( A + 2I 3 )u = 0 :

 − 1 1 − 1 x1   0   − x1 + x2 − x3 = 0
       x = x2
( A + 2 I 3 )u =  − 7 7 − 1 x2  =  0   − 7 x1 + 7 x2 − x3 = 0   1 .
 − 6 6 0  x   0  − 6 x  x3 = 0
  3     1 + 6 x2 =0

1
 
Hence the eigenvectors associated with the eigenvalue 0 = −2 are of the form   1 .
 0
 

To get the eigenvectors associated with 1 = 4 , we solve ( A − 4I 3 )u = 0 :

 − 7 1 − 1  x1   0  − 7 x1 + x2 − x3 = 0
       x =0
( A + 2 I 3 )u =  − 7 1 − 1  x 2  =  0   − 7 x1 + x2 − x3 = 0   1 .
 − 6 6 − 6  x   0  − 6 x  x 2 = x3
  3     1 + 6 x2 − 6 x3 = 0

 0
 
Hence the eigenvectors associated with the eigenvalue 1 = 4 are of the form   1 .
 1
 
Properties of Eigenvalues and Eigenvectors

At this point it might be a good idea to highlight several properties of eigenvalues and
eigenvectors. The following pertaint to the matrices we are dicussing here, only.

33
• the absolute value of a determinant (|detA|) is the product of the absolute values of
the eigenvalues of matrix A
• c = 0 is an eigenvalue of A if A is a singular (noninvertible) matrix
• If A is a nxn triangular matrix (upper triangular, lower triangular) or diagonal
matrix , the eigenvalues of A are the diagonal entries of A.
• A and its transpose matrix have same eigenvalues.
• Eigenvalues of a symmetric matrix are all real.
• Eigenvectors of a symmetric matrix are orthogonal, but only for distinct
eigenvalues.
• The dominant or principal eigenvector of a matrix is an eigenvector
corresponding to the eigenvalue of largest magnitude (for real numbers, largest
absolute value) of that matrix.
• For a transition matrix, the dominant eigenvalue is always 1.
• The smallest eigenvalue of matrix A is the same as the inverse (reciprocal) of the
largest eigenvalue of A-1; i.e. of the inverse of A.

34
Chapter Two

INVERSE FUNCTION

Any function f associates with each number x in its domain a unique number y in its range.
Although it is not always possible to reverse this procedure and still obtain a function, in
some cases this can be done. The resulting function is called the inverse of f .

Definition: Let f be a function. Then f has an inverse provided that there is a function
g such that the domain of g is the range of f and such that f ( x) = y if and only if
g ( y) = x for all x in the domain of f and all y in the range of f . In this case g is the

inverse of f and it is denoted by f −1 . Thus, 𝑓 (𝑥) = 𝑦 if and only if 𝑓 −1 (𝑦) = 𝑥 for all
x in the domain of f and all y in the range of f.

• To find f-1(x) from an algebraic function f(x), follow these steps:

o Interchange x and y in the equation y = f(x).
o Solve for y.
o Change 𝑦 by 𝑥
Example
9
Let 𝑓(𝑥) = 5 𝑥 + 32, 𝑥 ≥ −273.15. Then find a formula for the inverse of 𝑓, and determine

the domain of 𝑓 −1 .
Solution: Following the steps listed above, we obtain
9
𝑦= 𝑥 + 32
5
9
𝑥 = 𝑦 − 32
5
5
𝑥 = (𝑦 − 32)
9
5
Thus, 𝑓 −1 (𝑦) = (𝑦 − 32)
9

Because we customarily use x as the independent variable, we replace y by x to obtain;

35
 5
𝑓 −1 (𝑥) = (𝑥 − 32)
9
Since f is continuous and increasing, and since 𝑓(−273.15) = −459.67,the range of 𝑓 is
[−459.67, ∞). Therefore, the domain of 𝑓 −1 is [−459.67, ∞).
Consequently 𝑓 −1 is completely defined by the formula
5
𝑓 −1 (𝑥) = 9 (𝑥 − 32) for 𝑥 ≥ −459.67.

Example
Let 𝑓(𝑥) = 8𝑥 3 − 1. Write the formula for the inverse of 𝑓.
Solution: Following the steps listed above, we obtain
𝑦 = 8𝑥 3 − 1
𝑦+1
= 𝑥3
8
1 1
𝑥 = (𝑦 + 1) ⁄3
2
1 1⁄
Thus, 𝑓 −1 (𝑦) = 2 (𝑦 + 1) 3

1 1⁄
Converting to the variable x, we have 𝑓 −1 (𝑥) = 2 (𝑥 + 1) 3

Remark: It is not always possible to find a simple formula for an inverse.

For example, 𝑓(𝑥) = 𝑥 7 + 8𝑥 3 + 4𝑥 − 2 has an inverse but 𝑓 −1 has no simple formula.

Properties of inverse function

Theorem: Let 𝑓 have an inverse. Then 𝑓 and 𝑓 −1 have the following properties:
a. 𝑓 −1 has an inverse, and (𝑓 −1 )−1 = 𝑓
b. 𝑓 −1 (𝑓(𝑥)) = 𝑥 for all 𝑥 in the domain of 𝑓
c. 𝑓(𝑓 −1 (𝑦)) = 𝑦 for all 𝑦 in the range of 𝑓.

Proof
To prove (a), recall first that the domain and range of 𝑓 are the range and domain,
respectively, of 𝑓 −1 . Furthermore,𝑓 −1 (𝑦) = 𝑥 if and only if 𝑓 (𝑥) = 𝑦 for all 𝑦 in the domain
of 𝑓 −1 and all 𝑥 in the range of 𝑓 −1 .

36
Hence 𝑓 −1 has an inverse, which is evidently f. To prove (b), simply substitute 𝑓 (𝑥) for 𝑦 in
the equation 𝑓 −1 (𝑦) = 𝑥. Similarly, substitution of 𝑓 −1 (𝑦) for 𝑥 in the equation 𝑓 (𝑥) = 𝑦
establishes (c).

We would like to have a simple criterion that tells us whether a given function has an
inverse.

A function 𝑓 has an inverse if and only if for any two numbers 𝑥1 and 𝑥2 in the
domain of f, if 𝑥1 ≠ 𝑥2 , then 𝑓 (𝑥1 ) ≠ 𝑓 (𝑥2 ). --------- (*)

Definition: A function that satisfies (*) is said to be one-to-one.

Horizontal Line Test: Any horizontal line will intersect the graph of a one-to-one function
in at most one point.

𝑓 has an inverse

Theorem: Every increasing function and every decreasing function has an inverse.

Proof

Let 𝑓 be increasing, and suppose that 𝑥1 and 𝑥2 are in the domain of 𝑓 with𝑥1 ≠ 𝑥2 . Then
either 𝑥1 < 𝑥2 , in which case 𝑓 (𝑥1 ) < 𝑓 (𝑥2 ) (because 𝑓 is increasing), or 𝑥1 > 𝑥2 , in which
case 𝑓 (𝑥1 ) > 𝑓 (𝑥2 ) (again because 𝑓 is increasing). In either case, 𝑓 (𝑥1 ) ≠ 𝑓 (𝑥2 ).
Therefore 𝑓 satisfies the criterion in (*) and hence has an inverse. The proof for decreasing
functions is similar.

37
Test of monotonicity

Suppose 𝑓 is continuous on [𝑎, 𝑏] and differentiable on (𝑎, 𝑏).

(a) If 𝑓 ′ (𝑥) > 0 for all 𝑥 in (𝑎, 𝑏), then 𝑓 is increasing on [𝑎, 𝑏].
(b) If 𝑓 ′ (𝑥) < 0 for all 𝑥 in (𝑎, 𝑏), then 𝑓 is decreasing on [𝑎, 𝑏].

Summary of Inverse Functions

1. For a one-to-one function, each x-value corresponds to only one y-value, and each y-
value corresponds to only one x-value (i.e., the function passes the vertical line test
and the horizontal line test).
2. If a function f is one-to-one, then f has an inverse function, which we write as f-1.
3. The domain of f is the range of f-1, and the range of f is the domain of f-1.
4. The graphs of f and f-1 are reflections of each other across the line y = x.
Example
For each function f, sketch the graphs of f and f –1 on the same coordinate system
a. 𝑓 (𝑥) = 8𝑥 3 – 1 b. 𝑓 (𝑥) = 𝑠𝑖𝑛 𝑥 𝑓𝑜𝑟 – 𝜋/2 ≤ 𝑥 ≤ 𝜋/2
Solution: The graphs of 𝑓 and 𝑓 −1 are reflections of each other across the line y = x.

38
Continuity and Differentiability of Inverse Function

continuity and differentiability. In this subsection we will show that the inverse of a continuous

function is continuous and, under mild restrictions, the inverse of a differentiable function is

differentiable.

Theorem: If f is continuous and one to one on an interval, thenf −1 is also continuous.

Proof:

Theorem: Let f be a one-to-one function on an open interval I, and let J = f(I). If f

is differentiable at xo  I and if f ' ( xo )  0 , then f -1 is differentiable at yo =f( xo) and

1
( f −1 )' ( yo ) =
f ' ( xo )

39
 f ( x) − f ( xo )
Proof: f ' ( xo ) = xlim . Since f ' ( xo )  0 and since f ( x)  f ( xo ) for x  xo
→x o x − xo

1 x − xo
we can write = lim . Let ε > 0 , then  δ > 0 such that
f ' ( xo ) x → xo f ( x) − f ( xo )

x − xo 1
if x − xo   , then −  (1)
f ( x) − f ( xo ) f ' ( xo )

Let g = f -1, then since g is continuous at yo, then   0 such that

if y − yo   , then g ( y) − g ( yo )  

 g ( y) − xo   (2)

From (1), (2) we get

g ( y) − xo 1
if y − yo   , then − 
f ( g ( y)) − f ( xo ) f ' ( xo )

g ( y ) − g ( yo ) 1
 − 
y − yo f ' ( xo )

1 g ( y) − g ( yo ) 1
Hence, = lim  g ' ( yo ) = .
f ' ( xo ) y → y o y − yo f ' ( xo )

1
Therefore; (𝑓 −1 )′ (𝑦0 ) =
𝑓′ (𝑥0 )

40
Example

Let 𝑓(𝑥) = 8𝑥 3 − 1. Then find (𝑓 −1 )′ (7).

Solution

Since 𝑓 ′ (𝑥) = 24𝑥 2 ≥ 0 ∀𝑥 ∈ ℜ, 𝑓 is increasing. Hence 𝑓 has an inverse.

Let 𝑓 −1 (7) = 𝑥0 . Then 𝑓(𝑥0 ) = 7 ⟺ 8𝑥03 − 1 = 7 ⟺ 𝑥0 = 1

1 1
By above theorem; (𝑓 −1 )′ (7) = = .
𝑓′ (1) 24

Example

Let 𝑓(𝑥) = 𝑥 7 + 8𝑥 3 + 4𝑥 − 2. Then find (𝑓 −1 )′ (−2).

Solution

Since 𝑓 ′ (𝑥) = 7𝑥 6 + 24𝑥 2 + 4 ≥ 0 ∀𝑥 ∈ ℜ, 𝑓 is increasing. Hence 𝑓 has an

inverse.

Let 𝑓 −1 (−2) = 𝑎. Then 𝑓(𝑎) = 𝑎7 + 8𝑎3 + 4𝑎 − 2 = −2

⟺ 𝑎7 + 8𝑎3 + 4𝑎 = 0

⟺ 𝑎(𝑎6 + 8𝑎2 + 4) = 0

41
 ⟺ 𝑎 = 0, 𝑏𝑒𝑐𝑎𝑢𝑠𝑒 𝑎6 + 8𝑎2 + 4 ≠ 0

1 1
Now (𝑓 −1 )′ (−2) = = .
𝑓′ (−2) 4

Remark: suppose 𝑦 = 𝑓(𝑥) and 𝑓 has an inverse, so 𝑥 = 𝑓 −1 (𝑦) and we have the

relation;

𝑑𝑥 1
=
𝑑𝑦 𝑑𝑦
𝑑𝑥

Example

𝑑𝑥 𝑑𝑥
Let 𝑦 = 𝑥 5 + 2𝑥. Then find and | .
𝑑𝑦 𝑑𝑦 𝑦=−3

Solution:

The function 𝑦 has an inverse because its derivative, 5𝑥 4 + 2, is positive function.

Using the above relation, we deduced that

𝑑𝑥 1 1
= = 4
𝑑𝑦 𝑑𝑦 5𝑥 + 2
𝑑𝑥

Since 𝑦 = 𝑥 5 + 2𝑥, it follows that 𝑦 = −3 for 𝑥 = −1. We conclude that

𝑑𝑥 1 1
|𝑦=−3 = =
𝑑𝑦 5(−1)4 + 2 7
42
Activity

Calculate (𝒇−𝟏 )′ (𝒄);

a. 𝑓 (𝑥) = 𝑥 + 𝑠𝑖𝑛 𝑥; 𝑐 = 0 c. 𝑓(𝑥) = 5𝑥 5 + 4𝑥 3 ; 𝑐 = 9

b. 𝒇(𝒙) = 𝒙 𝐥𝐧 𝒙 ; 𝒄 = 𝟐𝒆𝟐 d. 𝒇(𝒙) = 𝒙𝟑 − 𝟑𝒙𝟐 − 𝟏, 𝒙 ≥ 𝟐 ; 𝒄 = −𝟏

INVERSE TRIGONOMETRIC FUNCTION

The six basic trigonometric functions do not have inverses because their graphs repeat

periodically and hence do not pass the horizontal line test. To circumvent this problem

we will restrict the domains of the trigonometric functions to produce one-to-one functions

and then define the “inverse trigonometric functions” to be the inverses of these restricted

functions.

 
(1) y = sin-1 x = that number y such that sin y = x and - y
2 2
(2) y = cos-1 x = that number y such that cos y = x and 0  y  
 
(3) y = tan-1 x = that number y such that tan y = x and - <y<
2 2
(4) y = cot-1 x = that number y such that cot y = x and 0 < y < 
 3
(5) y = sec-1 x = that number y such that sec y = x and 0  y < or -   y <
2 2
 3
(6) y = csc-1 x = that number y such that csc y = x and 0 < y  or -  < y 
2 2

43
Derivatives and integrals of inverse Trigonometric functions d 1
-1
(1) dx sin x = 1 - x2

Here is a summary of the derivatives of the inverse d -1

-1
(2) dx cos x = 1 - x2
trigonometric functions. Let's see why some of
Proposition 1. Show that d -1
1
(3) dx tan x = 1 + x2
d -1 1
dx (sin x) = 1 - x2 d -1
-1
(4) dx cot x = 1 - x2
d -1 -1
dx (cos x) = 1 - x2 d -1
1
(5) dx sec x = x x2 - 1

d -1
-1
(6) dx csc x = x x2 - 1

44
Proof. Let y = sin-1x so that x = sin y. Use implicit differentiation. Take the derivative of both
dy dy 1 1
sides of this last formula with respect to x to get 1 = (cos y) dx. Thus dx = cos y = =
1 - sin2y
1
which proves the sin-1 formula. The cos-
1 - x2
d 1 du
1  (1) -1
dx sin u(x) =
follows from the sin-1 formula and cos-1x = 2 1 - u(x)2 dx

- sin-1x. // d -1
-1 du
(2) dx cos u(x) = 1 - u(x)2 dx
Using the chain rule we get the following
analogues of (10) – (15) when x is replaced by an (3)
d -1
1 du
dx tan u(x) = 1 + u(x)2 dx
expression involving x.
d - 1 du
dy (4) -1
Example 1. Find dx if y = sin-1(3x – 4) dx cot u(x) = 1 - u(x)2 dx

dy d d -1
1 du
(5) dx sec u(x) = u(x) u(x)2 - 1 dx
Solution. Using (16) one has dx = dx sin-1(3x –

1 d  3 d -1 du
2  -1
4) = (3x - 4) = . (6) dx csc u(x) = u(x) u(x)2 - 1 dx
1 - (3x - 4) dx  1 - (3x - 4)2
dy
Example 2. Find dx if y = sin-1( sin x)

dy d 1 d 
Solution. Using (16) one has dx = dx sin-1( sin x) =  (sin x)1/2 =
2 dx
1 - ( sin x) 

 1  d  cos x
  (1/2) (sin x)-1/2 dx(sin x) = .
 1 - sin x 2 1 - sin x sin x
dy 1 + 2cos x
Example 3. Find dx if y = cos-1 2 + cos x 
 
Solution. Using (17) one has
dy d -11 + 2cos x -1  d 1 + 2cos x
dx = dx cos  2 + cos x  =   
1 + 2cos x dx  2 + cos x 
2
1 -  2 + cos x 
 
- (2 + cos x)
=
(2 + cos x)2 - (1 + 2cos x)2
 d d 
(2 + cos x)dx(1 + 2cos x) - (1 + 2cos x)dx(2 + cos x)
 
 (2 + cos x)2 

45
 - (2 + cos x)
=
4 + 4cos x + cos2 x - 1 - 4cos x - 4cos2 x
(2 + cos x)(- 2sin x) - (1 + 2cos x)(- sin x)
 
 (2 + cos x)2 
-1 - 4 sin x - 2sin x cos x + sin x + 2sin x cos x
= 2  
3 - 3cos x  2 + cos x 

3 sin x 3 sin x 3
= = = 2 + cos x
3 - 3cos2 x (2 + cos x) 2
1 - cos x (2 + cos x)
For the last formula we are assuming x lies in an interval where sin x is non-negative. If x lies in
an interval in which sin x is negative then the answer would
1
- 3
(1)  -1
 1 - x2 dx = sin x
be 2 + cos x.
1
Each of the derivative formulas (10) – (15) has a (2)  1 + x2 dx = tan-1 x

corresponding integration formula. However the integration
1
formulas for the co functions don't give new integration (3) 
 dx = sec-1 x
x x2 - 1
formulas.

Example 4. Find 
1
 6 - 2x2 dx


Solution.  

1 1 1 1 x
 6 - 2x2 dx =  dx = dx. Let u = . Then du =

2
 x 2
 6
6 x 3
1- 3 1- 
   3

dx 1  3 1 1 x
and dx = 3 du and the integral becomes  2 du = sin-1u = sin-1 .
3 6 1-u 2 2  3
Hyperbolic Functions
The hyperbolic functions are combinations of exponential functions that have many similarities to
the trigonometric functions. In fact, for almost every formula involving a trigonometric function
there is an analogous formula involving a hyperbolic function. The hyperbolic functions are useful
for doing integrals and they also arise is the solution of problems involving mechanical vibrations
and electric circuits where one frequently encounters combinations of exponentials such as 2e3x –
5e-2x. For certain calculations it is convenient to express such combinations in terms of the two
special combinations

46
 ex - e-x ex + e-x
and
2 2
These two special combinations are the first two hyperbolic functions. The remaining four are
related to these in the same way as the remaining trigonometric functions are related to the sine
and cosine.

ex - e-x
(1) hyperbolic sine of x = sinh x = 2

ex + e-x
(2) hyperbolic cosine of x = cosh x = 2

sinh x ex - e-x 1 - e-2x

(3) hyperbolic tangent of x = tanh x = cosh x = ex + e-x = 1 + e-2x

1 cosh x ex + e-x
(4) hyperbolic cotangent of x = coth x = tanh x = sinh x = ex - e-x

1 2 Example 1.
(5) hyperbolic secant of x = sech x = cosh x = ex + e-x
e2 - e-2 7.38905 - 0.13533
 2 1 =  23.62686
sinh 2 =
(6)
2
hyperbolic cosecant of x = csch x = sinh x ex - e-x
Here are their graphs. As we go along we will explain why these graphs have the shape they do.
x sinh y x tanh y
x x

10 1.0
x cosh y
x
10
5 0.5

y y
3 2 1 1 2 3 3 2 1 1 2 3
6

5 0.5
4

10 1.0 2

y
3 2 1 1 2 3

d d
(7) dx sinh x = cosh x dx sin x = cos x

d d
(8) dx cosh x = sinh x dx cos x = - sin x

d 2
d 2
(9) dx tanh x = sech x dx tan x = sec x

d 2
d 2
(10) dx coth x = - csch x dx cot x = - csc x

d d
(11) dx sech x = - tanh x sech x dx sec x = tan x sec x
47
d d
(12) dx csch x = - coth x csch x dx csc x = - cot x csc x
We are particularly interested in the derivatives of the hyperbolic functions and integrals involving
the hyperbolic functions.
In order to prove these derivative formulas we will need the following formulas which are similar
to the corresponding formulas for the trigonometric functions.
(13) cosh2x – sinh2x = 1 cos2x + sin2x = 1
(14) sech2x + tanh2x = 1 sec2x - tan2x = 1
(15) coth2x – csch2x = 1 cot2x – csc2x = 1

Proposition 1. (13) – (15) hold.

Proof. To prove (13) one has
x -x x -x
e + e 2 e - e 2 e2x + 2 + e-2x e2x - 2 + e-2x
cosh2x – sinh2x =  2  -  2  = - = 1
    4 4
To prove (14), divide (13) by cosh2x. To prove (15), divide (13) by sinh2x. //
Proposition 2. (7) – (12) hold.
Proof. To prove (7) and (8) one has
d d ex - e-x x -x
e + e 
dx  2   2  = cosh x
sinh x = =
dx  
d d ex + e-x x -x
e - e 
cosh x = 
dx  2   =   = sinh x
dx  2 
Formulas (9) – (12) follow from (7), (8), (14) and
d du
(15) and the rules for derivatives. // (16) dx sinh u(x) = cosh u(x) dx
If we combine the formulas for the derivatives of
d du
the hyperbolic functions with the chain rule we (17) dx cosh u(x) = sinh u(x) dx
get the formulas at the right which are useful for d du
2
(18) dx tanh u(x) = sech u(x) dx
the situation where what is inside a hyperbolic is
an expression u(x) involving x instead of just x. d 2
du
(19) dx coth u(x) = - csch u(x) dx
dy
Example 2. Find dx if
d du
(20) dx sech u(x) = - tanh u(x) sech u(x) dx
a. y = cosh( ln x )
b. y = tan-1( sinh x ) d du
(21) dx csch u(x) = - coth u(x) csch u(x) dx
c. y = tanh( 1 + e2x )
d. y = sinh( sin-1 x )

48
 e. y = coth( sec-1 x )
dy d d sinh( ln x )
For a, using (17) one has dx = dx (cosh( ln x )) = sinh( ln x ) dx ( ln x ) = . For b,
x
dy d
using the chain rule one has dx = dx (tan-1( sinh x )) =
(22)  cosh x dx = sinh x

 1  d  cosh x 
1 + sinh2x dx ( sinh x ) = 1 + sinh2x. For c, using (18)
   
dy d d
(23)  sinh x dx = cosh x

one has dx = dx (tanh( 1 + e2x )) = sech2( 1 + e2x ) dx ( 1

+ e2x ) = 2e2x sech2( 1 + e2x ). (24)  2

 sech x dx = tanh x
Each of the differentiation formulas (7) – (12) has a
(25) 
 csch x dx = - coth x
2

corresponding integration formula. These are given in the

table at the right. 
(26)  tanh x sech x dx = - sech x
cosh( x)
Example 3. Find 
 dx
x (27) 
 coth x csch x dx = - csch x
dx dx
Let u = x = x1/2. Then du = (1/2)x-1/2 dx = . So 2 du = 
. So the integral becomes 
2 x x
cosh( x)
 2 cosh u du = 2 sinh u = 2 sinh( x).
dx = 
x

Problem 1. Find a. 
 sinh(1 + 4x) dx b. 
 tanh x dx

c.  2 2
 s sech (x ) dx d. 
 (sinh x) e
cosh x
dx

Now let's consider why the graphs of the hyperbolic functions look the way they do. First we
consider positive x.
First, note that sinh x > 0 for x > 0 since ex > e-x. Also cosh x > 0 for x > 0 since ex and e-x are
positive for all x. It follows that the remaining four hyperbolic functions are positive for positive
x.
Second, recall that a function with a positive derivative is increasing and a function with a negative
derivative is decreasing. So, it follows from the derivative formulas that sinh x, cosh x and tanh x
are increasing for x > 0 and coth x, sech x and csch x are decreasing for x > 0.
Third, note that as x increases e-x goes to zero. In fact e-4  0.02. So by the time x  4 one has sinh
ex ex
x  2 and coshh x  2 . In particular, sinh x and cosh x approach  as x approaches . On the

49
 1 - e-2x
other hand, it follows from tanh x = 1 + e-2x that tanh x → 1 as x → . From the reciprocal formulas

it follows that coth x → 1 and sech x → 0 and csch x → 0 as x → .

The graphs for negative x follow from the graphs for positive x and the fact that sinh x, tanh x, coth
x and csch x are odd functions and cosh x and sech x are even functions.
Here are the addition formulas for sinh x and cosh x. Note their similarity to the corresponding
formulas for sin x and cos x.
Proposition 3. For all x and y one has
(28) sinh(x + y) = sinh x cosh y + cosh x sinh y compare sin(x + y) = sin x cos y + cos x sin y
(29) cosh(x + y) = cosh x cosh y + sinh x sinh y
with
To prove (28) one has
cos(x + y) = cos x cos y - sin x sin y
x -x y -y x -x y -y
e - e e + e  e + e e - e 
sinh x cosh y + cosh x sinh y =  2  2  +  2  2 
     
ex+y + ex-y - ey-x - e-x-y ex+y - ex-y + ey-x - e-x-y ex+y - e-x-y
= + = = sinh(x + y)
4 4 2
To prove (29) one has
x -x y -y x -x y -y
e + e e + e  e - e e - e 
cosh x cosh y + sinh x sinh y =  2  2  +  2  2 
     
ex+y + ex-y + ey-x + e-x-y ex+y - ex-y - ey-x + e-x-y ex+y + e-x-y
= + = = cosh(x + y)
4 4 2
The following double angle formulas are a corollary. They will be useful when we do integrals.
Proposition 4. For all x one has
(30) sinh 2x = 2 sinh x cosh x compare sin 2x = 2 sin x cos y
2 2
(31) cosh 2x = cosh x + sinh x
with
2 cosh 2x - 1
(32) sinh x = cos 2x = cos2 x - sin2 x
2
cosh 2x + 1
(33) cosh2x = 2
1 - cos 2x
sin2x =
Proof. (30) and (31) follow from (28) and (29) by replacing y by x. (32) and2 (33) follow from
(31) and the identity cosh2x – sinh2x = 1. //
Example 4. A cable hangs from two poles that are 40 feet tall and separated by 100 feet. In the
1 + cos 2x
cos2x = 2
middle the cable is 38 feet off the ground. Find

50
 a. A formula for the height of the cable above the ground for a general point between the two
cables.
b. The angle the cable makes with the horizontal where it meets the pole.
Solution. Choose a coordinate system with the x axis lying along the ground, the y axis pointing
upward and half way between the poles. Then be base of the poles are at (- 50, 0) and (50, 0) and
the tops of the poles are at (- 50, 40) and (50, 40). The midpoint of the cable is at (0, 38). See
picture at right.
In more advanced books it is shown that the equation of the cable is
 x
y = c + a cosha
 
where c and a are numbers depending of heights of the poles and the distance between them.
Plugging in x = 0 and y = 38 we get
38 = c + a
Plugging in x = 50 and y = 40 we get
50
40 = c + a cosh a 
 
Subtracting these two equations we get
50
a cosh a  - a = 2
 
50
If we let u = a then this becomes

u
cosh u = 1 + 25

This is an equation that we can't solve algebraically, so we need to solve it numerically. Using
50
some mathematical software (e.g. www.wolframalpha.com), we get u  0.08000. So a  0.08000

1
 625 and c  38 – 625 = -587. Since a  0.0016, the equation of the cable is approximately

y = 625 cosh(0.0016x) - 587

dy
We need to find dx |x = 50. One has

dy d  x  x 1 x
dx = dx (c + a cosha) = a sinha a = sinha

51
 dy 50
| = sinh  a  = sinh u = sinh(0.08) = 0.08004
dx x = 50  
dy
 = tan-1 dx |x = 50 = tan-1(0.08004) = 0.0798 radians = 4.58

Inverse Hyperbolic Functions

The inverse hyperbolic functions are formed from the hyperbolic functions in the same way the
inverse trigonometric functions are formed from the regular trigonometric functions. The only
difference is the restrictions one places on the hyperbolic functions to make them one-to-one. To
summarize

(34) y = sinh-1 x = that number y such that sinh y = x

(35) y = cosh-1 x = that number y such that y  0 and cosh y = x

(36) y = tanh-1 x = that number y such that tanh y = x

(37) y = coth-1 x = that number y such that coth y = x

(38) y = sech-1 x = that number y such that y  0 and sech y = x

(39) y = csch-1 x = that number y such that csch y = x

As with the inverse trigonometric functions, we get the graphs of the inverse hyperbolic functions
by reflecting the graphs of the hyperbolic functions across the line y = x. Their graphs are the
following. (40) sinh-1 x = ln(x + x2 + 1)

As opposed to the inverse trigonometric (41) cosh-1 x = ln(x + x2 - 1)

functions, there are formulas for the inverse 1 1+x
(42) tanh-1 x = ln
2  1-x 
hyperbolic involving logarithms. At the right are
1  x +1
(43) coth-1 x = ln
these formulas. 2  x - 1

(44) sech-1 x = ln
1 + 1 - x2
 x 

Example 5. Show how to obtain the formula (45) csch-1 x = ln

1 + x2 + 1
 x 
sinh-1 x = ln(x + x2 + 1).
Solution. We use the fact that y = sinh-1x is that number y such that sinh y = x. Using the definition
1 1
y -y ey - ey u-u
e -e y 1
of sinh y this gives x = 2 = 2 . Now let u = e . So x = 2 and 2x = u - u. This implies

52
 2x  4x2 +4
u2 - 2xu - 1 = 0. Using the quadratic formula, this gives u = = x  x2 +1. We choose
2

the + sign since u > 0. So ey = u = x + x2 +1. Taking ln of both sides gives y = ln(x + x2 + 1).
In calculus one of the main applications of hyperbolic functions is to do certain integrals. In order
to see how they are used, we first have to look at the derivative formulas for the inverse hyperbolic
functions.

d -1
1 d -1
1
(46) dx sinh x = dx sin x =
1 + x2 1 - x2

d 1 d -1
(47)comparedx cosh-1 x = -1
dx cos x =
x2 - 1 1 - x2
with
d -1
1 d -1
1
(48) dx tanh x = 1 - x2 dx tan x = 1 + x2

d -1
1 d -1
-1
(49) dx coth x = 1 - x2 dx cot x = 1 - x2

d -1
-1 d -1
1
(50) dx sech x = x 1 - x2 dx sec x = x x2 - 1
d 1
- 1 -1 x =
Example 6. dShow-1that dx sinh d -1
(51) dx csch x = | x| 2 1 + x2 -1
dx csc x = x x2 - 1
1+x
Solution. Let y = sinh-1x so that x = sinh y. Use implicit differentiation. Take the derivative of
dy dy 1
both sides of this last formula with respect to x to get 1 = (cosh y) dx. Thus dx = cosh y. Using

dy 1 1
cosh2x – sinh2x = 1 gives dx = 2 = .
1 + sinh y 1 + x2

6.5 Indeterminate forms and L’Hospital Rule

In general, if we have a limit of the form
𝑓(𝑥)
lim 𝑔(𝑥)
𝑥→𝑎
Where both 𝑓(𝑥) → 0 and 𝑔(𝑥) → 0 as 𝑥 → 𝑎, then this limit may or may not exist and is called
0
an Indeterminate form of type 0.
Similarly, if both 𝑓(𝑥) → ∞(𝑜𝑟 − ∞) and 𝑔(𝑥) → ∞ (𝑜𝑟 − ∞), then the limit may or may not
exist and is called an indeterminate form of type ∞⁄∞.
We introduce a systematic method, known as L’Hospital Rule, for the evaluation of indeterminate
forms.
L’Hospital Rule

53
 Suppose 𝑓 and 𝑔 are differentiable and 𝑔′ (𝑥) ≠ 0 on an open interval I that contains 𝑎
(except possibly at 𝑎). Suppose that
lim 𝑓(𝑥) = 0 and lim 𝑔(𝑥) = 0
𝑥→𝑎 𝑥→𝑎
or that lim 𝑓(𝑥) = ±∞ and lim 𝑔(𝑥) = ±∞
𝑥→𝑎 𝑥→𝑎
0 ∞
(In other words, we have an indeterminate form of type 0 𝑜𝑟 ∞). Then
′
𝑓(𝑥) 𝑓 (𝑥)
lim = lim ′
𝑥→𝑎 𝑔(𝑥) 𝑥→𝑎 𝑔 (𝑥)
If the limit on the right side exists (𝑜𝑟 𝑖𝑠 ∞ 𝑜𝑟 − ∞).
2𝑥 −1
Example 1: Find lim .
𝑥→0 𝑥
Solution
Since lim 2𝑥 − 1 = 0 𝑎𝑛𝑑 lim 𝑥 = 0,
𝑥→0 𝑥→0
we can apply L’Hospital Rule
𝑑 𝑥
2𝑥 − 1 (2 − 1) 2𝑥 𝑙𝑛2
lim = lim 𝑑𝑥 = lim = 𝑙𝑛2
𝑥→0 𝑥 𝑥→0 𝑑 𝑥→0 1
(𝑥)
𝑑𝑥
𝑒𝑥
Example 2: Calculate lim 𝑥 2.
𝑥→∞
Solution
We have lim 𝑒 𝑥 = ∞ 𝑎𝑛𝑑 lim 𝑥 2 = ∞,
𝑥→∞ 𝑥→∞
so L’Hospital Rule gives
𝑒𝑥 𝑒𝑥
lim = lim
𝑥→∞ 𝑥 2 𝑥→∞ 2𝑥
Since 𝑒 𝑥 → ∞ 𝑎𝑛𝑑 2𝑥 → ∞ 𝑎𝑠 𝑥 → ∞, the limit on the right side is also indeterminate, but a
second application of L’Hospital Rule gives
𝑒𝑥 𝑒𝑥 𝑒𝑥
lim = lim = lim =∞
𝑥→∞ 𝑥 2 𝑥→∞ 2𝑥 𝑥→∞ 2
Example 3: Evaluate lim+ 𝑥 ln 𝑥.
𝑥→0
Solution
1
The given limit is indeterminate because 𝑥 → 0+ while ln 𝑥 → −∞. Writing 𝑥 = 1 ,
𝑥
we have 1/𝑥 → ∞ as 𝑥 → 0+ ,
so L’Hospital Rule gives
1
ln 𝑥
lim+ 𝑥 ln 𝑥 = lim+ 1 = lim+ 𝑥
−1
𝑥→0 𝑥→0 𝑥
𝑥→0 𝑥2
= lim+(−𝑥) = 0
𝑥→0
Indeterminate Differences
If lim 𝑓(𝑥) = ∞ and lim 𝑔(𝑥) = ∞, then the limit
𝑥→𝑎 𝑥→𝑎
lim [𝑓(𝑥) − 𝑔(𝑥)]
𝑥→𝑎
Example 4: Compute lim
𝜋
(sec 𝑥 − tan 𝑥)
𝑥→( )−
2
Solution
𝜋
First notice that sec 𝑥 → ∞ and tan 𝑥 → ∞ as 𝑥 → ( 2 )− , so the limit is indeterminate.
Here we use a common denominator:
1 sin 𝑥
lim
𝜋 −
(sec 𝑥 − tan 𝑥) = lim𝜋 −
( − )
cos 𝑥 cos 𝑥
𝑥→( ) 𝑥→( )
2 2

54
 1−sin 𝑥
= lim −
𝑥→𝜋/2 cos 𝑥
−𝑐𝑜𝑠
= lim − − sin 𝑥 = 0
𝑥→𝜋/2
Note that the use of L’Hospital Rule is justified because 1 − sin 𝑥 → 0 and cos 𝑥 → 0 as 𝑥 →
𝜋/2− .
Example 5: Calculate lim+(1 + sin 4𝑥)cot 𝑥 .
𝑥→0
Solution
First notice that as 𝑥 → 0+ . We have 1 + sin 4𝑥 → 1 and cot 𝑥 → ∞, so the given limit is
indeterminate.
Let 𝑦 = (1 + sin 4𝑥)cot 𝑥
Then ln 𝑦 = ln[(1 + sin 4𝑥)cot 𝑥 ] = cot 𝑥 𝑙𝑛(1 + sin 4𝑥)
So, L’Hospital Rule gives
ln(1+sin 4𝑥)
lim+ ln 𝑦 = lim+
𝑥→0 𝑥→0 tan 𝑥
4 cos 4𝑥

= lim+ 1+sin 4𝑥
𝑠𝑒𝑐 2 𝑥
=4
𝑥→0
So far we have computed the limit of ln 𝑦,
but what we want the limit of y.
To find this we use the fact that 𝑦 = 𝑒 ln 𝑦 :
lim+(1 + sin 4𝑥)cot 𝑥 = lim+ 𝑦
𝑥→0 𝑥→0
= lim+ 𝑒 ln 𝑦 = 𝑒 4
𝑥→0

Chapter Three
TECHNIQUES OF INTEGRATION
A function 𝐹(𝑥) + 𝑐, 𝑤ℎ𝑒𝑟𝑒 𝑐 𝑖𝑠 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡, is called an antiderivative of 𝑓 on an interval 𝐼 if 𝐹 ′ (𝑥) = 𝑓(𝑥)
for all 𝑥 in 𝐼. The process of finding the antiderivative of a function is called integration.
Elementary integrations

55
In this chapter we develop techniques for using these basic integration formulas
to obtain indefinite integrals of more complicated functions.
Integration is not as straightforward as differentiation; there are no rules that
absolutely guarantee obtaining an indefinite integral of a function.
Integration by Parts
∫ 𝑓(𝑥)𝑔′ (𝑥)𝑑𝑥 = 𝑓(𝑥)𝑔(𝑥) − ∫ 𝑓 ′ (𝑥)𝑔(𝑥)𝑑𝑥

This is called the formula for integration by parts. It is perhaps easier to remember in the
following notation.
Let 𝑢 = 𝑓(𝑥) and 𝑣 = 𝑔(𝑥). Then 𝑑𝑢 = 𝑓 ′ (𝑥)𝑑𝑥 and 𝑑𝑣 = 𝑔′ (𝑥)𝑑𝑥, so, by substitution rule, the
formula for integration by parts becomes
∫ 𝑢 𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣 𝑑𝑢
Example: Find ∫ 𝑥𝑠𝑖𝑛 𝑥 𝑑𝑥.
Solution
Suppose we choose 𝑓(𝑥) = 𝑥 𝑎𝑛𝑑 𝑔′ (𝑥) = sin 𝑥.
Then 𝑓 ′ (𝑥) = 1 𝑎𝑛𝑑 𝑔(𝑥) = − cos 𝑥.
Thus, using above Formula, we have
∫ 𝑥𝑠𝑖𝑛 𝑥 𝑑𝑥 = 𝑓(𝑥)𝑔(𝑥) − ∫ 𝑓 ′ (𝑥)𝑔(𝑥)𝑑𝑥
= 𝑥(− cos 𝑥) − ∫(− cos 𝑥) 𝑑𝑥
= −𝑥𝑐𝑜𝑠 𝑥 + ∫ cos 𝑥 𝑑𝑥
= −𝑥𝑐𝑜𝑠 𝑥 + sin 𝑥 + 𝐶
Example: Evaluate ∫ ln 𝑥 𝑑𝑥
Solution
Here we do not have much choice for 𝑢 and 𝑑𝑣. Let
𝑢 = ln 𝑥 𝑑𝑣 = 𝑑𝑥
1
Then 𝑑𝑢 = 𝑑𝑥 𝑣=𝑥
𝑥
Integrating by parts, we get
𝑑𝑥
∫ ln 𝑥 𝑑𝑥 = 𝑥 𝑙𝑛 𝑥 − ∫ 𝑥 𝑥
= 𝑥 𝑙𝑛 𝑥 − ∫ 𝑑𝑥
= 𝑥𝑙𝑛 𝑥 − 𝑥 + 𝐶
2 𝑥
Example: Find ∫ 𝑥 𝑒 𝑑𝑥.
Solution
Let 𝑢 = 𝑥2 𝑑𝑣 = 𝑒 𝑥 𝑑𝑥
Then 𝑑𝑢 = 2𝑥 𝑑𝑥 𝑣 = 𝑒𝑥
Integration by parts gives
∫ 𝑥 2 𝑒 𝑥 𝑑𝑥 = 𝑥 2 𝑒 𝑥 − 2 ∫ 𝑥𝑒 𝑥 𝑑𝑥 -------------------- (1)
We use integration by parts a second time to find ∫ 𝑥𝑒 𝑥 𝑑𝑥
This time with 𝑢 = 𝑥 and 𝑑𝑣 = 𝑒 𝑥 𝑑𝑥.
Then 𝑑𝑢 = 𝑑𝑥, 𝑣 = 𝑒 𝑥 , and
∫ 𝑥𝑒 𝑥 𝑑𝑥 = 𝑥𝑒 𝑥 − ∫ 𝑒 𝑥 𝑑𝑥 = 𝑥𝑒 𝑥 − 𝑒 𝑥 + 𝐶
Putting this in (1), we get
∫ 𝑥 2 𝑒 𝑥 𝑑𝑥 = 𝑥 2 𝑒 𝑥 − 2 ∫ 𝑥𝑒 𝑥 𝑑𝑥
= 𝑥 2 𝑒 𝑥 − 2(𝑥𝑒 𝑥 − 𝑒 𝑥 + 𝐶)
= 𝑥 2 𝑒 𝑥 − 2𝑥𝑒 𝑥 + 2𝑒 𝑥 + 𝐶1 𝑤ℎ𝑒𝑟𝑒 𝐶1 = −2𝐶

56
 1
Example: Calculate ∫0 𝑡𝑎𝑛−1 𝑥𝑑𝑥.
Solution
Let 𝑢 = 𝑡𝑎𝑛−1 𝑥 𝑑𝑣 = 𝑑𝑥
𝑑𝑥
Then 𝑑𝑢 = 𝑣=𝑥
1+𝑥 2
By using the Formula, we get
1 1 1 𝑥
∫0 𝑡𝑎𝑛−1 𝑥𝑑𝑥 = 𝑥𝑡𝑎𝑛−1 𝑥] 0 − ∫0 1+𝑥2 𝑑𝑥
1 𝑥
= 1 ∙ 𝑡𝑎𝑛−1 1 − 0 ∙ 𝑡𝑎𝑛−1 0 − ∫0 𝑑𝑥
1+𝑥 2
𝜋 1 𝑥
= 4 − ∫0 1+𝑥2 𝑑𝑥
To evaluate this integral we use the substitution 𝑡 = 1 + 𝑥 2
Then 𝑑𝑡 = 2𝑥𝑑𝑥, 𝑠𝑜 𝑥𝑑𝑥 = 𝑑𝑡/2. When 𝑥 = 0, 𝑡 = 1; when 𝑥 = 1, 𝑡 = 2; so
1 𝑥 1 2 𝑑𝑡 1 2
∫0 1+𝑥 2 𝑑𝑥 = 2 ∫1 𝑡 = 2 ln |𝑡|] 1
1 1
= 2 (ln 2 − ln 1) = 2 ln 2
1 𝜋 1 𝑥 𝜋 ln 2
Therefore ∫0 𝑡𝑎𝑛−1 𝑥𝑑𝑥 = 4 − ∫0 1+𝑥2 𝑑𝑥 = 4 − 2
3.1. Integration by Substitution
When the integration process is not immediately obvious, it may be possible to ‘reduce’ the integral to a
well-known form by means of substitution.

This method is the first to be considered, whenever we try to obtain any integral. The purpose of this
method is to change the integrand into an expression of basic integral forms. In principle, the process of
integration by substitution can be done through five steps as follows:

Step 1: Make appropriate choice of u, let u = g (x)

du
Step 2: Obtain = g ' ( x)
dx

Step 3: Substitute

u = g (x) , du = g ' ( x)dx

After this stage, the whole integral must be in terms of u. This means

no more term in x can remain. If this step fails, we need to make

another appropriate choice of u.

57
Step 4: Evaluate the integral obtained in terms of u.

Step 5: Substitute u with g(x), so that the final answer will be in terms of x.

The method of substitution is widely used with a variety of integrals, but all such integrals have one thing
in common: Within the integrand appear both a function and (some multiple of) its derivative.

(i) Function of a Linear Function of x

We are very required to integrate functions like those in standard list, but where x is replaced by a linear

 (5x − 4)  x dx except that x is replaced by (5x − 4) . If

6 6
function of x, e.g. dx , which is very much


we put u to stand for (5 x − 4) , the integral becomes u dx and therefore we can complete the
6

operation, we must change the variables, thus:

u 
dx
6
dx = u 6 du
du

dx du dx 1
Now can be found from substitution u = (5x − 4) for = 5 , therefore = and the integral
du dx du 5
becomes:

61 1 u7
   
dx 1 6
u dx = u
6
du = u  du =
6
u du =  +c
du 5 5 5 7

Finally, we must express u in terms of the original variables, x, so that:

(5 x − 4) 7 (5 x − 4) 7
 (5 x − 4) 6 dx =
57
+c =
35
+c

58
Example

Evaluate the following integrals by using the substitution u = 4 x + 1 .

 (4 x + 1)  (4 x + 1)
2
(a) dx (b) dx

Solutions:

du 1
(a) If u = 4 x + 1 , then = 4  du = dx . Substituting u and du give us
dx 4
1
 (4 x + 1) dx = 4  u du
1
= u2 + c
8
1
= (4 x + 1) 2 + c
8

By integrating directly will get

Remarks

4x 2
 (4 x + 1) dx = 2 + x + c = 2 x + x + c
2

Both answer are actually the same, because

1 1
(4 x + 1) 2 + c = (16 x 2 + 8 x + 1) + c
8 8

1
= 2x 2 + x + +c
8
1
= 2 x 2 + x + C, where C = +c
8

59
 du 1
(b) If u = 4 x + 1 , then = 4  du = dx . Substituting u and du give us
dx 4

1 2
 (4 x + 1) dx =
4
2
u du

1 3
= u +c
12
1
= (4 x + 1) 3 + c
12

Remarks
By integrating directly will get

16 x 3
 (4 x + 1) dx =  (16 x + 8x + 1) dx = 3 + 4 x + x + C
2 2

Which is equivalent to the previous because

1 1
(4 x + 1) 3 + c = (64 x 3 + 48 x 2 + 12 x + 1) + c
12 12

16 3 1
= x + 4x 2 + x + + c
3 12
16 1
= x 3 + 4 x 2 + x + C, where C = +c
3 12

Example

Evaluate the following integrals.

60
 5
c 1
 ax + b dx  e dx 
(c) (10 x − 9) 2 dx   4x + 3
4x
(a) (b) (d) 35 x dx (e) dx

Solutions

(a) If u = ax + b , then (b) If u = 4 x , then (c) If u = ax + b , then

du 1 du 1 du 1
= a  du = dx = 4  du = dx = a  du = dx
dx a dx 4 dx a
c 1 1 c 1 1
 ax + b dx = c  u  a du 1  ax + b dx = c  u  a du
e dx =  e u  du
4x

c 1 4 c 1
=  du 1 4x =  du
a u = e +c a u
4
c c
= ln u + c = ln ax + b + c
a a

Exercise Evaluate the following

(a) e
−3 x +5
dx 
(b) sec 2 (3x + 1) dx (c)
 cos(1 + 4x) dx

f ' ( x)
(ii) Integral of the form dx
f ( x)

2x + 3
Consider the integral
x 2
+ 3x − 5
dx . This is not one of our standard integrals, so how shall we

tackle it? This is an example of a type of integral which is very easy to deal with but which depends
largely on how keen your wits are. You will notice that if we differentiate the denominator, we obtain
the expression in the numerator. So, let u stand for the denominator, i.e. u = x 2 + 3x − 5 .

du
 = 2x + 3  du = (2 x + 3)dx
dx

The given integral can then be written in terms of u:

2x + 3
 
1
dx = du = ln u + c
x + 3x − 5
2
u

61
 If we now put back what u stands for in terms of x, we get

2x + 3
x 2
+ 3x − 5
dx = ln( x 2 + 3x − 5) + c

Example

Evaluate the following indefinite integrals by using the substitution method.

2x + 3 x 12 x 2 + 16
(a)  x 2 + 3x + 2 dx (b)  x 2 − 4 dx (c)  x 3 + 4 x dx

Solutions:

du
(a) If u = x 2 + 3x + 2 , then = 2 x + 3  du = (2 x + 3)dx
dx
2x + 3 1
x 2
+ 3x + 2
dx =   du = ln u + c = ln( x 2 + 3x + 2) + c
u

du 1
(b) If u = x 2 − 4 , then = 2 x  du = xdx
dx 2
x 1 1 1 1 1 1
 x 2 − 4 dx =  u  2 du = 2  u du = 2 ln u + c = 2 ln( x − 4) + c
2

(iii) Integral of the form

 f ( x) f ' ( x) dx

Example

62
Evaluate the following integrals.

 4 x(2 x  e (e 
ln x
(a)
2
− 3) 6 dx (b)
x x
− 1) 3 dx (c) dx
x

  2x( x + 2)
dx 5
(d) (e) dx
x ln x

Solutions:

du
(a) If u = 2 x 2 − 3 , then = 4 x  du = 4 x dx
dx
u7
 
1
4 x(2 x 2 − 3) 6 dx = u 6 du = + c = (2 x 2 − 3)7 + c
7 7

du
(b) If u = e x − 1 , then = e x  du = e x dx
dx
u4
 
1
e x (e x − 1) 3 dx = u 3 du = + c = (e x − 1) 4 + c
4 4

du 1 1
(c) If u = ln x , then =  du = dx
dx x x
u2
 
ln x 1 1
dx = ln x  dx =  u du = + c = (ln x) 2 + c
x x 2 2

du 1 1
(d) If u = ln x , then =  du = dx
dx x x

  
dx 1 1 1
=  dx =  du = ln u + c = ln(ln x) + c
x ln x ln x x u

du
(e) If u = x + 2 , then = 1  du = dx . Also x = u − 2
dx

63
  2x( x + 2) 
dx = 2(u − 2)u 5 du =  (2u − 4)u 5 du =  (2u 6 − 4u 5 )du
5

2u 7 4u 6 2u 7 2u 6
= − +c = − +c
7 6 7 3

Trigonometric Substitution
Big idea: A lot of wicked-looking integrals can be computed using trigonometric identities and
substitutions.

Big skill:. You should be able to find the antiderivative of products of powers of sine and cosine
functions and tangent and secant functions, as well as make an appropriate substitution for integrands
of the form a 2  x 2 and x2 − a2 .

 sin ( x ) cos ( x ) dx .
m n
I. Integrals of the form

1. For integrals of this form, you will either integrate by substitution, letting
u ( x ) = sin ( x ) or u ( x ) = cos ( x ) , or you will use a reduction formula. Which path you
take depends on whether m or n is even or odd…
m n u(x)
Odd Even u ( x ) = cos ( x ) ; see Case 1
Odd Odd Either one; see Case 1
Even Odd u ( x ) = sin ( x ) ; see Case 1
Even Even Use reduction formulas; see Case 2

2. Case 1: m or n an odd positive integer.

i. Isolate one factor of the odd power to serve as the differential.
ii. Replace the remaining even powers using the Pythagorean identity:
sin 2  + cos2  = 1
iii. Integrate by substitution.
iv. Practice: sin 4 ( x ) cos3 ( x ) dx =


64
3. Case 2: m and n both even positive integers.
i. Use the power reduction identities:
1
(1 − cos ( 2 x ) ) or cos 2 ( x ) = (1 + cos ( 2 x ) ) .
sin 2 ( x ) =
2
1
2
ii. Practice:  sin ( x ) dx =
2

iii. Practice: sin 4 ( x ) dx =



iv. Practice: sin 2 ( x ) cos2 ( x ) dx =



4. Don’t forget: The reduction formulas:

1 n −1
 cos ( x ) dx = n cos ( x ) sin ( x ) + n  cos ( x ) dx
n n −1 n−2
i.

1 n −1
 sin ( x ) dx = − n sin ( x ) cos ( x ) + n  sin ( x ) dx
n n −1 n−2
ii.

65
  tan ( x ) sec ( x ) dx .
m n
II. Integrals of the form

1. For integrals of this form, you will either integrate by substitution, letting
u ( x ) = tan ( x ) or u ( x ) = sec ( x ) , or you will use a reduction formula. Which path you
take depends on whether m or n is even or odd…
m n u(x)
Odd Even u ( x ) = sec ( x ) ; see Case 1
Or u ( x ) = tan ( x ) ; see Case 2
Odd Odd u ( x ) = sec ( x ) ; see Case 1
Even Odd Use reduction formulas; see Case 3
Even Even u ( x ) = tan ( x ) ; see Case 2

2. Case 1: m is an odd positive integer.

i. Isolate one factor of sec(x)tan(x) for the differential.
ii. Replace the remaining even powers of tan(x) using the Pythagorean identity:
tan 2 x = sec2 x − 1 .
iii. Integrate by substitution ( u(x) = sec(x) ).
 tan ( x ) sec ( x ) dx =
3 2
iv. Practice:

3. Case 2: n is an even positive integer.

i. Isolate one factor of sec2(x) for the differential.
ii. Replace the remaining even powers of sec(x) using the Pythagorean identity:
sec2 x = 1 + tan 2 x
iii. Integrate by substitution ( u(x) = tan(x) ).
 tan ( x ) sec ( x ) dx =
4 4
iv. Practice:

4. Case 3: m is an even positive integer and n is an odd positive integer.

66
 i. Replace the even powers of tan(x) using the Pythagorean identity:
tan 2 x = sec2 x − 1 .
ii. Use the following reduction formula to integrate the powers of sec(x):
1 n−2
 sec ( x ) dx = n − 1 sec ( x ) tan ( x ) + n − 1  sec ( x ) dx
n n−2 n−2

(This can be proven using integration by parts where

u(x) = secn-2(x) and dv = sec2(x). )
iii. Since n is odd, you will always be left with having to integrate sec(x), which you
can be done using the following formula:
sec ( x ) + tan ( x )
 sec ( x ) dx =  sec ( x )  sec ( x ) + tan ( x )
dx

sec2 ( x ) + tan 2 ( x )
= dx
sec ( x ) + tan ( x )
1 d
=  sec ( x ) + tan ( x )  dx
sec ( x ) + tan ( x ) dx
= ln sec ( x ) + tan ( x ) + c

 tan ( x ) sec ( x ) dx =
2 3
iv. Practice:

Expression Substitution Identity

𝜋 𝜋
𝑥 = a sin 𝜃, −2≤𝜃≤2
√𝑎2 − 𝑥 2 𝜋 𝜋 1 − 𝑠𝑖𝑛2 𝜃 = 𝑐𝑜𝑠 2 𝜃
𝑥 = 𝑎 tan 𝜃 , −2 < 𝜃 < 2
√𝑎2 + 𝑥 2 𝜋 1 + 𝑡𝑎𝑛2 𝜃 = 𝑠𝑒𝑐 2 𝜃
𝑥 = 𝑎 sec 𝜃, 0 ≤ 𝜃 𝑜𝑟 𝜋 ≤𝜃<
√𝑥 2 − 𝑎2 2 𝑠𝑒𝑐 2 𝜃 − 1 = 𝑡𝑎𝑛2 𝜃
3𝜋
2
√9−𝑥 2
Example: Evaluate ∫ 𝑥 2 𝑑𝑥.
𝜋 𝜋
Solution: Let 𝑥 = 3 sin 𝜃, where − 2 ≤ 𝜃 ≤ 2 .
Then 𝑑𝑥 = 3 cos 𝜃 𝑑𝜃 and √9 − 𝑥 2 = √9 − 9𝑠𝑖𝑛2 𝜃 = √9𝑐𝑜𝑠 2 𝜃
= 3| cos 𝜃| = 3 cos 𝜃
𝜋 𝜋
Note that cos 𝜃 ≥ 0 because − 2 ≤ 𝜃 ≤ 2
Thus the Inverse Substitution Rule gives
√9−𝑥 2 3 𝑐𝑜𝑠𝜃
∫ 𝑥2
𝑑𝑥 = ∫ 9 𝑠𝑖𝑛2 𝜃 3𝑐𝑜𝑠𝜃𝑑𝜃
𝑐𝑜𝑠2 𝜃
= ∫ 𝑠𝑖𝑛2 𝜃 𝑑𝜃 = ∫ 𝑐𝑜𝑡 2 𝜃𝑑𝜃
= ∫(𝑐𝑠𝑐 2 𝜃 − 1)𝑑𝜃
= −𝑐𝑜𝑡𝜃 − 𝜃 + 𝐶 ----------------- (1)
by using trigonometric identities,
√9−𝑥 2
cot 𝜃 = 𝑥
since sin 𝜃 = 𝑥/3,
𝑥
𝜃= 𝑠𝑖𝑛−1 ( 3 ) since sin 𝜃 = 𝑥/3,
Putting above values in (1), we get
√9−𝑥 2 √9−𝑥 2 𝑥
∫ 𝑥2
𝑑𝑥 =− 𝑥
− 𝑠𝑖𝑛−1 (3) + 𝐶

67
 𝑥
Example: Find ∫ 2 𝑑𝑥.
√𝑥 +4
Solution: It would be possible to use the trigonometric substitution 𝑥 = 2 𝑡𝑎𝑛 𝜃.
But the direct substitution 𝑢 = 𝑥 2 + 4 is simpler, because then 𝑑𝑢 = 2𝑥𝑑𝑥 and
𝑥 1 𝑑𝑢
∫ 2 𝑑𝑥 = ∫ = √𝑢 + 𝐶 = √𝑥 2 + 4 + 𝐶.
2
√𝑥 +4 √𝑢
𝑥
Example: Evaluate ∫ 𝑑𝑥.
√3−2𝑥−𝑥 2
Solution: We can transform the integrand into a function for which trigonometric substitution is
appropriate by first completing the square under the root sign:
3 − 2𝑥 − 𝑥 2 = 3 − (𝑥 2 + 2𝑥) = 3 + 1 − (𝑥 2 + 2𝑥 + 1)
= 4 − (𝑥 + 1)2
This suggests that we make the substitution 𝑢 = 𝑥 + 1.
Then 𝑑𝑢 = 𝑑𝑥 and 𝑥 = 𝑢 − 1, so
𝑥 𝑢−1
∫ 2
𝑑𝑥 = ∫ 2
𝑑𝑢
√3−2𝑥−𝑥 √4−𝑢
We now substitute 𝑢 = 2 sin 𝜃, giving 𝑑𝑢 = 2 cos 𝜃 and √4 − 𝑢2 = 2 cos 𝜃, so
𝑥 2 sin 𝜃−1
∫ 2
𝑑𝑥 = ∫ 2 cos 𝜃 2 cos 𝜃𝑑𝜃
√3−2𝑥−𝑥
= ∫(2 sin 𝜃 − 1)𝑑𝜃
= −2 cos 𝜃 − 𝜃 + 𝐶
𝑢
= −√4 − 𝑢2 − 𝑠𝑖𝑛−1 (2 ) + 𝐶
𝑥+1
= −√3 − 2𝑥 − 𝑥 2 − 𝑠𝑖𝑛−1 ( 2
)+ 𝐶
Integration of Rational functions by Partial Fractions
Let us consider a rational function
𝑃(𝑥)
𝑓(𝑥) =
𝑄(𝑥)
where 𝑃(𝑥) and 𝑄(𝑥) are polynomials. It is possible to express f as a sum of simpler fractions
provided that the degree of P is less than the degree of Q. Such a rational function is called
proper.
If 𝑃(𝑥) = 𝑎𝑛 𝑥 𝑛 + 𝑎𝑛−1 𝑥 𝑛−1 + ⋯ + 𝑎1 𝑥 + 𝑎0
Where 𝑎𝑛 ≠ 0, then the degree of P is 𝑛 and we write deg(𝑃) = 𝑛.
If 𝑓 is improper, that is, deg(𝑃) ≥ deg(𝑄), then we divide Q into P until a remainder 𝑅(𝑥) is
obtained such that deg(𝑅) < deg(𝑄). the division statement is
𝑃(𝑥) 𝑅(𝑥)
𝑓(𝑥) = 𝑄(𝑥) = 𝑆(𝑥) + 𝑄(𝑥)
where S and R are also polynomials.
𝑥 3 +𝑥
Example: Find ∫ 𝑥−1 𝑑𝑥.
Solution
Since the degree of the numerator is greater than the degree of the denominator, we first perform
the long division. This enables us to write
𝑥 3 +𝑥 2
∫ 𝑥−1
𝑑𝑥 = ∫(𝑥 2 + 𝑥 + 2 + 𝑥−1)𝑑𝑥
𝑥3 𝑥2
= 3
+ 2
+ 2𝑥 + 2 ln|𝑥 − 1| + 𝐶
𝑥 2 +2𝑥−1
Example: Evaluate ∫ 2𝑥 3 +3𝑥 2 −2𝑥 𝑑𝑥.
Solution
Since the degree of the numerator is less than the degree of the denominator, we do not need to
divide. We factor the denominator as
2𝑥 3 + 3𝑥 2 − 2𝑥 = 𝑥(2𝑥 2 + 3𝑥 − 2) = 𝑥(2𝑥 − 1)(𝑥 + 2)
Since the denominator has three distinct linear factors, the partial fraction decomposition of the
integrand (2) has the form

68
 𝑥 2 +2𝑥−1 𝐴 𝐵 𝐶
𝑥(2𝑥−1)(𝑥+2)
= 𝑥 + 2𝑥−1 + 𝑥+2 ----------------- (1)
To determine the values of A, B, and C, we multiply both sides of this equation by
𝑥(2𝑥 − 1)(𝑥 + 2), obtaining
𝑥 2 + 2𝑥 − 1 = 𝐴(2𝑥 − 1)(𝑥 + 2) + 𝐵𝑥(𝑥 + 2) + 𝐶𝑥(2𝑥 − 1) ------------ (2)
Expanding the right side of Equation 4 and writing it in the standard form for polynomials, we get
𝑥 2 + 2𝑥 − 1 = (2𝐴 + 𝐵 + 2𝐶)𝑥 2 + (3𝐴 + 2𝐵 − 𝐶)𝑥 − 2𝐴 ------------- (3)
The polynomials in Equation (3) are identical, so their coefficients must be equal.
2𝐴 + 𝐵 + 2𝐶 = 1
3𝐴 + 2𝐵 − 𝐶 = 2
−2𝐴 = −1
1 1 1
Solving, we get 𝐴 = 2 , 𝐵 = 5 𝑎𝑛𝑑 𝐶 = 10, and so
𝑥 2 +2𝑥−1 11 1 1 1 1
∫ 2𝑥 3 +3𝑥2 −2𝑥 𝑑𝑥 = ∫ [2 𝑥 + 5 2𝑥−1 − 10 𝑥+2] 𝑑𝑥
1 1 1
= 2 ln|𝑥| + 10 ln|2𝑥 − 1| − 10 ln|𝑥 + 2| + 𝐾
𝑥 4 −2𝑥 2 +4𝑥+1
Example: Find ∫ 𝑥 3 −𝑥 2 −𝑥+1 𝑑𝑥.
Solution
The first step is to divide. The result of long division is
𝑥 4 −2𝑥 2 +4𝑥+1 4𝑥
= 𝑥 + 1 + 𝑥 3 −𝑥2 −𝑥+1
𝑥 3 −𝑥 2 −𝑥+1
The second step is to factor the denominator 𝑄(𝑥) = 𝑥 3 − 𝑥 2 − 𝑥 + 1.
Since 𝑄(1) = 0, we know that 𝑥 − 1 is a factor and we obtain
𝑥 3 − 𝑥 2 − 𝑥 + 1 = (𝑥 − 1)(𝑥 2 − 1) = (𝑥 − 1)(𝑥 − 1)(𝑥 + 1) = (𝑥 − 1)2 (𝑥 + 1)
Since the linear factor 𝑥 − 1 occurs twice, the partial fraction decomposition is
4𝑥 𝐴 𝐵 𝐶
(𝑥−1)2
= + 2 +
(𝑥+1) 𝑥−1 (𝑥−1) 𝑥+1
Multiplying by (𝑥 − 1)2 (𝑥 + 1), we get
4𝑥 = 𝐴(𝑥 − 1)(𝑥 + 1) + 𝐵(𝑥 + 1) + 𝐶(𝑥 − 1)2
= (𝐴 + 𝐶)𝑥 2 + (𝐵 − 2𝐶)𝑥 + (−𝐴 + 𝐵 + 𝐶)
Now we equate coefficients:
𝐴 + 𝐶=0
𝐵 − 2𝐶 = 4
−𝐴 + 𝐵 + 𝐶 = 0
Solving, we obtain 𝐴 = 1, 𝐵 = 2 and 𝐶 = −1, so
𝑥 4 −2𝑥 2 +4𝑥+1 1 2 1
∫ 𝑥 3 −𝑥 2 −𝑥+1
𝑑𝑥 = ∫ 𝑥 + 1 + 𝑥−1 + (𝑥−1)2 − 𝑥+1 ]𝑑𝑥
𝑥2 2
= 2
+ 𝑥 + ln|𝑥 − 1| − 𝑥−1 − ln|𝑥 + 1| + 𝐾
𝑥2 2 𝑥−1
= 2
+ 𝑥 − 𝑥−1 + ln |𝑥+1| + 𝐾
4𝑥 2 −3𝑥+2
Example: Evaluate ∫ 4𝑥 2 −4𝑥+3 𝑑𝑥.
Solution
Since the degree of the numerator is not less than the degree of the denominator, we first divide
and obtain
4𝑥 2 −3𝑥+2 𝑥−1
= 1 + 4𝑥2 −4𝑥+3
4𝑥 2 −4𝑥+3
Notice that the quadratic 4𝑥 2 − 4𝑥 + 3 is irreducible because its discriminant is 𝑏 2 − 4𝑎𝑐 =
−32 < 0. This means it cannot be factored, so we do not need to use the partial fraction
technique.
To integrate the given function we complete the square in the denominator:
4𝑥 2 − 4𝑥 + 3 = (2𝑥 − 1)2 + 2

69
let 𝑢 = 2𝑥 − 1 then 𝑑𝑢 = 2 𝑑𝑥 and 𝑥 = (𝑢 + 1)/2, so
4𝑥 2 −3𝑥+2 𝑥−1
∫ 4𝑥 2 −4𝑥+3 𝑑𝑥 = ∫(1 + 4𝑥 2 −4𝑥+3)𝑑𝑥
1
1 (𝑢+1)−1
= 𝑥 + 2∫2 𝑢2 +2
𝑑𝑢
1 𝑢−1
=𝑥+ ∫
4 𝑢2 +2
𝑑𝑢
1 𝑢 1 1
=𝑥+ ∫
4 𝑢2 +2
𝑑𝑢 − 4 ∫ 𝑢2 +2 𝑑𝑢
1 1 1 𝑢
=𝑥+ ln(𝑢2 + 2) − ∙ 𝑡𝑎𝑛−1 ( ) + 𝐶
8 4 √2 √2
1 1 2𝑥−1
=𝑥+ 8
ln(4𝑥 2 − 4𝑥 + 3) − 4 2 𝑡𝑎𝑛−1 ( 2 ) + 𝐶
√ √
1−3𝑥+2𝑥 2 −𝑥 3
Example: Evaluate ∫ 𝑥(𝑥 2 +1)2 𝑑𝑥
Solution
The form of the partial fraction decomposition is
1−3𝑥+2𝑥 2 −𝑥 3 𝐴 𝐵𝑥+𝐶 𝐷𝑥+𝐸
𝑥(𝑥 2 +1)2
= 𝑥 + 𝑥 2 +1 + (𝑥 2 +1)2
Multiplying by 𝑥(𝑥 2 + 1) , we have 2

−𝑥 3 + 2𝑥 2 − 3𝑥 + 1 = 𝐴(𝑥 2 + 1)2 + (𝐵𝑥 + 𝐶)𝑥(𝑥 2 + 1) + (𝐷𝑥 + 𝐸)𝑥

= 𝐴(𝑥 4 + 2𝑥 2 + 1) + 𝐵(𝑥 4 + 𝑥 2 ) + 𝐶(𝑥 3 + 𝑥) + 𝐷𝑥 2 + 𝐸𝑥
= (𝐴 + 𝐵)𝑥 4 + 𝐶𝑥 3 + (2𝐴 + 𝐵 + 𝐷)𝑥 2 + (𝐶 + 𝐸)𝑥 + 𝐴
If we equate coefficients, we get the system
𝐴 + 𝐵 = 0 𝐶 = −1 2𝐴 + 𝐵 + 𝐷 = 2 𝐶 + 𝐸 = −3 𝐴 = 1
Which has the solution A= 1, 𝐵 = −1, 𝐶 = −1, 𝐷 = 1, and 𝐸 = −2. Thus
1−3𝑥+2𝑥 2 −𝑥 3 1 𝑥+1 𝑥−2
∫ 𝑑𝑥 = ∫( − + 2 2 ) 𝑑𝑥
𝑥(𝑥 2 +1)2 𝑥 𝑥 2 +1 (𝑥 +1)
𝑑𝑥 𝑥 𝑑𝑥 𝑥𝑑𝑥 𝑑𝑥
= ∫ 𝑥 − ∫ 𝑥 2 +1 𝑑𝑥 − ∫ 𝑥 2 +1 + ∫ (𝑥 2 +1)2 − 2 ∫ (𝑥2 +1)2
1 1 𝑑𝑥
= ln|𝑥| − ln(𝑥 2 + 1) − 𝑡𝑎𝑛−1 𝑥 − 2 − 2 ∫ 2 2
2 2(𝑥 +1) (𝑥 +1)
2
To evaluate the final integral we substitute 𝑥 = tan 𝜃. Then we have 𝑑𝑥 = 𝑠𝑒𝑐 𝜃𝑑𝜃 and
𝑥 2 + 1 = 𝑠𝑒𝑐 2 𝜃, so
𝑑𝑥 𝑠𝑒𝑐 2 𝜃
∫ (𝑥 2 +1)2 = ∫ 𝑠𝑒𝑐 4 𝜃 𝑑𝜃 = ∫ 𝑐𝑜𝑠 2 𝜃𝑑𝜃
1 𝜃 sin 2𝜃
= ∫(1 + cos 2𝜃)𝑑𝜃 = + +𝐾
2 2 4
𝜃 sin 𝜃 cos 𝜃
= 2
+ 2
+𝐾
1 𝑥 1
= 2
(𝑡𝑎𝑛−1 𝑥 + 2 ∙ 2 ) + 𝐾
√𝑥 +1 √𝑥 +1
1 𝑥
= 2
(𝑡𝑎𝑛−1 𝑥 + 𝑥 2 +1) + 𝐾
Thus
1−3𝑥+2𝑥 2 −𝑥 3 1 1 𝑥
∫ 𝑑𝑥 = ln|𝑥| − ln(𝑥 2 + 1) − 𝑡𝑎𝑛−1 𝑥 − − 𝑡𝑎𝑛−1 𝑥 − + 𝐾1
𝑥(𝑥 2 +1)2 2 2(𝑥 2 +1) 𝑥 2 +1
|𝑥| 2𝑥+1
= ln − 2𝑡𝑎𝑛−1 𝑥 − + 𝐾1 where 𝐾1 = −2𝐾
√𝑥 2 +1 2(𝑥 2 +1)

Improper Integrals
Definite integrals when either the integrands or the interval of integrations are unbounded are
called improper integrals, as opposed to integrals of continuous functions over bounded intervals,
which are called proper integrals.
Integrals with unbounded integrands

70
We say that f is unbounded near c if f is unbounded either on every open interval of the form (c,
1 1
x) or on every open interval of the form (x, c). For example, 𝑥 and 𝑥 are unbounded near 0.
√
We now consider a function f that is continuous at every point in (a, b] and unbounded near a.
𝑏
By assumption f is continuous on the interval [c, b] for any c in (a, b), so that ∫𝑐 𝑓(𝑥)𝑑𝑥 is
𝑏 𝑏
defined for such c. If the one sided limit lim ∫𝑐 𝑓(𝑥)𝑑𝑥 exists, then we define ∫𝑎 𝑓(𝑥)𝑑𝑥 by the
𝑐−𝑎
𝑏 𝑏 𝑏
formula ∫𝑎 𝑓(𝑥)𝑑𝑥 = lim ∫𝑐 𝑓(𝑥)𝑑𝑥 and say that the integral ∫𝑎 𝑓(𝑥)𝑑𝑥 converges. Otherwise
𝑐−𝑎
𝑏
we say that ∫𝑎 𝑓(𝑥)𝑑𝑥 diverges and do not assign any number to the integral.
1 1
Example 1: Show that ∫0 𝑑𝑥 converges and compute its value.
√𝑥
1 1 1
Solution: is continuous at every point in (0, 1] and is unbounded near 0. Thus ∫0 𝑑𝑥 =
√𝑥 √𝑥
1
lim ∫𝑐 𝑓(𝑥)𝑑𝑥 , provided that the limit exists. Now for 0 < 𝑐 < 1 we have
𝑐→0
1 1
∫𝑐 𝑥 𝑑𝑥 = 2(1 − √𝑐) and lim 2(1 − √𝑐) = 2
√ 𝑐→0
1 1 1 1
Consequently ∫0 𝑑𝑥 converges, and ∫0 𝑑𝑥 = 2
√𝑥 √𝑥
This means that the shaded region has area 2.

x
1

11
Example 2: Show that ∫0 𝑑𝑥 diverges.
𝑥
If f is continuous at every point of (a, b) and is unbounded near both a and b, we say that
𝑏
∫𝑎 𝑓(𝑥)𝑑𝑥
𝑑 𝑏
Converges if for some point d in (a, b) both the integrals ∫𝑎 𝑓(𝑥)𝑑𝑥 and ∫𝑑 𝑓(𝑥)𝑑𝑥 converge.
𝑏
Otherwise, we say that ∫𝑎 𝑓(𝑥)𝑑𝑥 diverges.
Integrals over unbounded intervals
∞ 𝑏
Integrals of the form ∫𝑎 𝑓(𝑥)𝑑𝑥 and ∫−∞ 𝑓(𝑥)𝑑𝑥 are also called improper integrals.If f is
𝑏
continuous on [a,∞ ), then ∫𝑎 𝑓(𝑥)𝑑𝑥 is a proper integral for any 𝑏 ≥ 𝑎.This fact allows us to say
∞ 𝑏
that ∫𝑎 𝑓(𝑥)𝑑𝑥 converges if lim ∫𝑎 𝑓(𝑥)𝑑𝑥 exists. In that event,
𝑏−∞
∞ 𝑏
∫ 𝑓(𝑥)𝑑𝑥 = lim ∫ 𝑓(𝑥)𝑑𝑥
𝑎 𝑏−∞ 𝑎
𝑏
As before, the integral diverges otherwise. The improper integral ∫−∞ 𝑓(𝑥)𝑑𝑥 is handled in an
analogous way.
∞ 1
Example: Show that ∫0 𝑥 2 𝑑𝑥 diverges.
1 1 ∞ 1
Solution: Let 𝑑 = 1. We will show that ∫0 𝑥2
𝑑𝑥, and hence ∫0 𝑥2
𝑑𝑥 ,diverges.
1 1 1 1
To determine that ∫0 𝑥 2 𝑑𝑥 diverges we compute ∫𝑐 𝑥 2 𝑑𝑥 for any c in (0, 1). Since

71
 1
1 1
∫ 2
𝑑𝑥 = − 1
𝑐 𝑥 𝑐
1 1 1 ∞ 1
And since lim 𝑐 − 1 = ∞, we know that ∫0 𝑥 2 𝑑𝑥 diverges. This implies that ∫0 𝑥 2 𝑑𝑥 diverges.
𝑐−0
∞ 𝑑 ∞
We say that ∫−∞ 𝑓(𝑥)𝑑𝑥 converges if both ∫−∞ 𝑓(𝑥)𝑑𝑥 and ∫𝑑 𝑓(𝑥)𝑑𝑥 converge, for some
∞ 𝑑 ∞
number d. In that case, ∫−∞ 𝑓(𝑥)𝑑𝑥 = ∫−∞ 𝑓(𝑥)𝑑𝑥 + ∫𝑑 𝑓(𝑥)𝑑𝑥 .
∞ 1
Example: Determine whether ∫−∞ 2 𝑑𝑥 converges.
𝑥 +1
Solution: The integrand is continuous on (−∞, ∞).To shows that the integral converges, we will
0 1 ∞ 1
show that∫−∞ 𝑥 2 +1 𝑑𝑥 and ∫0 𝑥 2 +1 𝑑𝑥 both converge.

72
CHAPTER FOUR

APPLICATION OF INTEGRATION

SURFACE AREA

A higher-dimensional version of the length of a curve is the area of a surface. It has been
known since antiquity that the surface area S of a cube of side s is given by 𝑆 = 6𝑠 2 and
the surface area S of a cylinder of radius r and height h is given by 𝑆 = 2𝜋𝑟ℎ.

More generally, suppose 𝑓 is defined on [a, b], and 𝑓(𝑥) ≥ 0 for 𝑎 ≤ 𝑥 ≤ 𝑏. We will
derive a formula for the area S of the surface obtained by revolving the graph of 𝑓 about the
x -axis.

Definition:

1. Let 𝑓 be nonnegative and continuously differentiable on [a, b]. The area of the surface
obtained by revolving the graph of 𝑓 about the x axis is defined by;
𝒃
𝑺 = ∫ 𝟐𝝅𝒇(𝒙)√𝟏 + (𝒇′ (𝒙))𝟐 𝒅𝒙
𝒂

2. Let 𝑓 be nonnegative and continuously differentiable on [c, d]. The area of the surface
obtained by revolving the graph of 𝑓 about the y axis is defined by;
𝒅
𝑺 = ∫ 𝟐𝝅𝒇(𝒚)√𝟏 + (𝒇′ (𝒚))𝟐 𝒅𝒚
𝒄

Example

Let 𝑓(𝑥) = 𝑥 3 , 0 ≤ 𝑥 ≤ 1. Find the area S of the surface obtained by revolving the graph of
𝑓 about the x axis.

Solution

Since 𝑓 ′ (𝑥) = 3𝑥 2 it follows from above definition

that;
𝒃
𝑺 = ∫ 𝟐𝝅𝒇(𝒙)√𝟏 + (𝒇′ (𝒙))𝟐 𝒅𝒙
𝒂
𝟏
= ∫ 𝟐𝝅𝑥 3 √𝟏 + (3𝑥 2 )𝟐 𝒅𝒙
𝟎
𝟏
= 𝟐𝝅 ∫ 𝑥 3 √𝟏 + 𝟗𝒙𝟒 𝒅𝒙
𝟎

73
 To evaluate the integral; let 𝒖 = √𝟏 + 𝟗𝒙𝟒 and
𝒅𝒖 = 𝟑𝟔𝒙𝟑 𝒅𝒙 . Now if x = 0 then u = 1, and if x = 1
𝟐𝝅 𝟏𝟎
then u = 10. Therefore; 𝑆 = 𝟑𝟔 ∫𝟏 √𝒖 𝒅𝒖
𝝅 𝟐 𝟑⁄ 𝟏𝟎
= ( 𝒖 𝟐 ) |𝟏
𝟏𝟖 𝟑
𝝅 𝟑
= (𝟏𝟎 ⁄𝟐 − 𝟏)
𝟐𝟕

Example

Finding Area for Revolution about the y-axis if the line segment 𝑥 = 1 − 𝑦, 0 ≤ 𝑦 ≤ 1, is
revolved about the y-axis to generate the cone.

Solution:

From the equation of line segment we have

𝑓(𝑦) = 1 − 𝑦, 𝑐 = 0 and 𝑑 = 1.
Hence the required area of S is;
𝒅
𝑺 = ∫ 𝟐𝝅𝒇(𝒚)√𝟏 + (𝒇′ (𝒚))𝟐 𝒅𝒚
𝒄
𝟏
= ∫ 𝟐𝝅(𝟏 − 𝒚)√𝟏 + (−𝟏))𝟐 𝒅𝒚
𝟎
𝟏
= 2√2𝜋 ∫ (𝟏 − 𝒚)𝒅𝒚
𝟎
1
𝑦2
= 2√2𝜋 [𝑦 − ]
2 0
= 𝜋√2

Arc Length

Suppose that a curve (arc) C is defined by the equation 𝑦 = 𝑓(𝑥), where 𝑓 is continuous and
𝑎 ≤ 𝑥 ≤ 𝑏. We obtain a polygonal approximation to C by dividing the interval [𝑎, 𝑏] into 𝑛
subintervals with endpoints 𝑥0 , 𝑥1 , … … . , 𝑥𝑛 and equal width Δx. If 𝑦𝑖 = 𝑓(𝑥𝑖 ) , then the point
𝑃𝑖 (𝑥𝑖 , 𝑦𝑖 ) lies on C and the polygon with vertices 𝑃0 , 𝑃1 , … … . , 𝑃𝑛 illustrated in the following
figure, is an approximation to C. The length L of C is approximately the length of this polygon
and the approximation gets better as we let n increase.

74
 Arc Length
If 𝑓 ′ is continuous on [𝑎, 𝑏] ,
then the length of the curve
𝑦 = 𝑓(𝑥), 𝑎 ≤ 𝑥 ≤ 𝑏, is
𝑏

𝐿 = ∫ √1 + (𝑓 ′ (𝑥))2 𝑑𝑥
𝑎

Example
1
Let 𝑓(𝑥) = ln 𝑥 − 8 𝑥 2 for 1 ≤ 𝑥 ≤ 2. Find the length L of the arc.

Solution

Since 𝒇′ (𝒙) = 𝟏⁄𝒙 − 𝒙⁄𝟒, it follows

from arc length formula that,
𝟐
𝟐
𝑳 = ∫ √𝟏 + (𝟏⁄𝒙 − 𝒙⁄𝟒) 𝒅𝒙
𝟏
𝟐
𝟏 𝟏 𝒙𝟐
= ∫√ 𝟐 + + 𝒅𝒙
𝒙 𝟐 𝟏𝟔
𝟏
𝟐
𝟐
= ∫ √(𝟏⁄𝒙 + 𝒙⁄𝟒) 𝒅𝒙
𝟏
𝟐

= ∫ 𝟏⁄𝒙 + 𝒙⁄𝟒 𝒅𝒙
𝟏
𝟐
𝒙𝟐
= [𝐥𝐧 𝒙 + ]
𝟖 𝟏
𝟑
= 𝐥𝐧 𝟐 +
𝟖
Remark:

If 𝑓 ′ is continuous on [𝑐, 𝑑] , then the length of the curve 𝑥 = 𝑓(𝑦), 𝑐 ≤ 𝑦 ≤ 𝑑, is

𝑏

𝐿 = ∫ √1 + (𝑓 ′ (𝑦))2 𝑑𝑦
𝑎

Example
Find the length of the arc of the parabola from 𝑥 = 𝑦 2 from (0,0) to (1,1).
75
Solution
Since 𝒇(𝒚) = 𝒚𝟐 , we have 𝒇′ (𝒚) = 𝟐𝒚 and the above remark gives;
𝟏
𝑳 = ∫𝟎 √1 + (𝑓 ′ (𝑦))2 𝑑𝑦
𝟏
= ∫𝟎 √1 + (2𝑦)2 𝑑𝑦
𝟏

= ∫ √1 + 4𝑦 2 𝑑𝑦
𝟎
𝟏 𝟏
We make trigonometric substitution 𝒚 = 𝟐 𝐭𝐚𝐧 𝜽, 𝒅𝒚 = 𝟐 𝒔𝒆𝒄𝟐 𝜽 𝒅𝜽 and
√1 + 4𝑦 2 = 𝑠𝑒𝑐𝜃
When 𝒚 = 𝟎, then 𝐭𝐚𝐧 𝜽 = 𝟎 ⇒ 𝜽 = 𝟎, 𝒚 = 𝟏,then 𝐭𝐚𝐧 𝜽 = 𝟐 ⇒ 𝜽 = 𝒕𝒂𝒏−𝟏 𝟐.
𝒕𝒂𝒏−𝟏 𝟐
Thus, 𝑳 = ∫𝟎 𝒔𝒆𝒄𝟑 𝜽𝒅𝜽
𝟏 −𝟏 𝟐
= [𝒔𝒆𝒄𝜽 𝒕𝒂𝒏𝜽 + 𝒍𝒏|𝒔𝒆𝒄𝜽 + 𝒕𝒂𝒏𝜽|]𝒕𝒂𝒏
𝟎
𝟒
√𝟓 𝒍𝒏(√𝟓 + 𝟐)
= +
𝟐 𝟒

Formula for a volume of revolution:

𝑏
𝑉 = ∫𝑎 𝜋[𝑓(𝑥)]2 𝑑𝑥
It applies only when the axis of rotation is the x – axis.
Example: Find the volume of the solid obtained by rotating about the 𝑥-axis the region under the
curve 𝑦 = √𝑥 from 0 to 1.
Solution: The region is shown in following Figure.

Taking 𝑎 = 0, 𝑏 = 1, and 𝑓(𝑥) = √𝑥 in Formula, we find that the volume of the solid is
1 1 𝑥2 1 𝜋
𝑉 = ∫0 𝜋(√𝑥)2 𝑑𝑥 = 𝜋 ∫0 𝑥𝑑𝑥 = 𝜋 ]
2 0
= 2
Definition
𝑑
𝑉 = ∫𝑐 𝜋[𝑔(𝑦)]2 𝑑𝑦
The above formula applies only when the axis of rotation is the y – axis.
Example: Find the volume of the solid obtained by rotating the region bounded by 𝑦 = 𝑥 3 , 𝑦 =
8, and 𝑥 = 0 around the y-axis.
Solution: Since the axis of rotation is the 𝑦-axis, we write the curve 𝑦 = 𝑥 3 in the form 𝑥 = 3√𝑦.
The region and the solid with a typical disk are sketched in following Figure.

76
Using Formula with 𝑐 = 0, 𝑑 = 8, 𝑎𝑛𝑑 𝑔(𝑦) = 3√𝑦, we have
8 8
𝑉 = ∫0 𝜋( 3√𝑦)2 𝑑𝑦 = 𝜋 ∫0 𝑦 2/3 𝑑𝑦
5
3 96𝜋
= 𝜋 [5 𝑦 3 ] 80 = 5
Volumes by Cylindrical Shells
𝑏
The volume of the solid is 𝑉 = ∫𝑎 2𝜋𝑥𝑓(𝑥)𝑑𝑥 where 0 ≤ 𝑎 < 𝑏

Example: Find the volume of the solid obtained by rotating about the 𝑦-axis the region bounded
by
𝑦 = 𝑥(𝑥 − 1)2 and 𝑦 = 0.
Solution: The region and solid are sketched in following figure.

By using formula, the volume of rotation as

1
𝑉 = ∫0 2𝜋𝑥[𝑥(𝑥 − 1)2 ]𝑑𝑥
1
= 2𝜋 ∫0 (𝑥 4 − 2𝑥 3 + 𝑥 2 )𝑑𝑥
𝑥5 𝑥4 𝑥3 𝜋
= 2𝜋 [ 5 − 2 4 + 3 ] 10 = 15
Example: Find the volume of the solid obtained by rotating about the 𝑦-axis the region between
𝑦 = 𝑥 and 𝑦 = 𝑥 2 .
Solution: We use Formula to find the volume obtained by rotating about the 𝑦-axis the region
under 𝑦 = 𝑥 from 0 to 1 and then we subtract the corresponding volume for the region under 𝑦 =
𝑥 2.

77
Thus
1 1
𝑉 = ∫0 2𝜋𝑥 ∙ 𝑥𝑑𝑥 − ∫0 2𝜋𝑥 ∙ 𝑥 2 𝑑𝑥
1 𝑥3 𝑥4 1 𝜋
= 2𝜋 ∫0 (𝑥 2 − 𝑥 3 )𝑑𝑥 = 2𝜋 [ 3 − ]
4 0
= 6

78
CHAPTER FIVE

DIFFERENTIAL EQUATION

PART ONE: ORDINARY DIFFERENTIAL EQUATIONS

“When mathematical modelling is used to describe physical, biological or chemical phenomena,one of the
most common results of the modelling process is a system of ordinary or partial differential equations.
Finding and interpreting the solutions of these differential equations is therefore a central part of applied
mathematics, and a thorough understanding of differential equations is essential for any applied
mathematician”
1.1 Basic concepts and ideas

• Equation: Equations describe the relations between the dependent and independent variables.
An equal sign "=" is required in every equation.
• Differential Equation: Equations that involve dependent variables and their derivatives with
respect to the independent variables are called differential equations.
• A Differential Equation is a Mathematical statement that contains one or more derivatives. It
states a relationship involving the rates of change of continuously changing quantities modeled
by functions. Differential equations are very common in physics, engineering, and all fields
involving quantitative study of change. They are used whenever a rate of change is known but
the process giving rise to it is not. The solution of a differential equation is generally a function
whose derivatives satisfy the equation.
• A differential equation is a mathematical equation for an unknown function in one or several
variables that relates the values of the function itself and its derivatives of various order.
Example:- 𝑦 ′ + 𝑒 𝑦 𝑐𝑜𝑠𝑥 = 0 𝑤ℎ𝑒𝑟𝑒 𝑦(𝑥)
𝜕𝑦 𝜕𝑦 𝜕3 𝑦
𝜕𝑥
= 6𝑦 𝜕𝑡 − 𝜕𝑡 3 𝑤ℎ𝑒𝑟𝑒 𝑦(𝑥, 𝑡)
• Differential equations are classified into several broad categories. The most important ones are
ordinary differential equations (ODEs), in which change depends on a single variable, and
partial differential equations (PDEs), in which change depends on several variables
• An ordinary differential equation (ODE) is a differential equation in which the unknown function
(dependent variable) is a function of a single independent variable.
Example:
𝑦 ′ = 𝑐𝑜𝑠𝑥
𝑦 ′′ + 4𝑦 = 0
𝑥 2 𝑦 ′′′ 𝑦 ′ + 2𝑒 𝑥 𝑦 ′′ = (𝑥 2 + 2)𝑦 2
𝑑2 𝑧 𝑑𝑧
𝑎 𝑑𝑡 2 + 𝑏 𝑑𝑡 + 𝑐𝑧 = cos 𝜔𝑡

In these examples the dependent variables are y and z, the independent variables are x and t
i.e. y is a function of x and z is a function of t

• Partial differential equation (PDE) is a differential equation in which the unknown function is a
function of multiple independent variables

79
 Example:
𝜕2 𝑢 𝜕𝑢
+𝑡 =0
𝜕𝑡 2 𝜕𝑥
𝜕𝑢 𝜕𝑢 𝜕3 𝑢
𝜕𝑥
= 6𝑢 𝜕𝑡
− 𝜕𝑡 3
𝜕𝑃 𝜕2 𝑃 𝜕2 𝑃
= 𝑘 ( 2 + 2)
𝜕𝑡 𝜕𝑥 𝜕𝑦

In these examples the dependent variables are u and P and the independent variables are x, y and t
i.e. u is the function of x and t. P is a function of x ,y and t

• Ordinary differential functions are further classified according to the order of the highest derivative of
the dependant variable with respect to the independent variable appearing in the equation.

Order: The order of a differential equation is the highest derivative that appears in the differential
equation.

Example

𝑌 ′ = sin 𝑥 is a first order ordinary differential equation

𝑦 ′′ + 2𝑦 = 0 is a second order ordinary differential equation

𝑥 2 𝑦 ′′′ + 2𝑒 𝑥 𝑦 ′ = (𝑥 2 + 2)𝑦 2 is a third order ordinary differental equation

Degree: The degree of a differential equation is the power of the highest derivative term.

Differential equation is linear if the unknown function and its derivatives appear to the power one
and products are not allowed and non linear otherwise.

Linear: A differential equation is called linear if there are no multiplications among dependent
variables and their derivatives. In other words, all coefficients are functions of independent
variables.

Linear ordinary differential equations can be written in the form

𝑎0 (𝑥)𝑦 (𝑛) + 𝑎1 (𝑥)𝑦 (𝑛−1) + ⋯ + 𝑎𝑛−1 (𝑥)𝑦 ′ + 𝑎𝑛 (𝑥)𝑦 = 𝑔(𝑥)

Non-linear: Differential equations that do not satisfy the definition of linear are non-linear.

Quasi-linear: For a non-linear differential equation, if there are no multiplications among all
dependent variables and their derivatives in the highest derivative term, the differential equation
is considered to be quasi-linear.

Example

1. 3𝑥 2 𝑦 ′′ + 2 ln 𝑥 𝑦 ′ + 𝑒 𝑥 𝑦 = 3𝑥 cos 𝑥 is a linear second order differential equation

2. 4𝑦𝑦 ′′′ − 𝑥 3 𝑦 ′ = 𝑒 2𝑥 is not linear because of 𝑦 𝑦′′′ product

80
 3. 𝑦 ′ + cos 𝑦 = 𝑥 is not linear because of cos 𝑦 it can be said quasi linear

Non linear differential equations are often very difficult or impossible to solve

Homogeneous: A differential equation is homogeneous if every single term contains the

dependent variables or their derivatives.

Non-homogeneous: Differential equations which do not satisfy the definition of homogeneous

are considered to be non-homogeneous.

Examples

In the first group of examples, let u be an unknown function of x, and c and ω are known
constants.

• Inhomogeneous (Non-homogeneous)first order linear constant coefficient ordinary

differential equation:

because of 𝑥 2

• Homogeneous second order linear ordinary differential equation:

every term contain either u or derivative of u

• Homogeneous second order constant coefficient linear ordinary differential equation

describing the harmonic oscillator:

• First order nonlinear ordinary differential equation:

non linear because of 𝑢2 and inhomogeneous(Non-homogeneous) because of 1

• Second order nonlinear ordinary differential equation describing the motion of a

pendulum of length L:

Solutions Differential Equations

81
General Solution: Solutions obtained from integrating the differential equations are called
general solutions. The general solution of a 𝑛𝑡ℎ order ordinary differential equation contains 𝑛
arbitrary constants resulting from integrating 𝑛 times.

Particular Solution: Particular solutions are the solutions obtained by assigning specific values
to the arbitrary constants in the general solutions.

Singular Solutions: Solutions that can not be expressed by the general solutions are called
singular solutions.

A solution of a differential equation which is of the form 𝑦 = ℎ(𝑥) is called explicit solution forms. If the
solution y is given by implicit relation 𝐻(𝑥, 𝑦) = 0 and satisfies the differential equation it is said to be
implicit solution forms.

Example
1
The solutions 𝑦(𝑡) = log(𝑒 −1 +sin 𝑡) and 𝑦 = − cos 𝑥 + 𝑐 are given in explicit forms and the solutions
2 3 4
𝑦 − ln 𝑦 = 𝑥 + 𝐶 and 4𝑦 − 3𝑥 + 𝐶 = 0 are given in implicit forms.

Homogenous linear differential equations are a further subclass for which the space of solutions is a linear
subspace .i.e. the sum of any solutions or multiple of solution is also a solution.

If the function 𝑓(𝑥 , 𝑦)is homogeneous 𝑓(𝑡𝑥 , 𝑡𝑦) = 𝑓(𝑥 , 𝑦) for any number t other wise non
homogenous.
𝑥𝑦
Example 𝑓(𝑥 , 𝑦) =
𝑥 2 +𝑦 2

(𝑡𝑥)(𝑡𝑦) 𝑡 2 𝑥𝑦 𝑥𝑦
𝑓(𝑡𝑥 , 𝑡𝑦) = 2 2
= = 2 = 𝑓(𝑥 , 𝑦)
(𝑡𝑥) + (𝑡𝑦) 𝑡 (𝑥 + 𝑦 ) 𝑥 + 𝑦 2
2 2 2

Hence homogeneous

−3𝑥 2 𝑦
Example 𝑓(𝑥 , 𝑦) = ln(𝑥 3 +4𝑥𝑦2 )

−3(𝑡𝑥)2 (𝑡𝑦) 𝑡 3 (−3𝑥 2 𝑦) −3𝑥 2 𝑦

𝑓(𝑡𝑥 , 𝑡𝑦) = ln ( ) = ln( 3 3 ) = ln( 3 ) = 𝑓(𝑥 , 𝑦)
(𝑡𝑥)3 + 4(𝑡𝑥)(𝑡𝑦)2 𝑡 (𝑥 + 4𝑥𝑦 2 ) 𝑥 + 4𝑥𝑦 2

Hence homogeneous

Example 𝑓(𝑢 , 𝑥) = 𝑐𝑢 + 𝑥 2

𝑓(𝑡𝑢 , 𝑡𝑥) = 𝑐(𝑡𝑢) + (𝑡𝑥)2 = 𝑡(𝑐𝑢 + 𝑡𝑥 2 ) ≠ 𝑓(𝑢, 𝑥)

Hence non-homogeneo us

82
Conditions

Initial Condition: Constrains that are specified at the initial point, generally time point, are
called initial conditions. Problems with specified initial conditions are called initial value
problems.

Boundary Condition: Constrains that are specified at the boundary points, generally space
points, are called boundary conditions. Problems with specified boundary conditions are called
boundary value problems.

Exactness

• In order to classify ordinary differential equations as exact or non exact first we rewrite the equation
into 𝑀(𝑥 , 𝑦)𝑑𝑥 + 𝑁(𝑥 , 𝑦)𝑑𝑦 = 0
𝜕𝑀 𝜕𝑁
This equation will be exact if = and non- exact otherwise.
𝜕𝑦 𝜕𝑥

Example:-

−(1+2𝑥𝑦 3 )
1. 𝑦 ′ = 3𝑥 2 𝑦 2
𝑑𝑦 −(1 + 2𝑥𝑦 3 )
=
𝑑𝑥 3𝑥 2 𝑦 2
3𝑥 𝑦 𝑑𝑦 + (1 + 2𝑥𝑦 3 )𝑑𝑥 = 0
2 2

𝑀(𝑥 , 𝑦) = 1 + 2𝑥𝑦 3 𝑁(𝑥 , 𝑦) = 3𝑥 2 𝑦 2

𝑑𝑀 𝑑𝑁
= 6𝑥𝑦 2 = 6𝑥𝑦 2
𝑑𝑦 𝑑𝑥
𝑑𝑀 𝑑𝑁
= hence Exact
𝑑𝑦 𝑑𝑥
2. (𝑥 3 + 3𝑥𝑦 2 )𝑑𝑥 + (3𝑥 2 𝑦 + 𝑦 3 )𝑑𝑦 = 0
𝑀(𝑥 , 𝑦) = 𝑥 3 + 3𝑥𝑦 2 𝑁(𝑥 , 𝑦) = 3𝑥 2 𝑦 + 𝑦 3
𝑑𝑀 𝑑𝑁
= 6𝑥𝑦 = 6𝑥𝑦
𝑑𝑦 𝑑𝑥
hence exact
3. 𝑦𝑑𝑥 − 𝑥𝑑𝑦 = 0
𝑀(𝑥 , 𝑦) = 𝑦 𝑁(𝑥 , 𝑦) = −𝑥

𝑑𝑀 𝑑𝑁
=1 = −1
𝑑𝑦 𝑑𝑥
𝑑𝑀 𝑑𝑁
≠
𝑑𝑦 𝑑𝑥
hence this is non exact
𝑑𝑦
• A differential equation of the form = 𝑓(𝑥 , 𝑦) is called separable if 𝑓(𝑥 , 𝑦) = ℎ(𝑥)𝑔(𝑦)
𝑑𝑥

83
 𝑑𝑦
i.e. 𝑑𝑥
= ℎ(𝑥)𝑔(𝑦) and non-separable otherwise
Example 1 9𝑦𝑦 ′ + 4𝑥 = 0
−4𝑥
𝑦′ = 9𝑦
𝑑𝑦 1 1
= −4𝑥 where ℎ(𝑥) = −4𝑥 and 𝑔(𝑦) =
𝑑𝑥 9𝑦 9𝑦

Hence the equation is separable

−2𝑥+5𝑦
Example 2 𝑦 ′ = 2𝑥+𝑦
𝑑𝑦 −2𝑥 + 5𝑦
=
𝑑𝑥 2𝑥 + 𝑦

Here we cannot separate the variables as a function of x and y

Hence the equation is non separable.

1.2. Separable Equations

𝑑𝑦
We will start by looking at a class of equations that can be rewritten in the form 𝑔(𝑦) 𝑑𝑡 = 𝑓(𝑡)

𝑑𝑦
These equations can be integrated with respect to t to give ∫ 𝑔(𝑦) 𝑑𝑡 = ∫ 𝑓(𝑡) 𝑑𝑡 and using the chain
𝑑𝑡
rule, ∫ 𝑔(𝑦)𝑑𝑦 = ∫ 𝑓(𝑡) 𝑑𝑡
𝑑𝑦
Example 1.Determine the solution of the differential equation + 𝑒 𝑦 cos 𝑡 = 0 subject to the condition
𝑑𝑡
y (0) =1.
𝑑𝑦 𝑑𝑦
Solution: - Firstly, we convert the equation to the form𝑔(𝑦) 𝑑𝑡 = 𝑓(𝑡), so that 𝑒 −𝑦 𝑑𝑡 = − cos 𝑡
integrating with respect to t we have −𝑒 −𝑦 = − sin 𝑡 + 𝐶

Using the boundary condition we find that𝐶 = −𝑒 −1 , and rearranging gives the solution

1
𝑦(𝑡) = log( −1 )
𝑒 + sin 𝑡

Example 2. Solve the differential equation 𝑦 ′ = sin 𝑥

𝑑𝑦
Solution:- 𝑑𝑥
= sin 𝑥 =>𝑑𝑦 = sin 𝑥 𝑑𝑥

∫ 𝑑𝑦 = ∫ sin 𝑥 𝑑𝑥

84
 𝑦 = − cos 𝑥 + 𝑐
2𝑥𝑦
Example 3. solve the differential equation 𝑦 ′ = 𝑦−1

𝑑𝑦 2𝑥𝑦
Solution 𝑑𝑥 = 𝑦−1
𝑦−1
𝑑𝑦 = 2𝑥𝑑𝑥
𝑦
𝑦 − ln 𝑦 = 𝑥 2 + 𝐶

𝑥3
Example 4 solve the differential equation 𝑦 ′ = 𝑦2

𝑑𝑦 𝑥3
=
𝑑𝑥 𝑦2
𝑦 2 𝑑𝑦 = 𝑥 3 𝑑𝑥
𝑦3 𝑥4
3
= 4
+𝐶
4𝑦 − 3𝑥 4 + 𝐶 = 0
3

Exercises (for discussion)

Find the general solutions

a. 𝑦 ′ sin 2𝑥 = 𝑦 cos 2𝑥
b. 𝑦 ′ = 𝑒 2𝑥 𝑐𝑜𝑠 2 𝑦
𝑦
c. 𝑦 ′ = 𝑥 𝑙𝑛 𝑥
d. 𝑦 ′ = 2𝑥(1 + 𝑦)
𝑦 2 −1
e. 𝑦 ′ = 𝑥

1.2. Equations Reducible To Separable

There are some non separable first order ordinary differential equations which can be made separable by a
𝑦 𝑦 𝑦
simple change of variables like the equations 𝑦 ′ = 𝑔( 𝑥 ) . Here g is a function of 𝑥 so if we let 𝑥 = 𝑢 we
get 𝑦 = 𝑥𝑢 (product rule of differentiation )
𝑑𝑢 𝑑𝑢
𝑛𝑜𝑤 𝑦 ′ = 𝑢 + 𝑥 𝑑𝑥
So we have 𝑢 + 𝑥 𝑑𝑥
= 𝑔(𝑢) now we can easily separate as a function of x and as
a function of u.
𝑑𝑢
i.e. 𝑢+𝑥 = 𝑔(𝑢)
𝑑𝑥

𝑑𝑢
𝑥 = 𝑔(𝑢) − 𝑢
𝑑𝑥
𝑑𝑥 𝑑𝑢
=
𝑥 𝑔(𝑢) − 𝑢

85
Hence evaluating this integral we obtain a general solution

Example: - Solve the differential equation 2𝑥𝑦𝑦 ′ − 𝑦 2 + 𝑥 2 = 0

Solution : Divide all sides by 𝑥 2

𝑦 𝑦 2
2 ( ) 𝑦′ − ( ) + 1 =
𝑥 𝑥
𝑦 𝑑𝑢
Let 𝑢 = 𝑥 => 𝑦 = 𝑢𝑥 𝑦 ′ = 𝑢 + 𝑥 𝑑𝑥

2𝑢(𝑢 + 𝑥𝑢′ ) − 𝑢2 + 1 = 0

2𝑢2 + 2𝑥𝑢𝑢′ − 𝑢2 + 1 = 0

𝑢2 + 1
= −2𝑥𝑢′
𝑢
𝑑𝑥 𝑑𝑢
− =
2𝑥 (𝑢 + 1)
𝑢
1 𝑢𝑑𝑢
− ln 𝑥 = ∫ 2 +𝐶
2 𝑢 +1
1
ln = ln|𝑢2 + 1| + 𝐶
𝑥
𝐾 𝑦2
= +1
𝑥 𝑥2
𝑦 2 + 𝑥 2 − 𝐾𝑥 = 0

Exercises(for discussion) solve the following equations

a. (2𝑥 − 4𝑦 + 5)𝑦 ′ + 𝑥 − 2𝑦 + 3 = 0 [ ℎ𝑖𝑛𝑡 𝑢𝑠𝑒 𝑢 = 𝑥 − 2𝑦]

′ 2
b. 𝑦 = (𝑦 + 𝑥) 𝑢𝑠𝑒 𝑢 = 𝑦 + 𝑥
𝑦−𝑥+1
c. 𝑦 ′ = 𝑢𝑠𝑒 𝑢 = 𝑦 − 𝑥
𝑦−𝑥+5
1.3.Exact Differential Equations

All the techniques we have reviewed so far were not of a general nature since in each case the
equations themselves were of a special form. So, we may ask, what to do for the general equation
𝑑𝑦
= 𝑓(𝑥, 𝑦 )
𝑑𝑥
Let us first rewrite the equation into 𝑀(𝑥 , 𝑦)𝑑𝑥 + 𝑁(𝑥 , 𝑦)𝑑𝑦 = 0
𝜕𝑀 𝜕𝑁
This equation will be called exact if = and nonexact otherwise.
𝜕𝑦 𝜕𝑥

86
The condition of exactness insures the existence of a function F(x,y) such that
𝜕𝐹
= 𝑀(𝑥, 𝑦)
𝜕𝑥
𝜕𝐹
= 𝑁(𝑥, 𝑦)
{ 𝜕𝑦
When the equation is exact, we solve it using the following steps:
1. Check that the equation is indeed exact;
2.Write down the system
𝜕𝐹
= 𝑀(𝑥, 𝑦)
𝜕𝑥
𝜕𝐹
= 𝑁(𝑥, 𝑦)
{ 𝜕𝑦
3.Integrate either the first equation with respect of the variable x or the second with respect of the
variable y. The choice of the equation to be integrated will depend on how easy the calculations
are. Let us assume that the first equation was chosen, then we get
𝑓(𝑥, 𝑦) = ∫ 𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑘(𝑦)
The function 𝑘(𝑦) should be there, since in our integration, we assumed that the variable y is
constant.
4.Use the second equation of the system to find the derivative of 𝑘(𝑦). Indeed, we have
𝜕𝐹 𝜕
= 𝜕𝑦 (∫ 𝑀(𝑥, 𝑦)𝑑𝑥) + 𝑘 ′ (𝑦) = 𝑁(𝑥, 𝑦)
𝜕𝑦
which implies
𝜕
𝑘 ′ (𝑦) = 𝑁(𝑥, 𝑦) − (∫ 𝑀(𝑥, 𝑦)𝑑𝑥)
𝜕𝑦

Note that k is a function of y only. Therefore, in the expression giving 𝑘(𝑦)the variable, x, should
disappear. Otherwise something went wrong!

5. Integrate to find 𝑘(𝑦) ;

6. Write down the function F(x,y);
7. All the solutions are given by the implicit equation F(x,y)=C
8. If you are given an IVP, plug in the initial condition to find the constant C.

Solving Exact First Order Ordinary Differential Equations

For exact first order ordinary differential equations the solution𝑢(𝑥, 𝑦) can be found by guessing
or in the following systematic way. I.e. from the definition of exact we have

𝑀(𝑥 , 𝑦)𝑑𝑥 + 𝑁(𝑥 , 𝑦)𝑑𝑦 = 0 ................................ .(1)

𝜕𝑢 𝜕𝑢
𝑑𝑢 = 𝜕𝑥 𝑑𝑥 + 𝜕𝑦 𝑑𝑦 of some function 𝑢(𝑥, 𝑦)

If we put 𝑑𝑢 = 0we have 𝑢(𝑥, 𝑦) = 𝐶 by mean value theorem corollary

𝜕𝑢 𝜕𝑢
𝑑𝑥 + 𝜕𝑦 𝑑𝑦 = 0 .......................................... (2)
𝜕𝑥

87
Comparing (1) and (2) we have

𝜕𝑢 𝜕𝑢
𝑀(𝑥 , 𝑦) = 𝑁(𝑥 , 𝑦) =
𝜕𝑥 𝜕𝑦

Then 𝜕𝑢 = 𝑀(𝑥, 𝑦)𝜕𝑥 𝜕𝑢 = 𝑁(𝑥, 𝑦)𝜕𝑦

𝑢 = ∫ 𝑀(𝑥, 𝑦)𝑑𝑥 + 𝐾(𝑦) 𝑢 = ∫ 𝑁(𝑥, 𝑦)𝑑𝑦 + 𝑙(𝑥)

Example:-Solve (𝑥 3 + 3𝑥𝑦 2 )𝑑𝑥 + (3𝑦𝑥 2 + 𝑦 3 )𝑑𝑦 = 0

𝑀(𝑥 , 𝑦) = 𝑥3 + 3𝑥𝑦2 𝑁(𝑥 , 𝑦) = 3𝑦𝑥2 + 𝑦3

𝑢 = ∫(𝑥3 + 3𝑥𝑦2 )𝑑𝑥 + 𝐾(𝑦) 𝑢 = ∫(3𝑦𝑥2 + 𝑦3 )𝑑𝑦 + 𝑙(𝑥)
𝑥 4 3𝑦 2 𝑥 2 3𝑦 2 𝑥 2 𝑦 4
𝑢= + + 𝐾(𝑦) 𝑢= + + 𝑙(𝑥)
4 2 2 4
𝑥 4 3𝑦 2 𝑥 2 3𝑦 2 𝑥 2 𝑦 4
+ + 𝐾(𝑦) = + + 𝑙(𝑥)
4 2 2 4
𝑦4
𝐾(𝑦) =
4
𝑥4
𝑙(𝑥) =
4
𝑥 4 3𝑦 2 𝑥 2 𝑦 4
𝑡ℎ𝑒𝑟𝑒 𝑓𝑜𝑟𝑒 𝑢 = + + =𝐶
4 2 4
Exercise (for discussion )
a. solve (sin 𝑥 cosh 𝑦)𝑑𝑥 − (cos 𝑥 sinh 𝑦)𝑑𝑦 = 0 with initial condition 𝑦(0) = 0
−(1+2𝑥𝑦 3 )
b. Solve 𝑦 ′ = 3𝑥 2 𝑦 2
c. Solve (𝑥 3 + 3𝑥𝑦 2 )𝑑𝑥
+ (3𝑥 2 𝑦 + 𝑦 3 )𝑑𝑦 = 0

1.4.Integrating Factors

You may ask, what do we do if the equation is not exact? In this case, one can try to find an
integrating factor which makes the given differential equation exact.

• Even though an equation 𝑃(𝑥 , 𝑦)𝑑𝑥 + 𝑄(𝑥 , 𝑦)𝑑𝑦 = 0 may not be exact, sometimes it is not too
difficult to find a function 𝑢, no where zero such that
𝑢(𝑥, 𝑦)𝑃(𝑥 , 𝑦)𝑑𝑥 + 𝑢(𝑥, 𝑦)𝑄(𝑥 , 𝑦)𝑑𝑦 = 0

is exact . Such a function is called integrating factor.

𝜕𝑀 𝜕𝑁
By definition of exact we have =
𝜕𝑦 𝜕𝑥
𝜕(𝑢𝑃) 𝜕(𝑢𝑄)
=
𝜕𝑦 𝜕𝑥

88
 𝜕𝑢 𝜕𝑃 𝜕𝑢 𝜕𝑄
𝑃+𝑢 = 𝑄+𝑢
𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑥
𝜕𝑢
Let the integrating factor u be u(x) i.e. 𝜕𝑦 = 0
Then
𝜕𝑃 𝜕𝑢 𝜕𝑄
𝑢 = 𝑄+𝑢
𝜕𝑦 𝜕𝑥 𝜕𝑥
1 𝜕𝑢 1 𝜕𝑃 𝜕𝑄
= ( − )
𝑢 𝜕𝑥 𝑄 𝜕𝑦 𝜕𝑥
Example Solve the differential equation 𝑦𝑑𝑥 − 𝑥𝑑𝑦 = 0 which is not exact
Hence we can find the integrating factor by letting P=y and Q= -x
𝜕(𝑢𝑃) 𝜕(𝑢𝑄)
=
𝜕𝑦 𝜕𝑥
𝜕𝑢 𝜕𝑃 𝜕𝑢 𝜕𝑄
𝑃+𝑢 = 𝑄+𝑢
𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑥
𝜕𝑢
Let the integrating factor u be u(y) i.e. 𝜕𝑥 = 0
𝜕𝑢 𝜕𝑃 𝜕𝑄
Then 𝜕𝑦 𝑃 + 𝑢 𝜕𝑦 = 𝑢 𝜕𝑥
1 𝜕𝑢 1 𝜕𝑄 𝜕𝑃
= 𝑃 ( 𝜕𝑥 − 𝜕𝑦 ),
𝑢 𝜕𝑦
1 𝜕𝑢 1 𝜕(−𝑥) 𝜕𝑦
= 𝑦( − 𝜕𝑦),
𝑢 𝜕𝑦 𝜕𝑥
𝑑𝑢 1
= 𝑦 (−1 − 1)𝑑𝑦,
𝑢
ln 𝑢 = −2 ln 𝑦 + 𝐶,
𝑘
𝑢 = 𝐾𝑦 −2 = 𝑦 2.

1 1 𝑥
but multiplying the equation by 𝑢(𝑥, 𝑦) = 𝑦 ≠ 0 makes it exact 𝑑𝑥 − 𝑑𝑦 =0
𝑦2 𝑦 𝑦2

1 −𝑥
𝑀(𝑥, 𝑦) = 𝑁(𝑥, 𝑦) =
𝑦 𝑦2
𝜕𝑀 −1 𝜕𝑁 −1
= 2 = 2
𝜕𝑦 𝑦 𝜕𝑥 𝑦
1 −𝑥
𝑢 = ∫ 𝑑𝑥 + 𝑘(𝑦) 𝑢 = ∫ 2 𝑑𝑦 + 𝑙(𝑥)
𝑦 𝑦
𝑥 𝑥
𝑢 = + 𝑘(𝑦) 𝑢 = + 𝑙(𝑥)
𝑦 𝑦
𝑘(𝑦) = 0
𝑙(𝑥) = 0
𝑥
There fore 𝑢 = =𝐶 𝑦 = 𝑘𝑥 𝑜𝑟 𝑥 = 𝑘𝑦
𝑦

Exercises (for discussion)

Find the integrating factors and solve

89
 a. (2𝑦 3 + 2)𝑑𝑥 + 3𝑥𝑦 2 𝑑𝑦 = 0
b. cos 𝑥 cos 𝑦 𝑑𝑥 − 2 sin 𝑥 sin 𝑦 𝑑𝑦 = 0
c. 2𝑥𝑦𝑑𝑥 + (4𝑦 + 3𝑥 2 )𝑑𝑦 = 0 with initial condition 𝑦(0.2) = −1.5

1.5. Linear First Order Differential Equations

Solving first order ordinary differential equations with constant coefficients

We will consider the simplest case first, when both p and q are constants

𝑑𝑦
+ 𝑝𝑦 = 𝑞
𝑑𝑥
There are a number of ways to solve this equation

Multiply both sides by 𝑒 𝑝𝑥 i.e.

𝑑𝑦 𝑝𝑥
𝑒 + 𝑝𝑦𝑒 𝑝𝑥 = 𝑞𝑒 𝑝𝑥
𝑑𝑥
𝑑 𝑑𝑦
but 𝑑𝑥 (𝑦𝑒 𝑝𝑥 ) = 𝑑𝑥 𝑒 𝑝𝑥 + 𝑝𝑦𝑒 𝑝𝑥 ( product rule of differentiation)

𝑑
(𝑦𝑒 𝑝𝑥 ) = 𝑞𝑒 𝑝𝑥
𝑑𝑥
Integrating both sides with respect to x we have
𝑞 𝑝𝑥
𝑦(𝑥)𝑒 𝑝𝑥 = 𝑒 +𝐶
𝑝
𝑞
𝑦(𝑥) = + 𝐶𝑒 −𝑝𝑥
𝑝

If the initial condition 𝑦(𝑎) = 𝑦𝑎 is given

𝑞
𝑦(𝑎) = 𝑦𝑎 = + 𝐶𝑒 −𝑝𝑎
𝑝
𝑞
𝐶 = (𝑦𝑎 − )𝑒 𝑝𝑎
𝑝
𝑞 𝑞
𝑦(𝑥) = + (𝑦𝑎 − )𝑒 𝑝𝑎 𝑒 −𝑝𝑥
𝑝 𝑝
𝑞 𝑞
𝑦(𝑥) = + (𝑦𝑎 − )𝑒 𝑝(𝑎−𝑥)
𝑝 𝑝
𝑑𝑦
Example 1) solve the equation 𝑑𝑥 + 3𝑦 = 6

90
here 𝑝 = 3 𝑎𝑛𝑑 𝑞 = 6

Multiply both sides by 𝑒 3𝑥

𝑑𝑦 3𝑥
𝑒 + 3𝑦𝑒 3𝑥 = 6𝑒 3𝑥
𝑑𝑥
𝑑(𝑦(𝑥)𝑒 3𝑥 )
= 6𝑒 3𝑥
𝑑𝑥
6
Integrating both sides we have 𝑦(𝑥)𝑒 3𝑥 = 3 𝑒 3𝑥 + 𝐶

𝑦(𝑥) = 2 + 𝐶𝑒 −3𝑥

If the initial condition 𝑦(0) = 0

0 = 𝑦(0) = 2 + 𝐶𝑒 −3𝑥0

𝐶 = −2

𝑦(𝑥) = 2 − 2𝑒 −3𝑥

Exercise for discussion

𝑑𝑥
a. − 5𝑥 = 6
𝑑𝑡
𝑑𝑥
b. + 4𝑥 = 0
𝑑𝑡

We now consider more general linear equation

𝑑𝑥
+ 𝑃(𝑡)𝑥 = 𝑞(𝑡)
𝑑𝑡
Multiply all sides by integrating factor I (t)

𝑑𝑥
𝐼(𝑡) + 𝑃(𝑡)𝑥(𝑡)𝐼(𝑡) = 𝑞(𝑡)𝐼(𝑡)
𝑑𝑡
𝑑 𝑑𝑥
[𝐼(𝑡)𝑥(𝑡)] = 𝐼(𝑡) + 𝐼(𝑡)𝑃(𝑡)𝑥(𝑡)
𝑑𝑡 𝑑𝑡
𝑑𝐼 𝑑𝑥 𝑑𝑥
𝑥(𝑡) + 𝐼(𝑡) = 𝐼(𝑡) + 𝐼(𝑡)𝑃(𝑡)𝑥(𝑡)
𝑑𝑡 𝑑𝑡 𝑑𝑡
𝑑𝐼
= 𝐼(𝑡)𝑃(𝑡)
𝑑𝑡
Now it is separable and we can solve as

91
 𝑑𝐼
= 𝑃(𝑡)𝑑𝑡
𝐼(𝑡)

ln 𝐼(𝑡) = ∫ 𝑝(𝑡)𝑑𝑡

𝐼(𝑡) = 𝑒 ∫ 𝑝(𝑡)𝑑𝑡

There fore this is the integrating factor for the general linear equation

𝑑𝑥
+ 𝑃(𝑡)𝑥 = 𝑞(𝑡)
𝑑𝑡
So the general solution is

𝑑𝑥
𝑒 ∫ 𝑝(𝑡)𝑑𝑡 + 𝑃(𝑡)𝑥(𝑡)𝑒 ∫ 𝑝(𝑡)𝑑𝑡 = 𝑞(𝑡)𝑒 ∫ 𝑝(𝑡)𝑑𝑡
𝑑𝑡
𝑑
[𝑥(𝑡)𝑒 ∫ 𝑝(𝑡)𝑑𝑡 ] = 𝑞(𝑡)𝑒 ∫ 𝑝(𝑡)𝑑𝑡
𝑑𝑡

𝑥(𝑡)𝑒 ∫ 𝑝(𝑡)𝑑𝑡 = ∫ 𝑞(𝑡) 𝑒 ∫ 𝑝(𝑡)𝑑𝑡 𝑑𝑡

Example. Solve the equation

𝑑𝑥 8
+ 3𝑥 = 𝑡 𝑥(0) = 𝑃(𝑡) = 3 𝑞(𝑡) = 𝑡
𝑑𝑡 9

Solution 𝐼(𝑡) = 𝑒 ∫ 3𝑑𝑡 = 𝑒 3𝑡

Multiply all sides by 𝑒 3𝑡

𝑑𝑥
𝑒 3𝑡 + 3𝑥𝑒 3𝑡 = 𝑡𝑒 3𝑡
𝑑𝑡
𝑑
[𝑥(𝑡)𝑒 3𝑡 ] = 𝑡𝑒 3𝑡
𝑑𝑡

𝑥(𝑡)𝑒 3𝑡 = ∫ 𝑡 𝑒 3𝑡 𝑑𝑡 + 𝐶

𝑡
𝑥(𝑡)𝑒 3𝑡 = 𝑒 3𝑡 + 𝐶
3
𝑡
𝑥(𝑡) = + 𝐶𝑒 −3𝑡
3

92
 8
Since 𝑥(0) = 9,

8 0
= 𝑥(0) = + 𝐶𝑒 −3𝑥0
9 3
𝑡 8 −3𝑡
𝑥(𝑡) = + 𝑒
3 9
Example Find the general solution of

𝑑𝑦
(𝑥 2 + 1) + 4𝑥𝑦 = 12𝑥
𝑑𝑥
To change it to the general formula divide all sides by (𝑥 2 + 1)

𝑑𝑦 4𝑥 12𝑥 4𝑥 12𝑥
+ 2 𝑦= 2 𝑃(𝑥) = 𝑞(𝑥) =
𝑑𝑥 𝑥 + 1 𝑥 +1 𝑥2+1 𝑥2 + 1

The integrating factor is (𝑥 2 + 1)2

And the solution is 𝑦 = 3 + 𝐾(𝑥 2 + 1)−2

Exercise for discussion

𝑑𝑦 𝑦
A. find the general solution + 𝑥 = 𝑥2
𝑑𝑥
𝑑𝑥
B. solve the initial value problem + 𝑡𝑥 = 4𝑡 𝑤𝑖𝑡ℎ 𝑥(0) = 2
𝑑𝑡
2
C. solve 𝑦 ′ + 𝑥 𝑦 = 4𝑥

Bernoulli equation

Some non linear differential equations can be reduced to linear form. The most famous of these
non linear differential equation is the Bernoulli equation which is given by the form 𝑦 ′ +
𝑅(𝑥)𝑦 = 𝑔(𝑥)𝑦 𝑛 where 𝑎 is any real number .

If 𝑛 = 0 𝑜𝑟 𝑛 = 1, the equation is linear otherwise it is non linear .then we set

𝑢(𝑥) = [𝑦(𝑥)]1−𝑛 Differentiating with respect to x we have

𝑢′ (𝑥) = (1 − 𝑛)[𝑦(𝑥)]1−𝑛−1 𝑦 ′ (𝑥) 𝑜𝑟 𝑢′ = (1 − 𝑛)𝑦 −𝑛 𝑦′

But 𝑦 ′ = 𝑔(𝑥)𝑦 𝑛 − 𝑅(𝑥)𝑦

𝑢′ = (1 − 𝑛)𝑦 −𝑛 [𝑔(𝑥)𝑦 𝑛 − 𝑅(𝑥)𝑦]

𝑢′ = (1 − 𝑛)𝑔(𝑥) − (1 − 𝑛)𝑅(𝑥)𝑦1−𝑛

93
 𝑢′ = (1 − 𝑛)[𝑔(𝑥) − 𝑅(𝑥)𝑦 1−𝑛 ] 𝑠𝑖𝑛𝑐𝑒 𝑢 = 𝑦 1−𝑛

𝑢′ = (1 − 𝑛)[𝑔(𝑥) − 𝑅(𝑥)𝑢(𝑥)]

𝑢′ + (1 − 𝑛)𝑅(𝑥)𝑢(𝑥) = (1 − 𝑛)𝑔(𝑥) this is a linear differential equation

So we can solve it!!

Example: Solve the special Bernoulli equation, called the Verhulst equation

𝑌 ′ − 𝐴𝑌 = −𝐵𝑌 2 (A, B are positive constants)

Solution: Let 𝑢(𝑥) = [𝑦(𝑥)]1−2 = 𝑦 −1

𝑢 = 𝑦 −1

𝑢′ = −𝑦 −2 𝑦 ′

𝑢′ = −𝑦 −2 (𝐴𝑦 − 𝐵𝑦 2 )

𝑢′ = [𝐵 − 𝐴𝑦 −1 ]

𝑢′ = 𝐵 − 𝐴𝑢

𝑢′ + 𝐴𝑢 = 𝐵

𝑑𝑢 − (𝐵 − 𝐴𝑢)𝑑𝑥 = 0

These is non exact hence we can use the formula to find the integrating factors

𝑃 = −(𝐵 − 𝐴𝑈) 𝑄=1

1 𝜕𝐹 1 𝜕𝑃 𝜕𝑄
= ( − )
𝐹 𝜕𝑥 𝑄 𝜕𝑢 𝜕𝑥
1 𝜕𝐹 1 𝜕 𝜕(1)
= ( (𝐴𝑢 − 𝐵) − )
𝐹 𝜕𝑥 1 𝜕𝑢 𝜕𝑥
1 𝜕𝐹
= 1(𝐴 − 0)
𝐹 𝜕𝑥
𝑑𝐹
= 𝐴𝑑𝑥
𝐹
ln 𝐹 = 𝐴𝑥 𝐹 = 𝑒 𝐴𝑥

𝑑𝑢 𝐴𝑥
𝑒 + 𝑒 𝐴𝑥 𝐴𝑢 = 𝐵𝑒 𝐴𝑥
𝑑𝑥
𝑑
[𝑢𝑒 𝐴𝑥 ] = 𝐵𝑒 𝐴𝑥
𝑑𝑥

94
 𝐵 𝐴𝑥
𝑢𝑒 𝐴𝑥 = 𝑒 +𝐶
𝐴
𝐵
𝑢= + 𝐶𝑒 −𝐴𝑥
𝐴
Then the general solution is given by
1 1
𝑦= =
𝑢 (𝐵 ) + 𝐶𝑒 −𝐴𝑥
𝐴
Exercise for discussion
a. solve𝑦 ′ + 𝑦 = 𝑦 2
𝑥
b. find the general solution 𝑦 ′ + 𝑦 = 𝑦
c. solve the DE 3𝑦 ′ + 𝑦 = (1 − 2𝑥)𝑦 4
d. * solve 𝑦 ′ 𝑐𝑜𝑠𝑦 + 𝑥𝑠𝑖𝑛𝑦 = 2𝑥

Non Linear Equation; Clairaut Equation

This is a non linear equation of the form 𝑦 = 𝑥𝑦 ′ + 𝑔(𝑦′)

In order to solve such equation let us differentiate with respect to x

We have 𝑦 ′ = 𝑦 ′ + 𝑥𝑦 ′′ + 𝑔′ (𝑦 ′ )𝑦 ′′

0 = 𝑦 ′′ [𝑥 + 𝑔′ (𝑦′)]

𝑦 ′′ = 0 𝑜𝑟 𝑥 + 𝑔′ (𝑦 ′ ) = 0

Solving 𝑦 ′′ = 0 gives the general solution 𝑦 = 𝑐1 𝑥 + 𝑐2

Solving 𝑥 + 𝑔′ (𝑦 ′ ) = 0 gives the singular solution

*Remark; the general solution of all Clairaut equations is 𝑦 = 𝑐1 𝑥 + 𝑐2

1
Example: Solve the Clairaut equation 𝑦 = 𝑥𝑦 ′ + 𝑦′

Derivation with respect to x gives

𝑦′′
𝑦 ′ = 𝑦 ′ + 𝑥𝑦 ′′ −
(𝑦′)2

1
0 = 𝑦 ′′ [𝑥 − ]
(𝑦′)2

Solving 𝑦 ′′ = 0 gives the general solution 𝑦 = 𝑐1 𝑥 + 𝑐2

95
 1
The singular solution is 𝑥 − (𝑦′)2 = 0

1
(𝑦′)2 =
𝑥
1
𝑦′ =
√𝑥

𝑦 = 2√𝑥 + 𝐶

Some Application Of First Order Linear Differential Equations

We can form ODE from basic physical application in engineering science

Example 1 Radio Active Exponential Decay

Experiments show that a radio active substance decomposes at a rate proportional to the amount
present. What can be said about the amount available at a later time? Starting with a given
amount of substance say, 2 grams at a certain time 𝑡 = 0

Example 2 Bacterial Growth

3
A culture has N0 number of bacteria at t=1 hr the number of bacteria is measured to be (2) N0 if
the rate of growth is proportional to the number of bacteria present, determine the time necessary
for the number of bacteria to triple

Example 3 Carbon Dating

1
A fossilized bone is found to contain 1000 the original amount of radioactive carbon C-14. given
the half life of C-14 is approximately 5600 years, Determine the age of the fossil

Example 4 Cooling And Warming Law

When a chicken is removed from an oven, its temperature is measured to be 3000F. Three
minutes later its temperature is 2000F. How long will it take for the chicken to cool off to a
temperature of 90.

2. Ordinary Differential Equation Of The Second Order

We shall now define and discuss linear ODEs of second order these equations have important
engineering applications, especially in connection with mechanical and electrical vibrations as
well as in wave motion, heat conduction, and other parts of physics

Ordinary differential equations of the second order is called linear if it has the form

96
 𝑦 ′′ + 𝑃1 (𝑥)𝑦 ′ + 𝑃0 (𝑥)𝑦 = 𝑟(𝑥)

If 𝑟(𝑥) = 0 then the equation becomes homogeneous and otherwise non homogeneous

Further more if the functions 𝑃1 (𝑥) and 𝑃0 (𝑥) are known as coefficients of the equation

Example: Identify the following equations (i.e. homogeneity and linearity)

a. 𝑦 ′′ + 4𝑦 = 𝑒 𝑥 sin 𝑥
b. (1 − 𝑥 2 )𝑦 ′′ − 2𝑥𝑦 ′ + 6𝑦 = 0
c. 𝑥(𝑦 ′′ 𝑦 + (𝑦 ′ )2 ) + 2𝑦 ′ 𝑦 = 0
d. 𝑦 ′′ = √(𝑦′)2 + 1

Operator notation and some of its properties

Let as define operator 𝐿 as 𝐷2 + 𝑎𝐷 + 𝑏 where 𝐷is the operator differentiation and 𝑎&𝑏 are
constants

𝐿(𝑦) = [𝐷2 + 𝑎𝐷 + 𝑏](𝑦) = 𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦

This operator is a linear operator because it satisfies

𝐿(𝛼𝑦 + 𝛽𝑤) = 𝛼𝐿(𝑦) + 𝛽𝐿(𝑤) for any constants 𝛼&𝛽 and any twice differentiable function
𝑦 𝑎𝑛𝑑 𝑤

The homogenous linear differential equation 𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 0 may now be written as

𝐿(𝑦) = 0 = [𝐷2 + 𝑎𝐷 + 𝑏](𝑦)

For homogeneous linear equations, we can obtain another new solution from known solutions by
scalar multiplication and addition

i.e. If 𝑦1 = 𝑒 𝑥 𝑎𝑛𝑑 𝑦2 = 𝑒 −𝑥 are solutions of 𝑦 ′′ − 𝑦 = 0 then 𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2

C1 andC2 are constants. This is called the linear combination of 𝑦1 &𝑦2

• Theorem:- The Fundamental Theorem For The Homogenous Equations

For a homogeneous linear differential equation, any linear combination of two solution
on an open interval I is again a solution on I. In particular for such an equation, sums and
constant multiples of solutions is again a solution.

Proof:-Let the homogenous linear equation be 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 and let 𝑦1 &𝑦2 be
solutions of this equation. Then by substitution 𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2 and its derivative 𝑦′ = 𝐶1 𝑦′1 +
𝐶2 𝑦′2 and 𝑦′′ = 𝐶1 𝑦′′1 + 𝐶2 𝑦′′2 in the equation we can show that Y is a solution.

97
Remark:- The above important theorem works only for homogenous ODEs only but it does not
hold for non homogeneous.

Example. Consider a non homogeneous second order linear equation 𝑦 ′′ + 𝑦 = 1

Now 𝑦1 = 1 + cos 𝑥 𝑎𝑛𝑑 𝑦2 = 1 + sin 𝑥 are solutions

But 𝑦 = 2𝑦1 + 𝑦2is not a solution

𝑦 = 2(1 + cos 𝑥) + (1 + sin 𝑥) = 3 + 2 cos 𝑥 + sin 𝑥

𝑦 ′ = −2 sin 𝑥 + cos 𝑥 𝑎𝑛𝑑 𝑦 ′′ = −2 cos 𝑥 − sin 𝑥

Substituting in the equation we have

(−2 cos 𝑥 − sin 𝑥) + (3 + 2 cos 𝑥 + sin 𝑥) = 1

3 ≠ 1 Hence y is not a solution

Reduction of order (how to find a second solution provided a solution)

Method of reduction of order is applicable to any equation of the form

𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 as follows

Let 𝑦1 be a solution &let 𝑦2 = 𝑢(𝑥)𝑦1 and the derivatives

𝑦′2 = 𝑢′𝑦1 + 𝑢𝑦1′ and 𝑦2 ′′ = 𝑢′′𝑦1 + 2𝑢′𝑦1′ + 𝑢𝑦1′′ into 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0,we have

𝑢′′ 𝑦1 + 𝑢′ (2𝑦1′ + 𝑝(𝑥)𝑦1 ) + 𝑢(𝑦1′′ + 𝑝(𝑥)𝑦1′ + 𝑞(𝑥)𝑦1 ) = 0,since we let 𝑦1 to be a solution

𝑦1′′ + 𝑝(𝑥)𝑦1′ + 𝑞(𝑥)𝑦1 = 0,then 𝑢′′ 𝑦1 + 𝑢′ (2𝑦1′ + 𝑝(𝑥)𝑦1 ) = 0

Now let 𝑢′ = 𝑧 => 𝑢′′ = 𝑧′

𝑧′𝑦1 + 𝑧(2𝑦1′ + 𝑝(𝑥)𝑦1 ) = 0, using separation of variables we have,

𝑧′ 2𝑦1′
= +𝑝
𝑧 𝑦1

Now integrating and taking constant zero

1
ln 𝑧 = −2 ln 𝑦1 − ∫ 𝑝 𝑑𝑥, ln 𝑧 = ln 𝑦1 −2 − ∫ 𝑝 𝑑𝑥 𝑧 = 𝑦 2 𝑒 − ∫ 𝑝𝑑𝑥
1

1 1
But 𝑧 = 𝑢′ => 𝑢′ = 𝑦 2 𝑒 − ∫ 𝑝𝑑𝑥 𝑢 = ∫ 𝑦 2 𝑒 − ∫ 𝑝𝑑𝑥 𝑑𝑥
1 1

𝑦2 1
=>𝑦2 = 𝑢𝑦1 , = ∫ 𝑦 2 𝑒 − ∫ 𝑝𝑑𝑥 𝑑𝑥
𝑦1 1

Example (𝑥 2 − 1)𝑦 ′′ − 2𝑥𝑦 ′ + 2𝑦 = 0

98
By inspection we have the first solution as 𝑦 = 𝑥 to find another independent solution,

we set 𝑦2 = 𝑢𝑦1 and write the equation in standard form (i.e. 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0)
2𝑥
𝑦 ′′ − (𝑥 2 −1) 𝑦 ′ + 2𝑦 = 0,

2𝑥
Thus − ∫ 𝑝 𝑑𝑥 = ∫ 𝑥 2 −1 𝑑𝑥 = ln|𝑥 2 − 1|

1 2 −1| 1
𝑢 = ∫ 𝑥 2 𝑒 ln|𝑥 𝑑𝑥 = ∫(1 − 𝑥 2 )𝑑𝑥 = 𝑥 + 𝑥

1
𝑦1 = 𝑥 𝑦2 = 𝑢𝑦1 = (𝑥 + ) 𝑥 = 𝑥 2 + 1
𝑥
Exercise for discussion

1. (𝑥 + 1)2 𝑦 ′′ − 2(𝑥 + 1)𝑦 ′ + 2𝑦 = 0 use 𝑦1 = 𝑥 + 1

′′ ′ − sin 𝑥
2. 𝑥𝑦 + 2𝑦 + 𝑥𝑦 = 0 use 𝑦1 = 𝑥≠0
𝑥
−1⁄
3. 𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ + (𝑥 2 − 1⁄4)𝑦 = 0 use 𝑦1 = 𝑥 2 cos 𝑥 𝑥>0
4. (𝑥 − 1)𝑦 ′′ − 𝑥𝑦 ′ + 𝑦 = 0 use 𝑦1 = 𝑒 𝑥
5.
2.1. Homogeneous Linear Differential Equations With Constant Coefficients

Consider 𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 0 where 𝑎 & 𝑏 are constants.

To solve this equation we recall from first order linear differential equations the equation of the
form 𝑌 ′ + 𝐾𝑌 = 0 with constant coefficient K has an exponential function as solution 𝑦 = 𝑒 −𝑘𝑥

From this letting 𝜆 = −𝐾 we have 𝑦 = 𝑒 𝜆𝑥

Substituting 𝑦 = 𝑒 𝜆𝑥 and its derivatives 𝑦′ = 𝜆𝑒 𝜆𝑥 and 𝑦′′ = 𝜆2 𝑒 𝜆𝑥 in the given second order
linear differential equation

𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 0

𝜆2 𝑒 𝜆𝑥 + 𝑎𝜆𝑒 𝜆𝑥 + 𝑏𝑒 𝜆𝑥 = 0

𝑒 𝜆𝑥 [𝜆2 + 𝑎𝜆 + 𝑏] = 0

Hence 𝑦 = 𝑒 𝜆𝑥 is a solution if 𝜆 is a solution of the quadratic equation

𝜆2 + 𝑎𝜆 + 𝑏 = 0

this is called the characteristic equation (or auxiliary equation) and its roots are
−𝑎±√𝑎2 −4𝑏
𝜆= 2

99
 −𝑎+√𝑎2 −4𝑏 −𝑎−√𝑎2 −4𝑏
i.e. 𝜆1 = 𝑎𝑛𝑑 𝜆2 =
2 2

Hence from these 𝑦1 = 𝑒 𝜆1 𝑥 and 𝑦2 = 𝑒 𝜆2 𝑥 are solutions

The general solution of second order ODEs is given by 𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2 and

To find the particular solution, unlike the first order ODEs (that need only a single initial value)
we need two initial or two boundary conditions to determine the values of 𝐶1 &𝐶2

Now from algebra we know that quadratic equation (characteristic equation) may have three
kinds of roots, depending on the sign of the discriminant 𝑎2 − 4𝑏 namely

Case 1 Two real roots if 𝑎2 − 4𝑏 > 0

Case 2 A real double root if 𝑎2 − 4𝑏 = 0

Case 3 Complex conjugate roots if 𝑎2 − 4𝑏 < 0

Case1:- The characteristic equation has two real roots 𝝀1and 𝝀2 i.e. 𝑎2 − 4𝑏 > 0

The solutions of the differential equation are 𝑦1 = 𝑒 𝜆1 𝑥 and 𝑦2 = 𝑒 𝜆2 𝑥

Since the equation is linear the general solution is given by 𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2

𝑦 = 𝐶1 𝑒 𝜆1 𝑥 + 𝐶2 𝑒 𝜆2 𝑥 is the solution if initial or boundary condition are given we can find the
particular solution.

Examples a. 𝑦 ′′ − 𝑦 = 0

Solution here let 𝑦 = 𝑒 𝜆𝑥

𝑦′′ = 𝜆2 𝑒 𝜆𝑥

𝑦 ′′ − 𝑦 = 𝜆2 𝑒 𝜆𝑥 − 𝑒 𝜆𝑥 = 0,
[𝜆2 − 1]𝑒 𝜆𝑥 = 0,

Since 𝑒 𝜆𝑥 can never be zero then 𝜆2 − 1 = 0

The discriminant √0 − 4(−1) = √4 = 2

−0 + √02 − 4(−1) −0 − √02 − 4(−1)

𝜆1 = =1 𝑎𝑛𝑑 𝜆2 = = −1
2 2
So 𝑦 = 𝑒 𝑥 and 𝑦 = 𝑒 −𝑥 are solutions and the general solution is given by 𝑦 = 𝐶1 𝑒 𝑥 + 𝐶2 𝑒 −𝑥

a. Solve the initial value problem 𝑦 ′′ + 𝑦 ′ − 2𝑦 = 0 𝑦(0) = 4 𝑦 ′ (0) = −5

100
Solution substitute 𝑦 = 𝑒 𝜆𝑥 and its derivatives 𝑦′ = 𝜆𝑒 𝜆𝑥 , 𝑦′′ = 𝜆2 𝑒 𝜆𝑥 in the given equation

𝜆2 𝑒 𝜆𝑥 + 𝜆𝑒 𝜆𝑥 − 2𝑒 𝜆𝑥 = 0 .

[𝜆2 + 𝜆 − 2]𝑒 𝜆𝑥 = 0 ,

Since 𝑒 𝜆𝑥 can never be zero then 𝜆2 + 𝜆 − 2 = 0

The discriminant √12 − 4(−2) = √9 = 3

−1 + √12 − 4(−2) −1 − √12 − 4(−2)

𝜆1 = =1 𝑎𝑛𝑑 𝜆2 = = −2
2 2
So 𝑦 = 𝑒 𝑥 and 𝑦 = 𝑒 −2𝑥 are solutions and

the general solution is given by 𝑦 = 𝐶1 𝑒 𝑥 + 𝐶2 𝑒 −2𝑥 and using the initial conditions

𝑦(0) = 4 = 𝐶1 + 𝐶2

𝑦 ′ = 𝐶1 𝑒 𝑥 − 2𝐶2 𝑒 −2𝑥 𝑦 ′ (0) = −5 = 𝐶1 − 2𝐶2

𝐶1 = 1 𝑎𝑛𝑑 𝐶2 = 3

Hence 𝑦 = 𝑒 𝑥 + 3𝑒 −2𝑥 is the particular solution satisfying the IVP

Exercises for discussion

1. Find the general solution of 𝑦 ′′ + 3𝑦 ′ + 2𝑦 = 0

2. Find the particular solution of 𝑦 ′′ − 9𝑦 = 0 𝑦(0) = 2 , 𝑦 ′ (0) = 0
−𝑎
Case 2:- The characteristic equation has double real root 𝜆 = 𝜆1 = 𝜆2 = that is 𝑎2 −
2
4𝑏 = 0
𝑎⁄ )𝑥
Hence the only solution is 𝑦1 = 𝑒 −( 2

To obtain a second solution we use reduction of order

Let 𝑦2 = 𝑢𝑦1 where u is a function of x, then substituting 𝑦2 = 𝑢𝑦1 and its derivatives

𝑦′2 = 𝑢′𝑦1 + 𝑢𝑦1′ ,and 𝑦′′2 = 𝑢′′𝑦1 + 2𝑢′𝑦1′ + 𝑢𝑦1′′ in the given DE 𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 0

(𝑢′′𝑦1 + 2𝑢′𝑦1′ + 𝑢𝑦1′′ ) + 𝑎(𝑢′𝑦1 + 𝑢𝑦1′ ) + 𝑏(𝑢𝑦1 ) = 0

Collecting like terms, we obtain

𝑢′′ 𝑦1 + 𝑢′ (2𝑦1′ + 𝑎𝑦1 ) + 𝑢(𝑦1′′ + 𝑎𝑦1′ + 𝑏𝑦1 ) = 0

101
Since 𝑦1 is a solution 𝑦1′′ + 𝑎𝑦1′ + 𝑏𝑦1 = 0
𝑎⁄ )𝑥 𝑎⁄ )𝑥
and since 𝑦1 = 𝑒 −( 2 𝑦1 ′ = −(𝑎⁄2)𝑒 −( 2

𝑦1′ = − 𝑎⁄2 𝑦1 , 2𝑦1′ + 𝑎𝑦1 = 0

We are left with only 𝑢′′ 𝑦1 = 0

𝑎⁄ )𝑥
Again 𝑦1 = 𝑒 −( 2 can never be zero 𝑢′′ = 0

=> 𝑢 = 𝐾1 𝑥 + 𝐾2 .
𝑎⁄ )𝑥 𝑎⁄ )𝑥
There fore 𝑦2 = 𝑢𝑦1 = (𝐾1 𝑥 + 𝐾2 )𝑒 −( 2 or taking 𝑢 = 𝑥 , 𝑦2 = 𝑥 𝑒 −( 2

Since the equation is linear the general solution is given by 𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2

𝑎 𝑎 𝑎
𝑦 = 𝐶1 𝑒 −( ⁄2)𝑥 + 𝐶2 𝑥𝑒 −( ⁄2)𝑥 𝑜𝑟 𝑦 = [𝐶1 + 𝐶2 𝑥]𝑒 −( ⁄2)𝑥 is the general solution if initial or
boundary condition are given we can find the particular solution.
𝑎⁄ )𝑥
Remark: 𝝀 must be a double root!! If 𝝀 is a simple root 𝑦 = [𝐶1 + 𝐶2 𝑥]𝑒 −( 2 is NOT a solution
of 𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 0

Example solve the general solution 𝑦 ′′ + 8𝑦 ′ + 16𝑦 = 0

Solution let 𝑦 = 𝑒 𝜆𝑥

Substitute 𝑦 = 𝑒 𝜆𝑥 and its derivatives 𝑦′ = 𝜆𝑒 𝜆𝑥 , 𝑦′′ = 𝜆2 𝑒 𝜆𝑥 in the given equation

𝑦 ′′ + 8𝑦 ′ + 16𝑦 = 0

𝜆2 𝑒 𝜆𝑥 + 8𝜆𝑒 𝜆𝑥 + 16𝑒 𝜆𝑥 = 0

[𝜆2 + 8𝜆 + 16]𝑒 𝜆𝑥 = 0

Since 𝑒 𝜆𝑥 can never be zero then the characteristic equation 𝜆2 + 8𝜆 + 16 = 0

(𝜆 + 4)2 = 0

𝝀= -4 is a double root this implies

Hence 𝑦1 = 𝑒 −4𝑥 and let 𝑦2 = 𝑢𝑦1 𝑦2′ = 𝑢′𝑦1 + 𝑢𝑦1′ 𝑦2′′ = 𝑢′′ 𝑦1 + 2𝑢′𝑦1′ + 𝑢𝑦1′′

Then
𝑦 ′′ + 8𝑦 ′ + 16𝑦 = 0

(𝑢′′ 𝑦1 + 2𝑢′𝑦1′ + 𝑢𝑦1′′ ) + 8(𝑢′𝑦1 + 𝑢𝑦1′ ) + 16𝑢𝑦1 = 0

102
Collecting like terms 𝑢′′ 𝑦1 + 𝑢′ (2𝑦1′ + 8𝑦1 ) + 𝑢(𝑦1′′ + 8𝑦1′ + 16𝑦1 ) = 0

Since 𝑦1 is a solution 𝑦1′′ + 8𝑦1′ + 16𝑦1 = 0

𝑦1 = 𝑒 −4𝑥 , 𝑦1′ = −4𝑒 −4𝑥 𝑦1′ = −4𝑦1 𝑦1′ + 4𝑦1 = 0

This is also 2(𝑦1′ + 4𝑦1 ) = 2𝑦1′ + 8𝑦1 = 0

Hence 𝑢′′ 𝑦1 = 0

=> 𝑢′′ = 0

𝑢 = 𝐾1 𝑥 + 𝐾2

Let 𝑢 = 𝑥, 𝑦2 = 𝑢𝑦1 , 𝑦2 = 𝑥𝑒 −4𝑥

There fore the general solution

𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2 ,

𝑦 = 𝐶1 𝑒 −4𝑥 + 𝐶2 𝑥𝑒 −4𝑥 , 𝑦 = [𝐶1 + 𝐶2 𝑥]𝑒 −4𝑥

Example: - Solve the initial value problem 𝑦 ′′ − 4𝑦 ′ + 4𝑦 = 0 𝑦(0) = 3 𝑦 ′ (0)=1

Solution let 𝑦 = 𝑒 𝜆𝑥

Substitute 𝑦 = 𝑒 𝜆𝑥 and its derivatives 𝑦′ = 𝜆𝑒 𝜆𝑥 , 𝑦′′ = 𝜆2 𝑒 𝜆𝑥 in the given equation

𝑦 ′′ − 4𝑦 ′ + 4𝑦 = 0

𝜆2 𝑒 𝜆𝑥 − 4𝜆𝑒 𝜆𝑥 + 4𝑒 𝜆𝑥 = 0

[𝜆2 − 4𝜆 + 4]𝑒 𝜆𝑥 = 0

Characteristic equation 𝜆2 − 4𝜆 + 4 = 0

(𝜆 − 2)2 = 0

𝝀=2 is a double root

𝑦1 = 𝑒 2𝑥 , hence 𝑦2 = 𝑥𝑦1 𝑙𝑒𝑡 𝑢 = 𝑥

𝑦2 = 𝑥𝑒 2𝑥 ,

𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2 = 𝐶1 𝑒 2𝑥 + 𝐶2 𝑥𝑒 2𝑥 = [𝐶1 + 𝐶2 𝑥]𝑒 2𝑥 ,

Then 𝑦(0) = [𝐶1 + 𝐶2 (0)]𝑒 2𝑥0 = 3

𝐶1 = 3,
103
𝑦 ′ = 2𝐶1 𝑒 2𝑥 + 2𝐶2 𝑥𝑒 2𝑥 + 𝐶2 𝑒 2𝑥 ,

𝑦 ′ (0) = 2𝐶1 + 𝐶2 = 1 𝐶2 = −5

𝑦 = [3 − 5𝑥]𝑒 2𝑥

Exercise for discussion

Solve the following questions

1. 𝑦 ′′ − 6𝑦 ′ + 9𝑦 = 0
2. 𝑦 ′′ + 10𝑦 ′ + 25𝑦 = 0

Case 3 :-The characteristic equation has two complex conjugate roots

i.e. 𝑎2 − 4𝑏 < 0

if the characteristic equation of 𝜆2 + 𝑎𝜆 + 𝑏 = 0 has the discriminant 𝑎2 − 4𝑏 < 0

we put √−1 = 𝑖from the roots we have, −(4𝑏 − 𝑎2 ) > 0

1
let 𝜔 = √𝑏 − 𝑎2 ⁄4 = 2 √𝑎2 − 4𝑏

−𝑎 1 −𝑎 1
then 𝜆1 = + 2 √𝑎2 − 4𝑏 𝜆2 = − 2 √𝑎2 − 4𝑏
2 2

−𝑎 1 −𝑎 1
𝜆1 = + 𝜔𝑖, 𝜆2 = − 𝜔𝑖
2 2 2 2
−𝑎 1 −𝑎
Hence 𝑦1 = 𝑒 𝜆1 𝑥 𝑦1 = 𝑒 ( 2 +2𝜔𝑖)𝑥 𝑦1 = 𝑒 ( 2 )𝑥 cos 𝜔𝑥
−𝑎 1 −𝑎
𝑦2 = 𝑒 𝜆2 𝑥 𝑦2 = 𝑒 ( 2 −2𝜔𝑖)𝑥 𝑦2 = 𝑒 ( 2 )𝑥 sin 𝜔𝑥
𝑦
Remark 𝑦1 and 𝑦2 are linearly independent so that 𝑦2 = tan 𝜔𝑥 is not constant since 𝜔 ≠ 0
1

−𝑎 −𝑎
The general solution is given by 𝑦 = 𝐶1 𝑦1 + 𝐶2 𝑦2 = 𝐶1 𝑒 ( 2 )𝑥 cos 𝜔𝑥 + 𝐶2 𝑒 ( 2 )𝑥 sin 𝜔𝑥
−𝑎
𝑦 = 𝑒 ( 2 )𝑥 [𝐴 cos 𝜔𝑥 + 𝐵 sin 𝜔𝑥]

Example 1 Find the solution for 𝑦 ′′ − 2𝑦 ′ + 10𝑦 = 0

𝜆2 − 2𝜆 + 10 = 0

−(−2) 1 −(−2) 1
𝜆1 = + √(−2)2 − 4(10) 𝜆2 = − √(−2)2 − 4(10)
2 2 2 2

104
 𝜆1 = 1 + 3𝑖 𝜆2 = 1 − 3𝑖

𝑦1 = 𝑒 𝜆1 𝑥 = 𝑒 (1+3𝑖)𝑥 = 𝑒 𝑥 𝑒 3𝑥𝑖 = 𝑒 𝑥 cos 3𝑥

𝑦2 = 𝑒 𝜆2 𝑥 = 𝑒 (1−3𝑖)𝑥 = 𝑒 𝑥 𝑒 −3𝑥𝑖 = 𝑒 𝑥 sin 3𝑥

There fore the general solution 𝑦 = 𝑒 𝑥 [𝐴 cos 3𝑥 + 𝐵 sin 3𝑥]

Example 2 Solve the initial value problem 𝑦 ′′ + 2𝑦 ′ + 5𝑦 = 0 𝑦(0) = 1 𝑦 ′ (0) = 5

Exercise for discussion

a. Solve the boundary value problem 𝑦 ′′ + 𝑦 = 0 𝑦(0) = 3 𝑦(𝜋) = −3

b. Solve the general solution of 𝑦 ′′ + 10𝑦 ′ + 29𝑦 = 0
c. Solve the initial value problem 𝑦 ′′ + 4𝑦 ′ + 4.25𝑦 = 0 𝑦(0) = 1 𝑦 ′ (0) = −2
1
d. Solve the boundary value problem 𝑦 ′′ − 2𝑦 ′ = 0 𝑦(0) = −1 𝑦 (2) = 𝑒 − 2
e. 𝑦 ′′ + 7𝑦 ′ + 10𝑦 = 0
f. 𝑦 ′′ − 8𝑦 ′ + 16𝑦 = 0
g. 𝑦 ′′ + 2𝑦 ′ + 3𝑦 = 0

Non Homogeneous Equation

Ordinary differential equations of the second order is called linear if it has the form

𝑦 ′′ + 𝑃1 (𝑥)𝑦 ′ + 𝑃0 (𝑥)𝑦 = 𝑟(𝑥) 𝑤ℎ𝑒𝑟𝑒 𝑟(𝑥) ≠ 0

Is known as non homogeneous

To solve such non homogeneous equations we need the following theorem

Theorem Relation between solutions as non homogeneous and homogeneous equations

a) The difference of two solutions of non homogenous equation on some open interval I is a
solution of homogeneous equation on I
b) The sum of a solution of non homogeneous and a solution of homogenous equation on I
is a solution of the non homogeneous on the interval I

Proof: - let non homogeneous equation be 𝑦 ′′ + 𝑃1 (𝑥)𝑦 ′ + 𝑃0 (𝑥)𝑦 = 𝑟(𝑥) and the
homogeneous equation be 𝑦 ′′ + 𝑃1 (𝑥)𝑦 ′ + 𝑃0 (𝑥)𝑦 = 0

a) If we denote 𝑦 ′′ + 𝑃1 (𝑥)𝑦 ′ + 𝑃0 (𝑥)𝑦 = 𝐿(𝑦) and let 𝑦 &𝑦̅ be any solution of 𝑦 ′′ +

𝑃1 (𝑥)𝑦 ′ + 𝑃0 (𝑥)𝑦 = 𝑟(𝑥)then
𝐿(𝑦) = 𝑟(𝑥) 𝐿(𝑦̅) = 𝑟(𝑥) the difference of these two solutions
𝑦 − 𝑦̅, (𝑦 − 𝑦̅) = 𝑦 ′ − 𝑦̅ ′ ,
′ (𝑦 − 𝑦̅)′′ = 𝑦 ′′ − 𝑦̅′′
𝐿(𝑦 − 𝑦̅) = 𝑦 ′′ − 𝑦̅ ′′ + 𝑃(𝑥)𝑦 ′ − 𝑃(𝑥)𝑦̅ ′ + 𝑞(𝑥)𝑦 − 𝑞(𝑥)𝑦̅

105
 = 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 − [𝑦̅ ′′ + 𝑃(𝑥)𝑦̅ ′ + 𝑞(𝑥)𝑦̅] = 𝑟(𝑥) − 𝑟(𝑥) = 0
b) Similarly for 𝑦 a solution of the non homogeneous equation and 𝑦 ∗ any solution of the
homogeneous
𝑦 + 𝑦 ∗, (𝑦 + 𝑦 ∗)′ = 𝑦 ′ + 𝑦 ∗′ (𝑦 + 𝑦 ∗)′′ = 𝑦 ′′ + 𝑦 ∗′′
𝐿(𝑦 + 𝑦 ∗) = 𝐿(𝑦) + 𝑙(𝑦 ∗) = 𝑟(𝑥) + 0 = 𝑟(𝑥).
2.2.Non Homogeneous Equations With Constant Coefficients

A general solution of the non homogenous equation on some open interval I is a solution of the
form

𝑦(𝑥) = 𝑦ℎ (𝑥) + 𝑦𝑝 (𝑥)

where 𝑦ℎ (𝑥) = 𝐶1 𝑦1 (𝑥) + 𝐶2 𝑦2 (𝑥) is a general solution of the homogeneous equation on I and
𝑦𝑝 (𝑥) is any solution of the non homogeneous equation on I containing NO arbitrary constants.

A particular solution of non homogeneous equation on I is a solution obtained from the general
solution 𝑦(𝑥) = 𝑦ℎ (𝑥) + 𝑦𝑝 (𝑥) by assigning specific values to the arbitrary constants𝐶1 & 𝐶2 of
𝑦ℎ (𝑥)

Theorem : Suppose that the coefficient 𝑃(𝑥), 𝑞(𝑥) 𝑎𝑛𝑑 𝑟(𝑥) are continuous on an open interval
I. Then every solution of non homogeneous equation on I is obtained by assigning suitable
values to the arbitrary constants in a general solution 𝑦(𝑥) = 𝑦ℎ (𝑥) + 𝑦𝑝 (𝑥) of the non
homogeneous equation on I.

2.2.1 The method of undetermined coefficient

Rules for solving non homogeneous equations using the method of undetermined coefficient

(A) Basic rule :- if 𝑟(𝑥)in 𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 𝑟(𝑥) (where 𝑎 & 𝑏 are constants) is one of the
function of the form exponential, polynomial, cosine sine or the sum or product of such
functions, choose the corresponding 𝑦𝑝 as the following and determine the undetermined
coefficient by substituting 𝑦𝑝 and its derivatives in the given non homogeneous equation.

106
(B) Modification rule:- If a term in the choice of 𝑦𝑝 happens to be a solution of the
homogeneous equation corresponding to 𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 𝑟(𝑥) then multiply the choice
𝑦𝑝 by 𝑥 (or 𝑥 2 if the solution corresponds to a double root of the characteristic equation
of the homogeneous equation).
(C) Sum rule:-If 𝑟(𝑥) is a sum of function listed then choose 𝑦𝑝 the sum of the
corresponding function listed

Example solve initial value problem

𝑦 ′′ − 4𝑦 ′ + 3𝑦 = 10𝑒 −2𝑥 𝑦(0) = 1 𝑦 ′ (0) = −3

Solution

Step I 𝑦 ′′ − 4𝑦 ′ + 3𝑦 = 0 the general solution of the homogeneous is

𝜆2 − 4𝜆 + 3 = 0

𝜆=3 𝜆=1

𝑦1 = 𝑒 𝑥 𝑦2 = 𝑒 3𝑥

And the general solution of the homogenous equation is

𝑦ℎ = 𝐶1 𝑒 𝑥 + 𝐶2 𝑒 3𝑥

Step II since 𝑒 −2𝑥 has derivatives 𝐾𝑒 −2𝑥

𝑦𝑝 = 𝐾𝑒 −2𝑥 , 𝑦𝑝′ = −2𝐾𝑒 −2𝑥 , 𝑦𝑝′′ = 4𝐾𝑒 −2𝑥 , 𝑠𝑢𝑏𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑛𝑔 𝑦𝑝 , 𝑦𝑝 ′&𝑦𝑝 ′′in the non
homogeneous equation we have,𝑦 ′′ − 4𝑦 ′ + 3𝑦 = 10𝑒 −2𝑥

𝑦𝑝′′ − 4 𝑦𝑝′ + 3𝑦𝑝 = 10𝑒 −2𝑥

107
 4𝐾𝑒 −2𝑥 − 4(−2𝐾𝑒 −2𝑥 ) + 3𝐾𝑒 −2𝑥 = 10𝑒 −2𝑥

2
𝐾= => 𝑦𝑝 = 2/3𝑒 −2𝑥
3
Hence the general solution of the non homogeneous equation is given by
𝑦(𝑥) = 𝑦ℎ (𝑥) + 𝑦𝑝 (𝑥)

𝑦(𝑥) = 𝐶1 𝑒 𝑥 + 𝐶2 𝑒 3𝑥 + 2/3𝑒 −2𝑥

Step III using initial conditions the particular solution is given by

2 4
𝑦(0) = 𝐶1 + 𝐶2 + = 1 𝑦 ′ (0) = 𝐶1 + 3𝐶2 − = −3 , using
3 3
4
these equations simultaneously we have𝐶1 = 3 , 𝐶2 = −1

𝑦 = −1𝑒 𝑥 + 4/3𝑒 3𝑥 + 2/3𝑒 −2𝑥

Example 2 solve 𝑦 ′′ + 4𝑦 = 8𝑥 2

Exercises for discussion

1. Solve the initial value problem 𝑦 ′′ − 2𝑦 ′ + 𝑦 = 𝑒 𝑥 + 𝑥 𝑦(0) = 1 𝑦 ′ (0) = 0

2. Solve the general solution 𝑦 ′′ − 3𝑦 ′ + 2𝑦 = 𝑒 𝑥
3. Solve the general solution 𝑦 ′′ + 2𝑦 ′ + 5𝑦 = 16𝑒 𝑥 + sin 2𝑥
4. Solve the general solution𝑦 ′′ − 4𝑦 ′ + 3𝑦 = 2 sin 𝑥 − 4 cos 𝑥
5. Solve the general solution𝑦 ′′ − 4𝑦 = 2 sinh 2𝑥 + 𝑥

Linear independence and Wronskians

Consider 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 , 𝑦(𝑥0 ) = 𝑘𝑜 𝑦 ′ (𝑥0 ) = 𝑘1 with continuous but arbitrary

variable coefficient p and q.

Theorem existence and uniqueness theorem for initial value problem

If 𝑝(𝑥)& 𝑞(𝑥) are continuous functions on some open interval I, and 𝑥0 is in I, Then the IVP has
a unique solution 𝑦(𝑥)on I.

The general solution of 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 are made up of the basis 𝑦1 &𝑦2 that is a pair
of linearly independent solutions. If 𝐾1 𝑦1 (𝑥) + 𝐾2 𝑦2 (𝑥) = 0 𝑜𝑛 𝐼 => 𝐾1 = 0, 𝐾2 = 0

Then 𝑦1 &𝑦2 are linearly independent on I and if 𝐾1 𝑦1 (𝑥) + 𝐾2 𝑦2 (𝑥) = 0 on I with 𝐾1 and𝐾2

Not both zero then 𝑦1 = 𝐾𝑦2 or 𝑦2 = 𝑙𝑦1 hence linearly dependant.

To determine the linear independence &dependence of solution we use the Wronskian

determinant or briefly, the wronskian of two solutions 𝑦1 and𝑦2 , defined by

108
 𝑦1 𝑦2 ′ ′
𝑊(𝑦1 , 𝑦2 ) = [𝑦 ′ 𝑦2′ ] = 𝑦1 𝑦2 − 𝑦2 𝑦1
1

Theorem linear dependence and independence of solutions

Suppose that 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 has continuous coefficients 𝑃(𝑥)&𝑞(𝑥) on an open

interval I. Then the two solutions 𝑦1 and𝑦2 are linearly dependent if and only if 𝑊 = 0on some
𝑥0 in I. furthermore if 𝑊 = 0 for 𝑥 = 𝑥0 , then 𝑊 = 0 on I.

Hence if there is an 𝑥1 in I at which 𝑊 ≠ 0 then 𝑦1 and𝑦2 are linearly independent on I.

Remark A basis is of solution of 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 on an interval I is a pair 𝑦1 , 𝑦2 of

linearly independent solutions on I.

General solutions of homogeneous linear equation 𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 with continuous

coefficients 𝑃(𝑥)&𝑞(𝑥) on some open interval I, then every solution 𝑦(𝑥) is of the form 𝑦 =
𝐶1 𝑦1 (𝑥) + 𝐶2 𝑦2 (𝑥) where 𝑦1 𝑎𝑛𝑑𝑦2are basis of solutions and 𝐶1 𝑎𝑛𝑑𝐶2 are suitable constants.

Hence such equation does not have singular solutions.

2.2.2 Solution by variation of parameters

This method of variation of parameters apply to non homogeneous equations of the form 𝑦 ′′ +
𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 𝑟(𝑥) with arbitrary variable functions 𝑃(𝑥), 𝑞(𝑥)& 𝑟(𝑥)that are continuous
on some interval I

The method gives a particular solution 𝑦𝑝 on I in the form

𝑦2 𝑟(𝑥) 𝑦1 𝑟(𝑥)
𝑦𝑝 (𝑥) = −𝑦1 ∫ 𝑑𝑥 + 𝑦2 ∫ 𝑑𝑥
𝑊 𝑊

Where 𝑦1 &𝑦2 form basis solutions of the homogeneous equation𝑦 ′′ + 𝑃(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0, and
𝑊 = 𝑦1 𝑦2′ − 𝑦1′ 𝑦2 wroniskian determinant of 𝑦1 &𝑦2

Example: Solve the initial value problem

𝑦 ′′ + 𝑦 = sec 𝑥

Exercises for discussion

3⁄ 𝑥
a. Find the general solution of 𝑦 ′′ − 2𝑦 ′ + 𝑦 = 𝑥 2𝑒

b. 𝑦 ′′ + 4𝑦 = 2 sec 2𝑥

Cauchy Euler equations

109
Cauchy Euler differential equations homogeneous second order DE with variable coefficient is
of the form 𝑥 2 𝑦 ′′ + 𝑎𝑥𝑦 ′ + 𝑏𝑦 = 0 (𝑎, 𝑏 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡𝑠)

We can solve such equations by a pure algebraic multiplication i.e. let 𝑦 = 𝑥 𝑚 and its derivatives
𝑦 ′ = 𝑚𝑥 𝑚−1 , 𝑦 ′′ = 𝑚(𝑚 − 1)𝑥 𝑚−2

𝑥 2 𝑚(𝑚 − 1)𝑥 𝑚−2 + 𝑎𝑥𝑚𝑥 𝑚−1 + 𝑏𝑥 𝑚 = 0

𝑚(𝑚 − 1)𝑥 𝑚 + 𝑎𝑚𝑥 𝑚 + 𝑏𝑥 𝑚 = 0

𝑥 𝑚 [𝑚(𝑚 − 1) + 𝑎𝑚 + 𝑏] = 0

𝑚2 − 𝑚 + 𝑎𝑚 + 𝑏 = 0, this is known as the auxiliary equation

Since the auxiliary equation is quadratic we have three cases depending on the discriminant

Case 1 two real roots if (𝑎 − 1)2 − 4𝑏 > 0

Case 2 a real double root if (𝑎 − 1)2 − 4𝑏 = 0

Case 3 Complex conjugate roots if (𝑎 − 1)2 − 4𝑏 < 0

Case 1 two distinct real roots 𝑚1 &𝑚2 then 𝑦1 = 𝑥 𝑚1 𝑎𝑛𝑑 𝑦2 = 𝑥 𝑚2 here 𝑦1 &𝑦2 constitute a
basis of solution of the differential equation for all x. hence the general solution is given by

𝑦 = 𝐶1 𝑥 𝑚1 + 𝐶2 𝑥 𝑚2

Example solve 𝑥 2 𝑦 ′′ − 2𝑥𝑦 ′ + 2𝑦 = 0

Solution letting 𝑦 = 𝑥 𝑚 , we have the auxiliary equation 𝑚(𝑚 − 1) − 2𝑚 + 2 = 0

𝑚2 − 𝑚 − 2𝑚 + 2 = 𝑚2 − 3𝑚 + 2 = (𝑚 − 1)(𝑚 − 2) = 0

𝑚1 = 1 𝑚2 = 2 => 𝑦1 = 𝑥 𝑦2 = 𝑥 2

The general solution is 𝑦 = 𝐶1 𝑥 + 𝐶2 𝑥 2

Case 2 Double root

If the auxiliary equation 𝑚2 + (𝑎 − 1)𝑚 + 𝑏 = 0 has a double root i.e.𝑚 = 𝑚1 = 𝑚2 =

1/2(1 − 𝑎). Then the first solution is 𝑦1 = 𝑥1/2(1−𝑎) and a second solution 𝑦2 can be obtained by
the method of reduction of order

Let 𝑦2 = 𝑢𝑦1 where u is a function of x, then substituting 𝑦2 = 𝑢𝑦1 and its derivatives

𝑦′2 = 𝑢′𝑦1 + 𝑢𝑦1′ , and 𝑦′′2 = 𝑢′′𝑦1 + 2𝑢′𝑦1′ + 𝑢𝑦1′′ in the given DE 𝑥 2 𝑦 ′′ + 𝑎𝑥𝑦 ′ + 𝑏𝑦 = 0

110
 𝑥 2 (𝑢′′𝑦1 + 2𝑢′𝑦1′ + 𝑢𝑦1′′ ) + 𝑎𝑥(𝑢′𝑦1 + 𝑢𝑦1′ ) + 𝑏(𝑢𝑦1 ) = 0

𝑢′′𝑥 2 𝑦1 + 𝑢′ (2𝑥 2 𝑦1′ + 𝑎𝑥𝑦1 ) + 𝑢(𝑥 2 𝑦1′′ + 𝑎𝑥𝑦1′ + 𝑏𝑦1 ) = 0

Since 𝑦1 is a solution 𝑥 2 𝑦1′′ + 𝑎𝑥𝑦1′ + 𝑏𝑦1 = 0

1 1 1
1 (1−𝑎)−1 1 (1−𝑎) −1
𝑦1 = 𝑥 2(1−𝑎) , 𝑦1′ = 2 (1 − 𝑎)𝑥 2 = 2 (1 − 𝑎)𝑥 2 𝑥

(1−𝑎)
𝑦1′ = 𝑦1 , => 2𝑥𝑦1′ = (1 − 𝑎)𝑦1 2𝑥𝑦1′ + 𝑎𝑦1 = 𝑦1 multiplying this equation by 𝑥 we
2𝑥
have 𝑥(2𝑥𝑦1′ + 𝑎𝑦1 = 𝑦1 ) 2𝑥 2 𝑦1′ + 𝑎𝑥𝑦1 = 𝑥𝑦1

Thus the equation reduces to

𝑢′′𝑥 2 𝑦1 + 𝑢′ 𝑥𝑦1 = 0,
𝑦1 [𝑢′′𝑥 2 + 𝑢′ 𝑥] = 0, =>𝑦1 = 0 𝑎𝑛𝑑 𝑢′′𝑥 2 + 𝑢′ 𝑥 = 0
𝑢′′ −𝑥 −1
𝑢′′𝑥 2 = −𝑢′ 𝑥, = =
𝑢′ 𝑥2 𝑥

𝑘1
ln|𝑢′| = − ln|𝑥| + 𝑘, 𝑢′ = 𝑢 = 𝑘1 ln 𝑥 + 𝑘2
𝑥

Hence in the case of double root the basis of solution is

1 1
𝑦1 = 𝑥 2(1−𝑎) 𝑎𝑛𝑑 𝑦2 = (ln 𝑥) 𝑥 2(1−𝑎) ,

There fore the general solution is given by

1
𝑦 = 𝐶1 𝑥 𝑚 + 𝐶2 𝑥 𝑚 ln 𝑥, 𝑦 = 𝑥 2(1−𝑎) [𝐶1 + 𝐶2 ln 𝑥] 𝐶1 &𝐶2 are constants

Example: Solve the DE 𝑥 2 𝑦 ′′ + 9𝑥𝑦 ′ + 16𝑦 = 0

Solution: Let 𝑦 = 𝑥 𝑚 and its derivatives 𝑦 ′ = 𝑚𝑥 𝑚−1 , 𝑦 ′′ = 𝑚(𝑚 − 1)𝑥 𝑚−2

Substitute them in the given equation 𝑥 2 (𝑚(𝑚 − 1)𝑥 𝑚−2 ) + 9𝑥(𝑚𝑥 𝑚−1 ) + 16𝑥 𝑚 = 0

𝑚2 − 𝑚 + 9𝑚 + 16 = 0, 𝑚2 + 8𝑚 + 16 = 0 (𝑚 + 4)2 = 0 𝑚 = −4

𝑦1 = 𝑥 −4, and 𝑦2 = 𝑢 𝑦1 = 𝑥 −4 ln 𝑥

So the general solution is 𝑦 = 𝐶1 𝑥 −4 + 𝐶2 𝑥 −4 ln 𝑥 = 𝑥 −4 [𝐶1 + 𝐶2 ln 𝑥],

Case 3 complex conjugate roots

If the auxiliary equation 𝑚2 + (𝑎 − 1)𝑚 + 𝑏 = 0 has the discriminant (𝑎 − 1)2 − 4𝑏 < 0

Then the roots are complex conjugate roots ,i.e. if𝑚1 = 𝑃 + 𝑖𝑞, 𝑚2 = 𝑃 − 𝑖𝑞

111
 𝑚
Recall that 𝑦 = 𝑥 𝑚 = 𝑒 ln 𝑥 = 𝑒 𝑚 ln 𝑥 = 𝑒 (𝑃+𝑖𝑞) ln 𝑥 = 𝑒 𝑝 ln 𝑥 𝑒 𝑖𝑞 ln 𝑥 = 𝑥 𝑝 (cos(𝑞 ln 𝑥))

In this case we claim the basis of solution for the differential equation for all positive x is

𝑦1 = 𝑥 𝑃 𝑥 𝑖𝑞 = 𝑥 𝑝 cos(𝑞 ln 𝑥), and 𝑦2 = 𝑥 𝑃 𝑥 −𝑖𝑞 = 𝑥 𝑝 sin(𝑞 ln 𝑥)

Hence the general solution is 𝑦 = 𝑥 𝑝 (𝐴 cos(𝑞 ln 𝑥) + 𝐵 sin(𝑞 ln 𝑥))

Example solve 𝑥 2 𝑦 ′′ + 7𝑥𝑦 ′ + 13𝑦 = 0

Solution let 𝑦 = 𝑥 𝑚 and its derivatives 𝑦 ′ = 𝑚𝑥 𝑚−1 , 𝑦 ′′ = 𝑚(𝑚 − 1)𝑥 𝑚−2

Substitute them in the given equation 𝑥 2 (𝑚(𝑚 − 1)𝑥 𝑚−2 ) + 7𝑥(𝑚𝑥 𝑚−1 ) + 13𝑥 𝑚 = 0

𝑚2 − 𝑚 + 7𝑚 + 13 = 0, 𝑚2 + 6𝑚 + 13 = 0 (7 − 1)2 − 4(13) = −16 < 0

𝑚1 = −3 + 2𝑖 𝑚2 = −3 − 2𝑖.

𝑦1 = 𝑥 𝑚1 = 𝑥 −3+2𝑖 = 𝑥 −3 cos(2 ln 𝑥) 𝑎𝑛𝑑 𝑦2 = 𝑥 𝑚2 = 𝑥 −3−2𝑖 = 𝑥 −3 sin(2 ln 𝑥)

Hence the general solution is 𝑦 = 𝑥 −3 (𝐴 cos(2 ln 𝑥) + 𝐵 sin(2 ln 𝑥))

Exercises for discussion

Solve the following Cauchy Euler equations

−1
a. 4𝑥 2 𝑦 ′′ + 4𝑥𝑦 ′ − 𝑦 = 0 𝑦(4) = 2 𝑦 ′ (4) = 4
2 ′′ ′ ′ (1)
b. 𝑥 𝑦 − 𝑥𝑦 + 2𝑦 = 0 𝑦(1) = −1 𝑦 = −1
c. 𝑥 2 𝑦 ′′ + 3𝑥𝑦 ′ + 𝑦 = 0
1 1
d. 𝑦 ′′ + 𝑥 𝑦 ′ + (1 − 4𝑥 2 ) 𝑦 = 𝑥

112