MATHEMATICAL ANALYSIS

Division of Mathematics Faculty of Education

Suan Sunandha Rajabhat University

2022
MATHEMATICAL ANALYSIS

Thanatyod Jampawai
Assistant Professor of Mathematics
Suan Sunandha Rajabhat University

Division of Mathematics, Faculty of Education

Suan Sunandha Rajabhat University
Bangkok, Thailand
Update : November 2022
Contents

1 The Real Number System 1

1.1 Ordered field axioms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Well-Ordering Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.3 Completeness Axiom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.4 Functions and Inverse functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2 Sequences in R 37
2.1 Limits of sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.2 Limit theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.3 Bolzano-Weierstrass Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.4 Cauchy sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

3 Topology on R 71
3.1 Open sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.2 Closed sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.3 Limit points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4 Limit of Functions 85
4.1 Limit of Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.2 One-sided limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.3 Infinite limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

5 Continuity on R 103
5.1 Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.2 Intermediate Value Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

i
ii CONTENTS

5.3 Uniform continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

6 Differentiability on R 121
6.1 The Derivative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.2 Differentiability theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
6.3 Mean Value Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
6.4 Monotone function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

7 Integrability on R 149
7.1 Riemann integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
7.2 Riemann sums . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
7.3 Fundamental Theorem of Calculus . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

8 Infinite Series of Real Numbers 181

8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
8.2 Series with nonnegative terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
8.3 Absolute convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
8.4 Alternating series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Chapter 1

The Real Number System

1.1 Ordered field axioms

FIELD AXIOMS.
There are functions + and ·, defined on R2 , that satisfy the following properties for every a, b, c ∈ R:

F1 Closure Properties a + b and a · b belong to R.

F2 Associative Properties a + (b + c) = (a + b) + c
a · (b · c) = (a · b) · c
F3 Commutative Properties a + b = b + a and a · b = b · a
F4 Distributive Properties a · (b + c) = a · b + a · c
(b + c) · a = b · a + c · a
F5 Additive Identity There is a unique element 0 ∈ R such that
0 + a = a = a + 0 for all a ∈ R .
F6 Multiplicative Identity There is a unique element 1 ∈ R such that
1 · a = a = a · 1 for all a ∈ R.
F7 Additive Inverse For every x ∈ R there is a unique −x ∈ R such that
x + (−x) = 0 = (−x) + x.
F8 Multiplicative Inverse For every x ∈ R\{0} there is a unique x−1 ∈ R such that
x · (x−1 ) = 1 = (x−1 ) · x.
We shall frequently denote
1 a
a + (−b) by a − b, a · b by ab, a−1 by and a · b−1 by .
a b
2 CHAPTER 1. THE REAL NUMBER SYSTEM

The real number system R contains certain special subsets: the set of natural numbers

N := {1, 2, 3, ...}

obtained by begining with 1 and successively adding 1’s to form 2 := 1 + 1, 3 := 2 + 1, etc,; the
set of integers
Z := {..., −2, −1, 0, 1, 2, ...}

(Zahlen is German for number); the set of rationals (or fractions or quoteints)
{ }
p
Q := : p, q ∈ Z and q ̸= 0
q
and the set of irrationals
Qc := R\Q.

Equality in Q is defined by
m p
= if and only if mq = np.
n q
Recall that each of the sets N, Z, Q and R is a proper subset of the next; i.e.,

N ⊂ Z ⊂ Q ⊂ R.

Definition 1.1.1 Let a ∈ R and n ∈ N. Define

| · a {z
an = a · ... · a}
n− copies

a and n are called base and exponent, respectively.

Definition 1.1.2 Let a be a non-zero real number. Define

1
a0 = 1 and a−n = for n ∈ N
an
Theorem 1.1.3 Let a, b ∈ R and n, m ∈ Z. Then

1. (ab)n = an bn
( a )n an
2. = n where b ̸= 0
b b
3. an · am = am+n
an
4. = an−m where a ̸= 0
am
Proof. Excercise.
1.1. ORDERED FIELD AXIOMS 3

Theorem 1.1.4 Let a be a real number. Then

1. 0a = 0 3. −(−a) = a

2. (−1)a = −a 4. (a−1 )−1 = a where a ̸= 0

Proof. Let a be a real number. We first consider

0a = (0 + 0)a ( by F5 )

= 0a + 0a ( by F4 )

By F5, it implies that 0a = 0. This result leds to

0 = 0a ( by 1. )

= (1 + (−1))a ( by F7 )

= 1a + (−1)a ( by F4 )

= a + (−1)a ( by F6 )

By F7, (−1)a is an additive inverse of a. Thus, (−1)a = −a. This result leds to

0 = a + (−a)

So, a is an inverse of −a. Thus, a = −(−a). For a ̸= 0, by F8, we give

aa−1 = 1

Then, a is a multiplicative inverse of a−1 . So, a = (a−1 )−1 .

Theorem 1.1.5 Let a and b be real numbers. Then

−(ab) = a(−b) = (−a)b.

Proof. Let a and b be real numbers. We consider

0 = 0b ( by 1. in Theorem 1.1.4)

= (a + (−a))b ( by F7 )

= ab + (−a)b ( by F4 )

Then, (−a)b is an additive inverse of ab. So, (−a)b = −(ab).

Similary, we will show that a(−b) = −(ab).
4 CHAPTER 1. THE REAL NUMBER SYSTEM

Theorem 1.1.6 (Cancellation law) Let a, b and c be real numbers. Then

1. Cancellation law for addition if a + c = b + c, then a = b.

2. Cancellation law for multiplication if ac = bc and c ̸= 0, then a = b.

Proof. Let a, b and c be real numbers. Assume that a + c = b + c. Then,

a=a+0 ( by F5 )

= a + (c + (−c)) ( by F7 )

= (a + c) + (−c) ( by F2 )

= (b + c) + (−c) ( by assumption )

= b + (c + (−c)) ( by F2 )

=b+0 ( by F7 )

=b ( by F5 )

Next, we assume that ac = bc and c ̸= 0. Then c−1 ∈ R. We obtain

a = a1 ( by F6 )

= a(cc−1 ) ( by F8 )

= (ac)c−1 ( by F2 )

= (bc)c−1 ( by assumption )

= b(cc−1 ) ( by F2 )

= b1 ( by F8 )

=b ( by F6 )
1.1. ORDERED FIELD AXIOMS 5

Theorem 1.1.7 (Integral Domain) Let a and b be real numbers.

If ab = 0 , then a = 0 or b = 0.

Proof. Let a and b be real numbers. Suppose ab = 0 and a ̸= 0. By 1. in Theorem 1.1.4, we get

ab = 0 = a0

By cancellation for multiplication, b = 0.

ORDER AXIOMS.
There is a relation < on R2 that has the following properties for every a, b, c ∈ R.

O1 Trichotomy Property Given a, b ∈ R, one and only one of

the following statements holds:
a < b, b < a, or a = b
O2 Trasitive Property a < b and b < c imply a < c
O3 Additive Property a < b imply a + c < b + c
O4 Multiplicative Property O4.1 a < b and 0 < c imply ac < bc
O4.2 a < b and c < 0 imply bc < ac
We define in other cases:

• By b > a we shall mean a < b.

• By a ≤ b we shall mean a < b or a = b.

• If a < b and b < c, we shall write a < b < c.

• We shall call a number a ∈ R nonnegative if a ≥ 0 and positive if a > 0.

Example 1.1.8 Let x ∈ R. Show that if 0 < x < 1, then 0 < x2 < x

Proof. Let x be a real number such that 0 < x < 1. Then 0 < x and x < 1. By O4.1 and the fact
that x > 0, we obtain

0 = 0 · x < x · x = x2 and x2 = x · x < 1 · x = x

By O2, it implies that

0 < x2 < x.
6 CHAPTER 1. THE REAL NUMBER SYSTEM

Example 1.1.9 Let x, y ∈ R. Show that if 0 < x < y, then 0 < x2 < y 2

Proof. Let x and y be real numbers such that 0 < x < y. Then x > 0 and y > 0. By O4.1, we
obtain
0·x < x·x < y·x
0 < x2 < xy
and
0·y < x·y < y·y
0 < xy < y2.
Then 0 < x2 < xy and xy < y 2 . By Transitive Property, 0 < x2 < y 2 .

Theorem 1.1.10 Let a, b, c and d be real numbers.

If a < b and c < d, then a + c < b + d.

Proof. Let a, b, c and d be real numbers. Assume that a < b and c < d. By O3, we get

a + c < b + c and b + c < b + d.

By Transitive Property, a + c < b + d.

Theorem 1.1.11 Let a, b, c and d be real numbers.

If 0 < a < b and 0 < c < d, then ac < bd.

Proof. Let a, b, c and d be real numbers. Assume that 0 < a < b and 0 < c < d.
Then b > 0 and c > 0. By O4.1, we get

ac < bc and bc < bd.

By Transitive Property, ac < bd.

1.1. ORDERED FIELD AXIOMS 7

Theorem 1.1.12 If a ∈ R, prove that

a ̸= 0 implies a2 > 0.

In particular, −1 < 0 < 1.

Proof. Let a be a real number. Assume that a ̸= 0. By Trichotomy Property (O1), a > 0 or a < 0.
Case a > 0. By O4.1, a · a > 0 · a. So, a2 > 0.
Case a < 0. By O4.2, a · a > 0 · a. So, a2 > 0.
Moreover, we see that 1 ̸= 0. So, 1 = 12 > 0. By cancellation for addition,

1 + (−1) > 0 + (−1).

From F7, we obtain 0 > −1. Thus, −1 < 0 < 1.

Example 1.1.13 If x ∈ R, prove that x > 0 implies x−1 > 0.

Proof. Let x ∈ R such that x > 0. Then x−1 ̸= 0. By Theorem 1.1.12, (x−1 )2 > 0. Thus,

x−1 = x · x−2 = x · (x−1 )2 > 0 · (x−1 )2 = 0.

Example 1.1.14 If x ∈ R, prove that x < 0 implies x−1 < 0.

Proof. Let x ∈ R such that x < 0. Then x−1 ̸= 0. By Theorem 1.1.12, (x−1 )2 > 0. Thus,

x−1 = x · x−2 = x · (x−1 )2 < 0 · (x−1 )2 = 0.

8 CHAPTER 1. THE REAL NUMBER SYSTEM

Theorem 1.1.15 Let a and b be real numbers such that 0 < a < b. Then

1 1
< .
b a

1
Proof. Let a and b be real numbers such that 0 < a < b. Then ab > 0. So, > 0.
ab
By O4.1, we obtain
1 1 1
0· < a· < b·
ab ab ab
1 1
0 < < .
b a

Example 1.1.16 Let x and y be two distinct real numbers. Prove that
x+y
lies between x and y.
2
Proof. Let x and y be two distinct real numbers.
By Trinochomy rule, x ̸= y. WLOG x < y. Then x + x < x + y and x + y < y + y.
By transitive rule,

2x < x + y < 2y
x+y
x< < y.
2
1.1. ORDERED FIELD AXIOMS 9

ABSOLUTE VALUE.

Definition 1.1.17 (Absolute Value) The absolute value of a number a ∈ R is a the number




 a if a > 0


|a| = 0 if a = 0





−a if a < 0

Theorem 1.1.18 (Positive Definite) For all a ∈ R,

1. |a| ≥ 0 2. |a| = 0 if and only if a = 0

Proof. Let a be a real number.

1. Case a = 0. Then |a| = |0| = 0 ≥ 0.

Case a > 0. Then |a| = a > 0.

Case a < 0. Then |a| = −a = (−1)a > (−1)0 = 0.

Hence, |a| ≥ 0.

2. It’s obviously by definition.

Theorem 1.1.19 (Multiplicative Law) For all a, b ∈ R,

|ab| = |a||b|.

Proof. Let a and b be real numbers.

Case a = 0 or b = 0. Then ab = 0 and |a| = 0 or |b| = 0. So, |ab| = |0| = 0 = |a||b|.

Case a > 0 and b > 0. Then ab > 0, |a| = a and |b| = b. So, |ab| = ab = |a||b|.

Case a > 0 and b < 0. Then ab < 0, |a| = a and |b| = −b. So, |ab| = −ab = a(−b) = |a||b|.

Case a < 0 and b > 0. Then ab < 0, |a| = −a and |b| = b. So, |ab| = −ab = (−a)b = |a||b|.

Case a < 0 and b < 0. Then ab > 0, |a| = −a and |b| = −b. So, |ab| = ab = (−a)(−b) = |a||b|.

Hence, |ab| = |a||b|.

10 CHAPTER 1. THE REAL NUMBER SYSTEM

Theorem 1.1.20 (Symmetric Law) For all a, b ∈ R,

|a − b| = |b − a|.

Moreover, |a| = | − a|.

Proof. Let a and b be real numbers. By Multiplicative Law, it implies that

|a − b| = | − (−a) + (−b)| = |(−1)(−a) + (−1)b| = |(−1)(−a + b)|

= | − 1|| − a + b| = 1 · | − a + b| = | − a + b| = |b − a|.

For b = 0, we obtain |a| = | − a|.

1 1
Example 1.1.21 Show that = for all x ̸= 0.
x |x|
Proof. Let x be a non-zero real number.
1 1 1 1
Case x > 0. Then |x| = x and > 0. So, = = .
x x x |x|
1 1 1 1 1
Case x < 0. Then |x| = −x and < 0. So, =− = = .
x x x −x |x|

Theorem 1.1.22 Let a, b ∈ R. Show that

1. |a2 | = a2 2. a ≤ |a| a |a|

3. = when b ̸= 0
b |b|

Proof. Let a, b ∈ R. By Theorem 1.1.12, a2 ≥ 0. So, |a2 | = a2 .

Case a = 0. Then a = 0 ≤ 0 = |0| = |a|.

Case a > 0. Then a ≤ a = |a|.

Case a < 0. Then −a > 0. So, a < 0 < −a = |a|.

Thus, a ≤ |a|. Use Multiplicative law and Example 1.1.21 to 3, we have

a 1 1 |a|
= |ab−1 | = |a||b−1 | = |a| · = |a| · =
b b |b| |b|
1.1. ORDERED FIELD AXIOMS 11

Theorem 1.1.23 Let a ∈ R and M ≥ 0. Then

|a| ≤ M if and only if −M ≤ a ≤ M

Proof. Let a ∈ R and M ≥ 0.

Assume that |a| ≤ M . By definition, |a| = ±a. Then

a≤M and −a ≤ M .

We obtain a ≥ −M . Thus, −M ≤ a ≤ M .
Conversely, assume that −M ≤ a ≤ M . Then

−M ≤ a and a ≤ M.

So, M ≥ −a. Thus, |a| = ±a ≤ M.

Corollary 1.1.24 For all a ∈ R, −|a| ≤ a ≤ |a|.

Proof. Choose M = |a| ≥ 0 in Theorem 1.1.23, we obtain this Corollary.

INTERVAL.
Let a and b real numbers. A closed interval is a set of the form
[a, b] := {x ∈ R : a ≤ x ≤ b} (−∞, b] := {x ∈ R : x ≤ b}

[a, ∞) := {x ∈ R : a ≤ x} (−∞, ∞) := R,

and an open interval is a set of the form

(a, b) := {x ∈ R : a < x < b} (−∞, b) := {x ∈ R : x < b}

(a, ∞) := {x ∈ R : a < x} (−∞, ∞) := R.

By an interval we mean a closed interval, an open interval, or a set of the form

[a, b) := {x ∈ R : a ≤ x < b} or (a, b] := {x ∈ R : a < x ≤ b}

Notice, then, that when a < b, then intervals [a, b], [a, b), (a, b] and (a, b) correspond to line
segments on the real line, but when b < a, these interval are all the empty set.
12 CHAPTER 1. THE REAL NUMBER SYSTEM

Example 1.1.25 Solve |x − 1| ≤ 1 for x ∈ R in interval form.

Solution. By Theorem 1.1.23, −1 < x − 1 < 1. So,

0 < x < 2.

Thus, x ∈ (0, 2).

Example 1.1.26 Show that if |x| < 1, then |x2 + x| < 2.

Solution. Let |x| < 1. Then −1 < x < 1. So, 0 < x + 1 < 2. We obtain

−2 < 0 < x + 1 < 1 −→ |x + 1| < 2.

Therefore,

|x2 + x| = |x(x + 1)| = |x||x + 1| < 1 · 2 = 2.

1
Example 1.1.27 Show that if |x − 1| < 2, then > 1.
|x|
Solution. Let |x − 2| < 1. Then −1 < x − 2 < 1. So, 1 < x < 3. We obtain

|x| > 1.

1
Therefore, > 1.
|x|

Theorem 1.1.28 (Triangle Inequality) Let a, b ∈ R. Then,

|a + b| ≤ |a| + |b|.

Proof. Let a, b ∈ R. By Corollary 1.1.24,

−|a| ≤ a ≤ |a|
−|b| ≤ b ≤ |b|

Then, −(|a| + |b|) ≤ a + b ≤ |a| + |b|. Therefore, |a + b| ≤ |a| + |b|.

1.1. ORDERED FIELD AXIOMS 13

Theorem 1.1.29 (Apply Triangle Inequality) Let a, b ∈ R. Then,

1. |a − b| ≤ |a| + |b| 3. |a| − |b| ≤ |a + b|

2. |a| − |b| ≤ |a − b| 4. ||a| − |b|| ≤ |a − b|

Proof. Let a, b ∈ R.

1. By Triangle Inequality,

|a − b| = |a + (−b)| ≤ |a| + | − b| = |a| + |b|.

2. By Triangle Inequality,
|a| = |(a − b) + b| ≤ |a − b| + |b|.

Thus, |a| − |b| ≤ |a − b|.

3. By 2,
|a| − |b| = |a| − | − b| ≤ |a − (−b)| = |a + b|.

4. By 2, |a| ≤ |a − b| + |b|. By 3,

|b| − |a − b| ≤ |b + (a − b)| = |a|.

Then,
|b| − |a − b| ≤ |a| ≤ |a − b| + |b|
−|a − b| ≤ |a| − |b| ≤ |a − b|
Thus, ||a| − |b|| ≤ |a − b|.

Example 1.1.30 Show that if |x − 2| < 1, then |x| < 3.

Solution. Let |x − 2| < 1. By 3 in Theorem 1.1.29,

|x| − 2 = |x| − |2| < |x − 2| < 1.

Therefore, |x| < 1 + 2 = 3.

14 CHAPTER 1. THE REAL NUMBER SYSTEM

Theorem 1.1.31 Let x, y ∈ R. Then

1. x < y + ε for all ε > 0 if and only if x ≤ y

2. x > y − ε for all ε > 0 if and only if x ≥ y

Proof. Let x, y ∈ R.

1. Assume that x < y + ε for all ε > 0 and x > y. Then x − y > 0. By assumption, we get

x < y + (x − y) = x.

It is imposible. So,x < y + ε for all ε > 0 if and only if x ≤ y Conversely, suppose that
there is an ε > 0 such that x ≥ y + ε. So,

x≥y+ε>y+0=y

Thus, x > y. We conclude that if x ≥ y, then x < y + ε for all ε > 0.

2. Excercise.

Corollary 1.1.32 Let a ∈ R. Then

|a| < ε for all ε > 0 if and only if a = 0

Proof. Use Theorem 1.1.31 by x = |a| and y = 0. Thus,

|a| < 0 + ε for all ε > 0 if and only if |a| ≤ 0.

Since |a| ≥ 0, |a| = 0. By positive definite, a = 0. The proof is complete.

1.1. ORDERED FIELD AXIOMS 15

Exercises 1.1

1. Let a, b ∈ R. Prove that

1.1 −(a − b) = b − a 1.3 (−a)(−b) = ab

−a a a
1.2 a(b − c) = ab − ac 1.4 = = − when b ̸= 0
b −b b

2. Let a, b ∈ R. Prove that

2.1 If a + b = a, then x = 0.

2.2 If ab = b and b ̸= 0, then a = 1.

2.3 If a−1 = a and a ̸= 0, then a = −1 or a = 1.

3. Let a, b, c, d ∈ R. Prove that

3.1 if a < b < 0, then 0 < b2 < a2 .

1 1
3.2 if a < b < 0, then < .
b a
3.3 if a ≤ b and a ≥ b, then a = b.
√ √
3.4 if 0 < a < b, then a < b.

4. Solve each of the following inequality for x ∈ R.

4.1 |1 − 2x| ≤ 3 4.3 |x2 − x − 1| < x2

4.2 |3 − x| < 5 4.4 |x2 − x| < 2

√
5. Prove that if 0 < a < 1 and b = 1 − 1 − a, then 0 < b < a.
√
6. Prove that if a > 2 and b = 1 − 1 − a, then 2 < b < a.

7. Prove that |x| ≤ 1 implies |x2 − 1| ≤ 2|x − 1|.

8. Prove that −1 ≤ x ≤ 2 implies |x2 + x − 2| ≤ 4|x − 1|.

9. Prove that |x| ≤ 1 implies |x2 − x − 2| ≤ 3|x + 1|.

10. Prove that 0 < |x − 1| ≤ 1 implies |x3 + x − 2| < 8|x − 1|. Is this true if 0 ≤ |x − 1| < 1 ?
16 CHAPTER 1. THE REAL NUMBER SYSTEM

11. Let x, y ∈ R. Prove that if |x + y| = |x − y|, then x|y| + y|x| = 0.

12. Let x, y ∈ R. Prove that if |2x + y| = |x + 2y|, then |xy| = x2 .

a2 + 2
13. Let a ∈ R. Prove that √ ≥ 2.
a2 + 1
14. Prove that
(a1 b1 + a2 b2 )2 ≤ (a21 + a22 )(b21 + b22 )

for all a1 , a2 , b1 , b2 ∈ R

15. Let x, y ∈ R. Prove that x > y − ε for all ε > 0 if and only if x ≥ y.

16. Suppose that x, a, y, b ∈ R, |x − a| < ε and |y − b| < ε for some ε > 0. Prove that

16.1 |xy − ab| < (|a| + |b|)ε + ε2

16.2 |x2 y − a2 b| < ε(|a|2 + 2|ab|) + ε2 (|b| + 2|a|) + ε2

17. The positive part of an a ∈ R is defined by

|a| + a
a+ :=
2

and the negative part by

|a| − a
a− := .
2
17.1 Prove that a = a+ − a− and |a| = a+ + a− .
 

a : a ≥ 0 
0 : a≥0
+ −
17.2 Prove that a := and a := .

0 : a ≤ 0 
−a : a ≤ 0

a+b
18. Let a, b ∈ R. The arithmetic mean of a, b is A(a, b) := ,
√ 2
the geometric mean of a, b ∈ (0, ∞) is G(a, b) := ab,
2
and harmonic mean of a, b ∈ (0, ∞) is H(a, b) := −1 .
a + b−1
Show that

18.1 if a, b ∈ (0, ∞). Then H(a, b) ≤ G(a, b) ≤ A(a, b).

18.2 if 0 < a ≤ b. Then a ≤ G(a, b) ≤ A(a, b) ≤ b.

18.3 if 0 < a ≤ b. Then, G(a, b) = A(a, b) if and only if a = b.

1.2. WELL-ORDERING PRINCIPLE 17

1.2 Well-Ordering Principle

Definition 1.2.1 A number m is a least element of a set S ⊂ R if and only if

m ∈ S and m ≤ s for all s ∈ S.

WELL-ORDERING PRINCIPLE (WOP).

Every nonempty subset of N has a least element.

S ⊆ N ∧ S ̸= ∅ → ∃m ∈ S ∀s ∈ S, m ≤ s.

Theorem 1.2.2 (Mathematical Induction) Suppose for each n ∈ N that P (n) is a statement
that satisfies the following two properties:

(1) Basic step : P (1) is true

(2) Inductive step : For every k ∈ N for which P (k) is true, P (k + 1) is also true.

Then P (n) is true for all n ∈ N.

Proof. We will prove by contradiction. Assume that (1) and (2) are ture and there is an n0 ∈ N
such that P (n0 ) is false. Define

S = {n ∈ N : P (n) is false }.

Then, n0 ∈ S ⊆ N. By WOP, S has a least element, said m ∈ S.

Since (1) is true, m ̸= 1. Then m > 1 or m − 1 > 0. So, m − 1 ∈ N.
But m − 1 < m and m is the least element in S, so m − 1 ∈
/ S. Set

k = m − 1 ∈ N. We obtain P (k) is true.

By (2), P (k + 1) = P (m) is true. This contradicts m ∈ S.

18 CHAPTER 1. THE REAL NUMBER SYSTEM

Example 1.2.3 (Gauss’ formula) Prove that

∑
n
n(n + 1)
k=
k=1
2

for all n ∈ N.

∑
1
2 1(1 + 1)
Proof. For n = 1, we get k=1= = . So, (1) is true.
k=1
2 2
∑
n
n(n + 1)
Assume that k= . Then,
k=1
2

∑
n+1 ∑
n
n(n + 1) [n ] (n + 1)(n + 2)
k= k + (n + 1) = + (n + 1) = (n + 1) +1 = .
k=1 k=1
2 2 2

So, (2) is true. By Mathematical Induction, Gauss’ formula is proved.

Example 1.2.4 Prove that 2n > n for all n ∈ N.

Proof. We will prove by induction on n. For n = 1, it is clear 21 > 1. Assume that 2n > n for
some n ∈ N. By inductive hypothesis and the fact that n ≥ 1,

2n+1 = 2n · 2 > 2n = n + n ≥ n + 1.

So, 2n > n is true for n + 1. We conclude by induction that 2n > n holds for n ∈ N.

BINOMIAL FORMULA.

Definition 1.2.5 The notation 0! = 1 and n! = 1 · 2 · · · (n − 1) · n for n ∈ N (called factorial),

define the binomial coefficient n over k by
 
n n!
  :=
k (n − k)!k!

for 0 ≤ k ≤ n and n = 0, 1, 2, 3, ...

Theorem 1.2.6 If n, k ∈ N and 1 ≤ k ≤ n, then

     
n+1 n n
 = + 
k k−1 k
1.2. WELL-ORDERING PRINCIPLE 19

Proof. Let n, k ∈ N and 1 ≤ k ≤ n. We obtain

   
n n n! n!
 + = +
k−1 k (n − k + 1)!(k − 1)! (n − k)!k!
n!k n!(n − k + 1)
= +
(n − k + 1)!(k − 1)!k (n − k + 1)(n − k)!k!
n!k n!(n − k + 1) n!k + n!(n − k + 1)
= + =
(n − k + 1)!k! (n − k + 1)!k! (n − k + 1)!k!
 
n![k + (n − k + 1)] n!(n + 1) (n + 1)! n + 1
= = = = .
(n − k + 1)!k! (n − k + 1)!k! (n − k + 1)!k! k

Theorem 1.2.7 (Binomial formula) If a, b ∈ R and n ∈ N, then

 
∑
n
n
(a + b)n =   an−k bk
k=0 k

Proof. We will prove by induction on n. The formula is obvious for n = 1. Assume that the
formula is true for some n ∈ N. By inductive hypothesis,
 
∑ n
n
(a + b)n+1 = (a + b)(a + b)n = (a + b)   an−k bk
k=0 k
   
∑n
n ∑n
n
=   an−k+1 bk +   an−k bk+1
k=0 k k=0 k
   
∑n
n ∑
n−1
n
= an+1 +   an−k+1 bk +   an−k bk+1 + bn+1
k=1 k k=0 k
   
∑n
n ∑n
n
= an+1 +   an−k+1 bk +   an−k+1 bk + bn+1
k=1 k k=1 k−1
   
∑n
n n
=a n+1
+   +   an−k+1 bk + bn+1
k=1 k k−1
Thus, it follows from Theorem 1.2.6 that
   
∑n
n+1 ∑
n+1
n+1
(a + b)n+1 = an+1 +   an−k−1 bk + bn+1 =   an+1−k bk
k=1 k k=0 k
i.e., the formula is true for n + 1. We conclude by induction that the formula holds for n ∈ N.
20 CHAPTER 1. THE REAL NUMBER SYSTEM

Exercises 1.2

1. Prove that the following formulas hold for all n ∈ N.

∑
n ∑
n
n(4n2 − 1)
1.1 (3k − 1)(3k + 2) = 3n + 6n + n
3 2
1.3 (2k − 1)2 =
k=1 k=1
3
∑n [ ]2 ∑n
n(n + 1) a−1 1
1.2 k3 = 1.4 = 1 − , a ̸= 0
k=1
2 k=1
ak an

2. Use the Binomial Formula to prove each of the following.

 
∑n
n
2.1 2n =   for all n ∈ N.
k=1 k
2.2 (a + b)n ≥ an + aan−1 b for all n ∈ N and a, b ≥ 0.
( )n
1
2.3 1 + ≥ 2 for all n ∈ N.
n
3. Let n ∈ N. Write
(x + h)n − xn
h
as a sum none of whose terms has an h in the dennominator.
√
4. Suppose that 0 < x1 < 1 and xn+1 = 1 − 1 − xn for n ∈ N. Prove that 0 < xn+1 < xn < 1
holds for all n ∈ N.
√
5. Suppose that x1 ≥ 2 and xn+1 = 1 + xn − 1 for n ∈ N. Prove that 2 ≤ xn+1 ≤ xn ≤ x1
holds for all n ∈ N.
√
6. Suppose that 0 < x1 < 2 and xn+1 = 2 + xn for n ∈ N. Prove that 0 < xn < xn+1 < 2
holds for all n ∈ N.

7. Prove that each of the following inequalities hold for all n ∈ N.

7.1 n < 3n 7.2 n2 ≤ 2n + 1 7.3 n3 ≤ 3n

8. Let 0 < |a| < 1. Prove that |a|n+1 < |a|n for all n ∈ N.

9. Prove that 0 ≤ a < b implies an < bn for all n ∈ N.

1.3. COMPLETENESS AXIOM 21

1.3 Completeness Axiom

SUPREMUM.

Definition 1.3.1 Let A be a nonempty subset of R.

1. The set A is said to be bounded above if and only if

there is an M ∈ R such that a ≤ M for all a ∈ A

2. A number M is called an upper bound of the set A if and only if

a ≤ M for all a ∈ A

3. A number s is called a supremum of the set A if and only if

s is an upper bound of A and s ≤ M for all upper bound M of A

In this case we shall say that A has a supremum s and shall write s = sup A

Example 1.3.2 Fill the blanks of the following table.

Sets Bounded above Set of Upper bound Supremum

A = [0, 1] Yes [1, ∞) 1

A = (0, 1) Yes [1, ∞) 1

A = {1} Yes [1, ∞) 1

A = (0, ∞) No ∅ None

A = (−∞, 0) Yes [0, ∞) 0

A=N No ∅ None

A=Z No ∅ None
22 CHAPTER 1. THE REAL NUMBER SYSTEM

Example 1.3.3 Show that sup A = 1 where

1. A = [0, 1] 2. A = (0, 1)

Solution.

1. For A = [0, 1]. Since a ≤ 1 for all a ∈ A, 1 is an upper bound of A.

Let M be an upper bound of A. Then,

a≤M for all a ∈ A

Since 1 ∈ A, 1 ≤ M . Thus, sup A = 1.

2. For A = (0, 1). Since a < 1 ≤ 1 for all a ∈ A, 1 is an upper bound of A.

Suppose that there is an upper bound M0 of A such that M0 < 1. Then,

a < M0 for all a ∈ A

M0 + 1 M0 + 1
But 0 < a < M0 < < 1, so belongs to A. It is imposible because M0 is
2 2
an upper bound of A. Hence, there is no upper bound of A such that it is less that 1. We
conclude that sup A = 1.

Theorem 1.3.4 If a set has one upper bound, then it has infinitely many upper bounds.

Proof. Let M0 be an upper bound of a set A. We set

M := M0 + k for all k ∈ N.

Then, M > M0 for all k ∈ N. So, M is another upper bound of A depending on k.

This reason shows that it has infinitely many upper of A.
1.3. COMPLETENESS AXIOM 23

Theorem 1.3.5 If a set has a supremum, then it has only one supremum.

Proof. Let s1 and s2 be suprema of the same of a set A. Then, s1 and s2 are upper bounds of A.
By definition of supremum, we obtain

s1 ≤ s2 and s2 ≤ s1 .

Therefore, s1 = s2 .

Theorem 1.3.6 (Approximation Property for Supremum (APS)) If A has a supremum

and ε > 0 is any positive number, then there is a point a ∈ A such that

sup A − ε < a ≤ sup A

Proof. We will prove by contradiction. Assume that A has an infimum, say s. Suppose that there
a positve ε0 > 0 such that

a ≤ s − ε0 or a > s for all a ∈ A

In this case a > s, it is imposible beacause s is an upper bound of A.

From a ≤ s − ε0 for all a ∈ A, it means that s − ε0 is an upper bound of A. But

s − ε0 < s

It’s imposible because s is the least upper bound of A.

Theorem 1.3.7 If A ⊂ N has a supremum, then sup A ∈ A.

Proof. Assume that A ⊂ N has a supremum, say s. Apply APS to choose an x0 ∈ A such that

s − 1 < x0 ≤ s.

If x0 = s, then s ∈ A. In this case s − 1 < x0 < s. Apply again APS to choose x1 ∈ A such that

x0 < x1 < s
0 < x1 − x0 < s − x0 .
24 CHAPTER 1. THE REAL NUMBER SYSTEM

Since x0 , x1 ∈ N and x0 ̸= x1 , x1 − x0 ≥ 1. From s − 1 < x0 and x1 < s, we get

(s − 1) + x1 < x0 + s.

So, x1 − x0 < 1. It contradicts to x1 − x0 ≥ 1. Thus, this case is false.

COMPLETENESS AXIOM.
If A is a nonempty subset of R that is bounded above, then A has a supremum.

Theorem 1.3.8 The set of natural numbers is not bounded above.

Proof. Suppose that N is bounded above. Since N is not a nonempty set by Completeness Axiom,
N has a supremum, say s. Then

n≤s for all n ∈ N.

If n ∈ N, then n + 1 ∈ N. So, n + 1 ≤ s for all n ∈ N, i.e.,

n≤s−1 for all n ∈ N.

Thus, s − 1 is an upper bound of N. We obatain s ≤ s − 1 or 0 < −1. It is imposible.

Theorem 1.3.9 (Archimedean Properties (AP)) For each x ∈ R, the following statements
are true.

1. There is an integer n ∈ N such that x < n.

1
2. If x > 0, there there is an integer n ∈ N such that < x.
n

Proof. Suppose that there is an x ∈ R such that x ≥ n for all n ∈ N. It means that x is an upper
bound of N. This is contradiction Theorem 1.3.8. Thus, part 1 is proved.
1 1
Next, we assume that x > 0. Then ∈ R. By 1, there is an n ∈ N such that < n. Thus,
x x
1
< x.
n

The proof of Archimedean Properties is complete.

1.3. COMPLETENESS AXIOM 25

Theorem 1.3.10 Let x ∈ R. Then

1
|x| < for all n ∈ N if and only if x = 0
n

1
Proof. Let x ∈ R. Assume that |x| < for all n ∈ N. Let ε > 0. By AP, there an N ∈ N such
n
1
that < ε. By assumption, we obtain
N
1
|x| < < ε.
N

From Corollary 1.1.32 , it implies that x = 0. Conversely, it is obvious.

{ }
1
Example 1.3.11 Let A = : n ∈ N . Prove that sup A = 1.
n
1
Proof. For each n ∈ N, we get n ≥ 1. So, ≤ 1. Thus, 1 is an upper bound of A.
n
Let M be any upper bound of A. Then

a≤M for all a ∈ A.

For n = 1, we have 1 =∈ A. So, 1 ≤ M . Hence, sup A = 1.

1
1
{ }
n
Example 1.3.12 Let A = : n ∈ N . Prove that sup A = 1.
n+1
n
Proof. Since 0 < n < n + 1 for all n ∈ N, <1 for all n ∈ N.
n+1
Thus, 1 is an upper bound of A.
Suppose that that there is an upper bound u0 of A such that u0 < 1.
Since u0 < 1, 1 − u0 > 0. By AP, there is n0 ∈ N such that

1
< 1 − u0 .
n0
1 1
Since n0 + 1 > n0 > 0, < . We obtain
n0 + 1 n0
1
< 1 − u0
n0 + 1
1 n0
u0 < 1 − =
n0 + 1 n0 + 1

So, u0 is not upper bound of A. This is contradiction. Therefore, sup A = 1.

26 CHAPTER 1. THE REAL NUMBER SYSTEM

Theorem 1.3.13 If x ∈ R, then there is an n ∈ Z such that

n − 1 ≤ x < n.

Proof. Let x ∈ R. If x = 0, we choose n = 1. We are done.

Case 1. x > 0. Define S = {n ∈ N : n > x} ⊆ N. By AP, S ̸= ∅. From WOP, S has the least
element, say n0 . Since n0 − 1 < n0 , n0 − 1 ∈
/ A. So, n0 − 1 ≤ x. Thus,

n0 − 1 ≤ x < n 0 .

The proof is complete in this case.

Case 2. x < 0. Then −x > 0. By Case 1, there is an m ∈ N such that m − 1 ≤ −x < m. Then

−m < x ≤ −m + 1.

If x = −m + 1, we choose n = −m + 2. So,

n − 1 = −m + 1 = x < n or n − 1 ≤ x < n.

If −m < x < −m+1, we choose n = −m+1. So, n−1 < x < n. It implies that n−1 ≤ x < n.

Theorem 1.3.14 (Density of Rationals) If a, b ∈ R satisfy a < b, then there is a rational

number r such that
a < r < b.

Proof. Let a, b ∈ R such that a < b. Then b − a > 0.

1
By AP, there is an N ∈ N such that < b − a. It follows that
n

na + 1 < nb.

By Theorem 1.3.13, there is an m ∈ Z such that m − 1 ≤ na < m. It implies that

na < m ≤ na + 1 < nb.

m
Set r := . We obtain a < r < b.
n
1.3. COMPLETENESS AXIOM 27

√
Theorem 1.3.15 2 is irrational.

√
Proof. Assume that 2 is a rational number. Then there are two integers p and q such that
√ p
2 = when q ̸= 0 and gcd(p, q) = 1.
q
We have 2q 2 = p2 . It implies that p is an even number. Then there is an k ∈ Z such that p = 2k.
So,

2q 2 = (2k)2 = 4k 2

q 2 = 2k 2

It implies again that q is an even number. Thus, gcd(p, q) ̸= 1. This is contradiction.

Theorem 1.3.16 (Density of Irrationals) If a, b ∈ R satisfy a < b, then there is an irrational

number t such that
a < t < b.

a b
Proof. Let a, b ∈ R such that a < b. Then √ < √ . By the Density of Rational, there is an
2 2
a b
r ∈ Q such that √ < r < √ . It follows that
2 2
√
a < r 2 < b.
√
If r ̸= 0, then t := r 2 is irrational (see Exercise). It is done.
a b
Case r = 0. By the Density of Rational, there is an s ∈ Q such that √ < 0 < s < √ . It follows
2 2
that
√
a < s 2 < b.
√
Set t = s 2, irrational. Thus, the proof is complete.
28 CHAPTER 1. THE REAL NUMBER SYSTEM

INFIMUM.

Definition 1.3.17 Let A be a nonempty subset of R.

1. The set A is said to be bounded below if and only if

there is an m ∈ R such that m ≤ a for all a ∈ A

2. A number m is called a lower bound of the set A if and only if

m≤a for all a ∈ A

3. A number ℓ is called an infimum of the set A if and only if

ℓ is a lower bound of A and m ≤ ℓ for all lower bound m of A

In this case we shall say that A has an infimum s and shall write ℓ = inf A

4. A is said to be bounded if and only if it is bounded above and below.

Example 1.3.18 Fill the blanks of the following table.

Sets Bounded below Set of Lower bound Infimum Bounded

A = [0, 1] Yes (−∞, 0] 0 Yes

A = (0, 1) Yes (−∞, 0] 0 Yes

A = {1} Yes (−∞, 1] 1 Yes

A = (0, ∞) Yes (−∞, 0] 0 No

A = (−∞, 0) No ∅ None No

A=N Yes (−∞, 1] 1 No

A=Z No ∅ None No
1.3. COMPLETENESS AXIOM 29

Example 1.3.19 Show that inf A = 0 where

1. A = [0, 1] 2. A = (0, 1)

Solution.

1. For A = [0, 1]. Since a ≥ 0 for all a ∈ A, 1 is a lower bound of A.

Let m be a lower bound of A. Then,

m≤a for all a ∈ A

Since 0 ∈ A, 0 ≤ M . Thus, inf A = 0.

2. For A = (0, 1). Since a > 0 ≥ 0 for all a ∈ A, 0 is a lower bound of A.

Suppose that there is a lower bound m0 of A such that m0 > 0. Then,

m0 ≤ a for all a ∈ A

m0 m0
But 0 < < m0 ≤ a, so belongs to A. It is imposible because m0 is a lower bound
2 2
of A. Hence, there is no lower bound of A such that it is greater that 0. We conclude that
inf A = 0.
{ }
1
Example 1.3.20 Let A = : n ∈ N . Prove that inf A = 0.
n
1
Proof. For each n ∈ N, we get n > 0. So, > 0. Thus, 0 is a lower bound of A.
n
Suppose that that there is a lower bound m0 of A such that m0 > 0.
By AP, there is n0 ∈ N such that

1
< m0 .
n0

So, m0 is not lower bound of A. This is contradiction. Therefore, inf A = 0.

30 CHAPTER 1. THE REAL NUMBER SYSTEM
{ }
n 1
Example 1.3.21 Let A = : n ∈ N . Prove that inf A = .
n+1 2
1 1
Proof. Let n ∈ N. Then n ≥ 1. So, ≤ 1 or 1 + ≤ 2. We obtain
n n
1 1 n
≤ 1 = .
2 1+ n
n+1
1
Thus, 2
is a lower bound of A.
Let m0 be any lower bound of A. Then

m0 ≤ a for all a ∈ A.
1 1
For n = 1, we have that = belongs to A.
2 1+1
1
m0 ≤ .
2
1
Therefore, inf A = .
2

Theorem 1.3.22 (Approximation Property for Infimum (API)) If A has an infimum and
ε > 0 is any positive number, then there is a point a ∈ A such that

inf A ≤ a < inf A + ε.

Proof. Assume that A has an infimum, say ℓ0 . Suppose that there a positve ε0 > 0 such that

a < ℓ0 or a ≥ ℓ0 + ε0 for all a ∈ A

In this case a < ℓ0 , it is imposible beacause ℓ0 is a lower bound of A.

From a ≥ ℓ0 + ε0 for all a ∈ A, it means that ℓ0 + ε0 is a lower bound of A. But

ℓ0 + ε0 > ℓ0

It’s imposible because ℓ0 is the greatest lower bound of A.

1.3. COMPLETENESS AXIOM 31

Exercises 1.3

1. Find the infimum and supremum of each the following sets.

{ }
1.1 A = [0, 2) (−1)n
1.7 A= 1+ :n∈N
n
1.2 A = {4, 3, 1, 5} { }
n+1
1.3 A = {x ∈ R : |x − 1| < 2} 1.8 A= :n∈N
n
{ 2 }
1.4 A = {x ∈ R : |x + 1| < 1} n +n
1.9 A= :n∈N
n2 + 1
1.5 A = {1 + (−1)n : n ∈ N}
{ } { }
1 n(−1)n + 1
1.6 A = − (−1) : n ∈ N
n
1.10 A= :n∈N
n n+2

2. Find inf A and sup A with proving them.

{ }
2.1 A = [−1, 1] n
2.5 A = :n∈N
n+2
{ }
2.2 A = (−1, 2] n−2
2.6 A = :n∈N
n+2
{ }
2.3 A = (−1, 0) ∪ (1, 2) n
2.7 A = :n∈N
n2 + 1
2.4 A = {1, 2, 3} 2.8 A = {(−1)n : n ∈ N}
{ }
n
3. Let A = 1 − : n ∈ N . What are supremum and infimum of A ? Verify (proof)
n2 + 2
your answers.
{ }
n
4. Let A = 2 − : n ∈ N . What are supremum and infimum of A ? Verify (proof)
n2 + 1
your answers.

5. If a set has one lower bound, then it has infinitely many lower bounds.

6. Prove that if A is a nonempty bounded subset of Z, then both sup A and inf A exist and
belong to A.

7. Prove that for each a ∈ R and each n ∈ N there exists a rational rn such that

1
|a − rn | < .
n
32 CHAPTER 1. THE REAL NUMBER SYSTEM

8. Let r be a rational number and s be an irrational number. Prove that

8.1 r + s is an irrational number.

8.2 if r ̸= 0, then rs is always an irrational number.

√
9. Let K ∈ Qc and a, b, x, y ∈ Z. Prove that

√ √
if a + b K = x + y K, then a = x and b = y.

10. Show that a lower bound of a set need not be unique but the infimum of a given set A is
unique.

11. Show that if A is a nonoempty subset of R that is bounded below, then A has a finite
infimum.

12. Prove that if x is an upper bound of a set A ⊆ R and x ∈ A, then x is the supremum of A.

13. Suppose E, A, B ⊂ R and E = A ∪ B. Prove that if E has a supremum and both A and B
are nonempty, then SupA and sup B both exist, and sup E is one of the numbers SupA or
sup B.

14. (Monotone Property) Suppose that A ⊆ B are nonempty subsets of R. Prove that

14.1 if B has a supremum, then sup A ≤ sup B

14.2 if B has an infimum, then inf B ≤ inf A

15. Define the reflection of a set A ⊆ R by

−A := {−x : x ∈ A}

Let A ⊆ R be nonempty. Prove that

15.1 A has a supremum if and only if −A has and infimum, in which case

inf(−A) = − sup A.

15.2 A has an infimum if and only if −A has and supremum, in which case

sup(−A) = − inf A.
1.4. FUNCTIONS AND INVERSE FUNCTIONS 33

1.4 Functions and Inverse functions

Review notation f : X → Y that means a fuction form X to Y , each x ∈ X is assigned a

unique y = f (x) ∈ Y , there is nothing that keeps two x’s from being assigned to the same y, and
nothing that say every y ∈ Y corresponds to some x ∈ X, i.e., f is a fuction if and only if for each
(x1 , y1 ), (x2 , y2 ) belong to f ,

if x1 = x2 , then y2 = y2 .

Definition 1.4.1 Let f be a function from a set X into a set Y .

1. f is said to be one-to-one (1-1) on X if and only if

x1 , x2 ∈ X and f (x1 ) = f (x2 ) imply x1 = x2 .

2. f is said to take X onto Y if and only if

for each y ∈ Y there is an x ∈ X such that y = f (x).

Example 1.4.2 Show that f (x) = 2x + 1 is 1-1 from R onto R.

Solution. Let x1 and x2 be reals such that f (x1 ) = f (x2 ). Then,

2x1 + 1 = 2x2 + 1

2x1 = 2x2

x1 = x2

y−1
So, f is 1-1. Let y ∈ R. Choose x = ∈ R. Then,
2
( )
y−1
f (x) = 2x + 1 = 2 +1=y
2

Thus, f takes R onto R.

34 CHAPTER 1. THE REAL NUMBER SYSTEM

Theorem 1.4.3 Let X and Y be sets and f : X → Y . Then f is 1-1 from X onto Y if and only
if there is a unique function g from Y onto X that satisfies

1. f (g(y)) = y, y∈Y

and

2. g(f (x)) = x, x∈X

Proof. Suppose that f is 1-1 and onto. For each y ∈ Y choose the unique x ∈ X such that
f (x) = y, and define
g(y) := x.

It is clear that g take Y onto X. By construction, 1 and 2 are satisfied.

Conversely, suppose that there a function g from Y onto X that satisfies 1 and 2.
Let x1 , x2 ∈ X and f (x1 ) = f (x2 ). Then it follows from 2 that

x1 = g(f (x1 )) = g(f (x2 )) = x2 .

Thus f is 1-1 on X. Let y ∈ Y and choose x = g(y). Then 1 implies that

f (x) = f (g(y)) = y.

Thus f takes X onto Y .

Finally, suppose that h is another function that satisfies 1 and 2, and y ∈ Y . Choose x ∈ X
such that f (x) = y. Then, by 2,

h(y) = h(f (x)) = x = g(f (x)) = g(y);

i.e., h = g on Y . It follows that the function is unique.

1.4. FUNCTIONS AND INVERSE FUNCTIONS 35

If f is 1-1 from a set X onto a set Y , we shall say that f has an inverse function. We shall
call the function g given in Theorem 1.4.3 the inverse of f , and denote it by f −1 . Then

f (f −1 (y)) = y and f −1 (f (x)) = x.

Example 1.4.4 Find inverse function of f (x) = 2x + 1.

Solution. By Example 1.4.2, f is 1-1 from R onto R. Then,

f −1 (2x + 1) = f −1 (f (x)) = x
x−1
Substitue x := . We obtain
2
( )
−1 −1 x−1 x−1
f (x) = f 2· +1 = .
2 2
Example 1.4.5 Let f (x) = ex − e−x .

1. Show that f is 1-1 from R onto R.

2. Find a formula of f −1 (x).

Solution. Let x1 , x2 ∈ R such that x1 ̸= x2 . WLOG x1 > x2 . Then ex1 > ex2 .
Since −x1 < −x2 , e−x1 < e−x2 . We obtain

ex2 + e−x1 > ex1 + e−x2

f (x2 ) = ex2 − e−x2 > ex1 − e−x1 = f (x1 )

( √ )
y+ y 2 +4
Then f (x1 ) ̸= f (x2 ). Thus f is 1-1 on R. Let y ∈ R. Choose x = ln 2
. Then
( √
2
) ( √
2
) √
ln y+ 2y +4 − ln y+ 2y +4 y+ y2 + 4 2
f (x) = e −e = − √ = y.
2 y + y2 + 4
Thus, f takes R onto R. Consider

f −1 (ex − e−x ) = f −1 (f (x)) = x.

( √ )
x+ x2 +4
Substitue x := ln 2
. We obtain
( ( √ ) ( √ )) )( √
2 2
−1 ln x+ 2x +4 − ln x+ 2x +4 x2 + 4 x+
f e −e = ln
2
( √ ) ( √ )
−1 x + x2 + 4 2 x + x2 + 4
f − √ = ln
2 x + x2 + 4 2
( √ )
−1 x + x2 + 4
f (x) = ln .
2
36 CHAPTER 1. THE REAL NUMBER SYSTEM

Exercises 1.4

1. For each of the following, prove f is 1-1 from A onto A. Find a formula for f −1 .

1.1 f (x) = 3x − 7 : A=R

1.2 f (x) = x2 − 2x − 1 : A = (1, ∞)

1.3 f (x) = 3x − |x| + |x − 2| : A=R

1.4 f (x) = x|x| : A=R

1
1.5 f (x) = e x : A = (0, ∞)

1.6 f (x) = tan x : A = (− π2 , π2 )

x
1.7 f (x) = : A = [−1, 1]
x2 + 1
2. Let f (x) = x2 ex where x ∈ R. Show that f is 1-1 on (0, ∞).
2

3. Suppose that A is finite and f is 1-1 from A onto B. Prove that B is finite.

4. Prove that there a fuction f that is 1-1 from {2, 4, 6, ...} onto N.

5. Prove that there a fuction f that is 1-1 from {1, 3, 5, ...} onto N.

6. Suppose that n ∈ N and ϕ : {1, 2, ..., n} → {1, 2, ..., n}.

6.1 Prove that ϕ is 1-1 if and only if ϕ in onto.

6.2 Suppose that A is finite and f : A → A. Prove that

f is 1-1 on A if and only if f takes A onto A.

∑
n
7. Let f : {1, 2, ..., n} → {1, 2, ..., n} be a 1-1 function. Show that f (x) = n!.
x=1
Chapter 2

Sequences in R

2.1 Limits of sequences

An infinite sequence (more briefly, a sequence) is a function whose domain in N. A sequence

f whose term are xn := f (n) will be defined by

x1 , x2 , x3 , ... or {xn }n∈N or {xn }∞

n=1 or {xn }.

Example 2.1.1 Use notation to represents the following sequences.

1. 1, 2, 3, ... represents the sequence {n}n∈N

2. 1, −1, 1, −1, ... represents the sequence {(−1)n }

1
Example 2.1.2 Sketch graph of {xn } and guess xn if n go to infinity where xn =
n
Y

1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
X
1 2 3 4 5 6 7 8 9 10

By the graph, we will see that xn approaches to ZERO as n go to infinity.

38 CHAPTER 2. SEQUENCES IN R

Definition 2.1.3 A sequence of real numbers {xn } is said to converge to a real number a ∈ R
if and only if for every ε > 0 there is an N ∈ N such that

n≥N implies |xn − a| < ε.

We shall use the following phrases and notations interchangeably:

(a) {xn } converges to a; (d) xn → a as n → ∞;

(b) xn converges to a; (e) the limit of {xn } exists and equals a.

(c) lim xn = a;
n→∞
Y

a+ε
a

a−ε

X
1 2 3 ··· N N + 1 ···

Theorem 2.1.4 lim k = k where k is a constant.

n→∞

Proof. Let k be a constant and ε > 0. We can choose whatever N ∈ N such that for each n ≥ N ,
we always obtain
|k − k| = 0 < ε.

So, lim k = k.
n→∞
1
Example 2.1.5 Prove that → 0 as n → ∞.
n
1
Proof. Let ε > 0. By AP, there is an N ∈ N such that < ε.
N
1 1
Let n ∈ N such that n ≥ N . Then ≤ . We obtain
n N
1 1 1
−0 = ≤ < ε.
n n N
1
Thus, → 0 as n → ∞.
n
2.1. LIMITS OF SEQUENCES 39

n
Example 2.1.6 Prove that lim =1
n→∞ n+1
1
Proof. Let ε > 0. By AP, there is an N ∈ N such that < ε.
N
1 1 1
Let n ∈ N such that n ≥ N . Then n + 1 > n ≥ N . So, < ≤ . We obtain
n+1 n N
n n − (n + 1) −1 1 1 1
−1 = = = < ≤ < ε.
n+1 n+1 n+1 n+1 n N
n
Thus, → 1 as n → ∞.
n+1
1
Example 2.1.7 Prove that → 0 as n → ∞
2n
1
Proof. Let ε > 0. By AP, there is an N ∈ N such that < ε.
N
1 1 1
Let n ∈ N such that n ≥ N . By Example 1.2.4, 2n > n. So, n < ≤ . We obtain
2 n N
1 1 1 1
− 0 = < ≤ < ε.
2n 2n n N
1
Thus, → 0 as n → ∞.
2n
1
Example 2.1.8 Prove that lim =0
n→∞ n2

√ 1 √
Proof. Let ε > 0. Then ε > 0. By AP, there is an N ∈ N such that < ε.
N
1 1
Let n ∈ N such that n ≥ N . Since n ≥ N > 0, n2 ≥ N 2 . Then 2 ≤ 2 . We obtain
n N
1 1 1
2
− 0 = 2 ≤ 2 < ε.
n n N
1
Thus, → 0 as n → ∞.
n2
(√ √ )
Example 2.1.9 Prove that lim n+1− n =0
n→∞

1
Proof. Let ε > 0. Then ε2 > 0. By AP, there is an N ∈ N such that < ε2 .
N
√ √ 1 1
Let n ∈ N such that n ≥ N . Since n ≥ N > 0, n ≥ N . Then √ ≤ √ .
n N
√ √ √ √ 1 1
Since n + 1 > 0, n + 1 + n > n. Then √ √ < √ . We obtain
n+1+ n n
√ √
√ √ √ √ n+1− n
n + 1 − n − 0 = ( n + 1 − n) · √ √
n+1− n
1 1 1
=√ √ < √ ≤ √ < ε.
n+1+ n n N
√ √
Thus, n + 1 − n → 0 as n → ∞.
40 CHAPTER 2. SEQUENCES IN R

Example 2.1.10 If xn → 1 as n → ∞. Prove that

2xn + 1 → 3 as n → ∞.

Proof. Assume that xn → 1 as n → ∞.

Let ε > 0. By assumption, there is an N ∈ N such that
ε
n≥N implies |xn − 1| < .
2
Let n ∈ N such that n ≥ N . Then

ε
|(2xn + 1) − 3| = |2(xn − 1)| = 2|xn − 1| < 2 · = ε.
2

Thus, 2xn + 1 → 3 as n → ∞.

Example 2.1.11 If xn → −1 as n → ∞. Prove that

(xn )2 → 1 as n → ∞.

Proof. Assume that xn → −1 as n → ∞.

Given ε = 1. There is an N1 ∈ N such that

n ≥ N1 implies |xn + 1| < 1.

Then, |xn | − |1| = |xn | − | − 1| ≤ |xn − (−1)| = |xn + 1| ≤ 1. So, |xn | < 2.
Let ε > 0. By assumption, there is an N2 ∈ N such that
ε
n ≥ N2 implies |xn + 1| < .
3
Let n ∈ N. Choose N = max{N1 , N2 }. For each n ≥ N , we obtain

|(xn )2 − 1| = |(xn − 1)(xn + 1)| = |xn − 1||xn + 1|

ε ε
< (|xn | + 1) < (2 + 1) = ε.
3 3

Thus, (xn )2 → 1 as n → ∞.
2.1. LIMITS OF SEQUENCES 41

Example 2.1.12 Assume that xn → 1 as n → ∞. Show that

1
→ 1 as n → ∞.
xn

Proof. Assume that xn → 1 as n → ∞.

1
Given ε = . There is an N1 ∈ N such that
2
1
n ≥ N1 implies |xn − 1| < .
2
1 1
Then 1 = |1 − xn + xn | ≤ |1 − xn | + |xn | ≤ 1
+ |xn |. So, ≤ |xn |. We get ≤ 2.
2
2 |xn |
Let ε > 0. There is an N2 ∈ N such that
ε
n ≥ N2 implies |xn − 1| < .
2
Let n ∈ N. Choose N = max{N1 , N2 }. For each n ≥ N . We obtain

1 1 − xn 1 ε
−1 = ≤ · |xn − 1| < 2 · = ε.
xn xn |xn | 2

1
Thus, → 1 as n → ∞.
xn

Example 2.1.13 Assume that xn → 1 as n → ∞. Show that

1 + (xn )2
→ 1 as n → ∞
xn + 1

Proof. Assume that xn → 1 as n → ∞.

Given ε = 1. There is an N1 ∈ N such that

n ≥ N1 implies |xn − 1| < 1.

Then |xn | − 1 ≤ |xn − 1| ≤ 1. So, |xn | ≤ 2. We consider

2 = |2 − xn + xn | = |1 − xn + 1 + xn | ≤ |1 − xn | + |1 + xn | ≤ 1 + |1 + xn |

1 ≤ |1 + xn |
1
≤ 1.
|1 + xn |

Let ε > 0. There is an N2 ∈ N such that

ε
n ≥ N2 implies |xn − 1| < .
2
42 CHAPTER 2. SEQUENCES IN R

Let n ∈ N. Choose N = max{N1 , N2 }. For each n ≥ N . We obtain

1 + (xn )2 (xn )2 − xn xn (xn − 1)

−1 = =
xn + 1 xn + 1 xn + 1
|xn ||xn − 1| 1
≤ = |xn | · · |xn − 1|
|xn + 1| |xn + 1|
ε
≤ 2 · 1 · = ε.
2
1 + (xn )2
Hence, Thus, → 1 as n → ∞.
xn + 1

Theorem 2.1.14 A sequence can have at most one limit.

Proof. Assume that a sequence {xn } converges to both a and b. We will show that a = b by
Corollary 1.1.32. Let ε > 0. By assumption, there are N1 , N2 ∈ N such that
ε
n ≥ N1 implies |xn − a| <
2
and
ε
n ≥ N2 implies |xn − b| < .
2
Choose N = max{N1 , N2 }. For each n ≥ N , we obtain

ε ε
|a − b| = |(a − xn ) + (xn − b)| ≤ |xn − a| + |xn − b| < + = ε.
2 2

Hence, a − b = 0 or a = b. We conclude that the sequence {xn } can have at most one limit.

Example 2.1.15 Show that the limit {(−1)n }n∈N has no limit or does not exist (DNE).

Proof. Suppose that (−1)n → 1 as n → ∞. Given ε = 1. There is an N ∈ N such that

n≥N implies |(−1)n − a| < 1.

Since (−1)n = ±1, |1 − a| < 1 and |1 + a| = | − 1 − a| < 1. We have

2 = |1 + 1| = |(1 − a) + (1 + a)| ≤ |1 − a| + |1 + a| < 1 + 1 = 2.

It is imposible because 2 < 2. Thus, {(−1)n }n∈N has no limit.

2.1. LIMITS OF SEQUENCES 43

SUBSEQUENCES.

Definition 2.1.16 By a subsequence of a sequence {xn }n∈N , we shall mean a sequence of the
form

{xnk }k∈N , where each nk ∈ N and n1 < n2 < n3 < ...

Example 2.1.17 Give examples for two subsequences of the following sequences.

Sequences Subsequences

1, −1, 1, −1, 1, −1, ... 1, 1, 1, ...

−1, −1, −1, ...

{n}n∈N 1, 3, 5, ...

2, 4, 6, ...

Consider {xn }. We may interest a formula of nk depending on k. Choose a subsequence {xnk }

where nk = 2k − 1 for k = 1, 2, 3, .... Then

{xn1 , xn2 , xn3 , ...} = {x1 , x3 , x5 , ...}.

Theorem 2.1.18 If {xn }n∈N converges to a and {xnk }k∈N is any subsequence of {xn }n∈N , then

xnk converges to a as k → ∞.

Proof. Assume that xn → a as n → ∞. Let {xnk } be a subsequence of {xn }.

Let ε > 0. By assumption, there is an N ∈ N such that

n≥N implies |xn − a| < ε.

Since nk ∈ N and n1 < n2 < n3 < ..., it is clear that

nk ≥ k for all k ∈ N.

Let k ∈ N such that k ≥ N . We have nk > k ≥ N . So,

|xnk − a| < ε.

Thus, xnk converges to a as k → ∞.

44 CHAPTER 2. SEQUENCES IN R

Example 2.1.19 Show that the limit {cos(nπ)}n∈N has no limit.

Solution. Choose two subsequences of {cos(nπ)}n∈N to be

nk = 2k and nk = 2k − 1.

If nk = 2k, then cos(nk π) = cos(2kπ) = 1. So, cos(2kπ) → 1 as k → ∞.

If nk = 2k − 1, then cos(nk π) = cos(2k − 1)π = −1. So, cos(2k − 1)π → −1 as k → ∞.
We will see that two subsequences coverges to different limits. Thus, {cos(nπ)}n∈N DNE.

BOUNDED SEQUENCES.

Definition 2.1.20 Let {xn } be a sequence of real numbers.

1. {xn } is said to be bounded above if and only if

there is an M ∈ R such that xn ≤ M for all n ∈ N

2. {xn } is said to be bounded below if and only if

there is an m ∈ R such that m ≤ xn for all n ∈ N

3. {xn } is said to be bounded if and only if it is both above and below or

there a K > 0 such that |xn | ≤ K for all n ∈ N

Example 2.1.21 Show that the following sequence is bounded above or bounded below or bounded.

Sequences Bounded below Bounded above Bounded

{n}n∈N Yes No No

1 ≤ n for all n ∈ N

{−n}n∈N No Yes No

−n ≤ 1 for all n ∈ N

{(−1)n }n∈N Yes Yes Yes

−1 ≤ (−1)n for all n ∈ N (−1)n ≤ 1 for all n ∈ N |(−1)n | ≤ 1 for all n ∈ N

2.1. LIMITS OF SEQUENCES 45

Theorem 2.1.22 (Bounded Convergent Theorem (BCT)) Every convergent sequence is

bounded.

Proof. Assume that xn → a as n → ∞. Given ε = 1. There is an N ∈ N such that

n≥N implies |xn − a| < 1.

Then, |xn | − |a| ≤ |xn − a| < 1. So, |xn | ≤ 1 + |a|.

Choose K = max{|x1 |, |x2 |, |x3 |, ..., |xN |, 1 + |a|}. We obtain

|xn | ≤ K for all n ∈ N.

Thus, xnk is bounded.

Example 2.1.23 Show that the limit {n}n∈N does not exist.

Solution. Suppose that {n}n∈N converges. By BCT, there is a K > 0 such that

n = |n| ≤ K for all n ∈ N (2.1)

Since K ∈ R, by AP, there is an N ∈ N such that K < N . By (2.1), n = N , we have N ≤ K.

It is imposible because
N ≤ K < N.

Thus, {n}n∈N DNE.

Example 2.1.24 Assume that xn → 1 as n → ∞. Use BCT to prove that

(xn )2 → 1 as n → ∞.

Proof. Assume that xn → 1 as n → ∞. By BCT, there is a K > 0 such that

|xn | ≤ K for all n ∈ N.

Let ε > 0. By assumption, there is an N ∈ N such that

ε
n≥N implies |xn − 1| < .
K +1
Let n ∈ N such that n ≥ N , we obtain

|(xn )2 − 1| = |(xn − 1)(xn + 1)| = |xn − 1||xn + 1|

ε ε
< (|xn | + 1) < (K + 1) = ε.
3 K +1
Thus, (xn )2 → 1 as n → ∞.
46 CHAPTER 2. SEQUENCES IN R

Exercises 2.1

1. Prove that the following limit exist.

1 5+n
1.1 3 + as n → ∞ 1.5 as n → ∞
n n2
( )
1 3
1.2 2 1 − as n → ∞ 1.6 π − √ as n → ∞
n n
2n + 1 n(n + 2)
1.3 as n → ∞ 1.7 as n → ∞
1−n n2 + 1
n2 − 1 n
1.4 as n → ∞ 1.8 3 as n → ∞
n2 n +1

2. Suppose that xn is sequence of real numbers that converges to 2 as n → ∞.

Use Definition 2.1.3, prove that each of the following limit exists.

2.1 2 − xn → 0 as n → ∞ 1
2.4 →1 as n → ∞
xn − 1
2.2 3xn + 1 → 7 as n → ∞
2 + x2n
2.3 (xn )2 + 1 → 5 as n → ∞ 2.5 →3 as n → ∞
xn

3. Assume that {xn } is a convergent sequence in R. Prove that lim (xn − xn+1 ) = 0.
n→∞

4. If xn → a as n → ∞, prove that xn+1 → a as n → ∞.

5. If xn → +∞ as n → ∞, prove that xn+1 → +∞ as n → ∞.

6. Prove that {(−1)n } has some subsequences that converge and others that do not converge.

7. Find a convergent subsequence of n + (−1)3n n.

8. Suppose that {bn } is a sequence of nonnegative numbers that converges to 0, and {xn } is a
real sequence that satisfies |xn − a| ≤ bn for large n. Prove that xn converges to a.
xn
9. Suppose that {xn } is bounded. Prove that → 0 as n → ∞ for all k ∈ N.
nk
10. Suppose that {xn } and {yn } converge to same point. Prove that xn − yn → 0 as n → ∞

11. Prove that xn → a as n → ∞ if and only if xn − a → 0 as n → ∞.

2.2. LIMIT THEOREMS 47

2.2 Limit theorems

Theorem 2.2.1 (Squeeze Theorem) Suppose that {xn }, {yn }, and {wn } are real sequences.
If xn → a and yn → a as n → ∞, and there is an N0 ∈ N such that

xn ≤ wn ≤ yn for all n ≥ N0 ,

then wn → a as n → ∞.

Proof. Let {xn }, {yn }, and {wn } be real sequences. Assume that xn → a and yn → a as n → ∞
and there is an N0 ∈ N such that

xn ≤ wn ≤ yn for all n ≥ N0 .

Let ε > 0. By assumption, there are N1 , N2 ∈ N such that

n ≥ N1 implies |xn − a| < ε or a − ε < xn < a + ε

and
n ≥ N2 implies |yn − a| < ε or a − ε < yn < a + ε.

Let n ∈ N. Choose N = max{N0 , N1 , N2 }. For each n ≥ N , we obtain

a − ε < xn ≤ wn ≤ yn < a + ε.

It implies that |wn − a| < ε. We conclude that wn → a as n → ∞.

Example 2.2.2 Use the Squeeze Theorem to prove that

sin(n2 )
lim = 0.
n→∞ 2n

Solution. By the sine fucntion property,

−1 ≤ sin(n2 ) ≤ 1 for all n ∈ N.

1 sin(n2 ) 1
Then, − n
≤ n
≤ n . From
2 2 2
1 1
lim − n
= 0 and lim n = 0.
n→∞ 2 n→∞ 2
sin(n2 )
By the Squeeze Theorem, we conclude that lim = 0.
n→∞ 2n
48 CHAPTER 2. SEQUENCES IN R

Theorem 2.2.3 Let {xn }, and {yn } be real sequences. If xn → 0 and {yn } is bounded, then

xn yn → 0 as n → ∞.

Proof. Let {xn }, and {yn } be real sequences. Assume that xn → 0 as n → ∞ and {yn } is bounded.
Then there is a K > 0 such that

|yn | ≤ K for all n ∈ N.

Let ε > 0. By assumption, there is an N ∈ N such that

ε
n≥N implies |xn | = |xn − 0| < .
K
Let n ∈ N. For each n ≥ N , we obtain

ε
|xn yn − 0| = |xn ||yn | < · K = ε.
K

Hence, xn yn → 0 as n → ∞.
cos(1 + n)
Example 2.2.4 Show that lim = 0.
n→∞ n2
Solution. By the cosine fucntion property,

| cos(1 + n)| ≤ 1 for all n ∈ N.

So, {cos(1 + n)} is bounded. From

1
lim = 0.
n→∞ n2

cos(1 + n)
By Theorem 2.2.3, we conclude that lim = 0.
n→∞ n2
2.2. LIMIT THEOREMS 49

Theorem 2.2.5 Let A ⊆ R.

1. If A has a finite supremum, then there is a sequence xn ∈ A such that

xn → sup A as n → ∞.

2. If A has a finite infimum, then there is a sequence xn ∈ A such that

xn → inf A as n → ∞.

Proof. Exercise for 1. We will prove 2. Suppose A has a finite infimum. By API, there is x ∈ A
such that

inf A ≤ x ≤ inf A + ε for all ε > 0.

We construct a sequence {xn } by

ε1 = 1, ∃x1 ∈ A such that inf A ≤ x1 ≤ inf A + 1

1 1
ε2 = , ∃x2 ∈ A such that inf A ≤ x2 ≤ inf A +
2 2
1 1
ε3 = , ∃x3 ∈ A such that inf A ≤ x3 ≤ inf A +
3 3
..
.
1 1
εn = , ∃xn ∈ A such that inf A ≤ xn ≤ inf A +
n n

Thus, {xn } is a sequence in A and satisfies

1
inf A ≤ xn < inf A +
n

By the Squeez Theorem,

( )
1
lim inf A ≤ lim xn ≤ lim inf A +
n→∞ n→∞ n→∞ n
inf A ≤ lim xn ≤ inf A
n→∞

Therefore, lim xn = inf A.

n→∞
50 CHAPTER 2. SEQUENCES IN R

Theorem 2.2.6 (Additive Property) Suppose that {xn } and {yn } are real sequences.
If {xn } and {yn } are convergent, then

lim (xn + yn ) = lim xn + lim yn .

n→∞ n→∞ n→∞

Proof. Assume that xn → a and yn → b as n → ∞. Let ε > 0. By assumption, there are

N1 , N2 ∈ N such that
ε
n ≥ N1 implies |xn − a| <
2
and
ε
n ≥ N2 implies |yn − b| < .
2
Let n ∈ N. Choose N = max{N1 , N2 }. For each n ≥ N , we obtain

ε ε
|(xn + yn ) − (a + b)| = |(xn − a) + (yn − b)| ≤ |xn − a| + |yn − b| < + = ε.
2 2

Thus, lim (xn + yn ) = a + b = lim xn + lim yn .

n→∞ n→∞ n→∞

Theorem 2.2.7 (Scalar Multiplicative Property) Let α ∈ R. If {xn } is a convergent sequence,

then
lim (αxn ) = α lim xn .
n→∞ n→∞

Proof. Assume that xn → a as n → ∞.

|α|
Let ε > 0 and α ∈ R. Then |α| + 1 > |α| ≥ 0. So, < 1. By assumption, there is an N ∈ N
|α| + 1
such that
ε
n≥N implies |xn − x| < .
|α| + 1
Let n ∈ N. For each n ≥ N , we obtain

ε |α|
|αxn − αx| = |α||xn − x| < |α| · = ε < 1 · ε = ε.
|α| + 1 |α| + 1

Thus, lim (αxn ) = αa = α lim xn .

n→∞ n→∞
2.2. LIMIT THEOREMS 51

Theorem 2.2.8 (Multiplicative Property) Suppose that {xn } and {yn } are convergent se-
quences. Then
( )( )
lim (xn yn ) = lim xn lim yn .
n→∞ n→∞ n→∞

Proof. Assume that xn → a and yn → b as n → ∞. By BCT, {xn } is bounded, i.e., there is a

K > 0 such that

|xn | ≤ K for all n ∈ N.

Let ε > 0. By assumption, there are N1 , N2 ∈ N such that

ε
n ≥ N1 implies |xn − a| <
2(|b| + 1)
and
ε
n ≥ N2 implies |yn − b| < .
2K
Let n ∈ N. Choose N = max{N1 , N2 }. For each n ≥ N , we obtain

|xn yn − ab| = |xn (yn − b) + (xn − a)b| ≤ |xn ||yn − b| + |xn − a||b|
ε ε ε ε |b| ε ε
<K· + |b| = + · < + · 1 = ε.
2K 2(|b| + 1) 2 2 (|b| + 1) 2 2

Thus, lim xn yn = ab = lim xn · lim yn .

n→∞ n→∞ n→∞

Theorem 2.2.9 (Reciprocal Property) Suppose that {xn } is a convergent sequence.

1 1
lim =
n→∞ xn lim xn
n→∞

where lim xn ̸= 0 and xn ̸= 0.

n→∞

Proof. Assume that {xn } converges to a such that a ̸= 0.

2
Given ε = . There is an N1 ∈ N such that
|a|
|a|
n ≥ N1 implies |xn − a| < .
2
52 CHAPTER 2. SEQUENCES IN R

|a| |a|
Then |a| = |a − xn + xn | ≤ |xn − a| + |xn | ≤ + |xn |. So, ≤ |xn |, i.e.,
2 2
1 2
≤ .
|xn | |a|

Let ε > 0. There is an N2 ∈ N such that

|a|2
n ≥ N2 implies |xn − a| < ε.
2
Let n ∈ N. Choose N = max{N1 , N2 }. For each n ≥ N , We obtain

1 1 a − xn 1 |xn − a| 2 |a|2
− = ≤ · < · ε = ε.
xn a axn |xn | |a| |a| 2|a|
1 1
Therefore, lim = .
n→∞ xn lim xn
n→∞

Theorem 2.2.10 (Quotient Property) Suppose that {xn } and {yn } are convergent sequences.
Then
xn lim xn
lim = n→∞
n→∞ yn lim yn
n→∞

where lim yn ̸= 0 and yn ̸= 0.

n→∞

Proof. The proof of Theorem is result from Multiplicative Property and Reciprocal Property.
n2 + n − 3
Example 2.2.11 Find the limit lim .
n→∞ 1 + 3n2

Solution.

n2 + n − 3 n2 (1 + n1 − n32 )
lim = lim
n→∞ 1 + 3n2 n→∞ n2 ( n12 + 3)
1+ − n32
1
n
= lim 1
n→∞
n2
+3
1 1
lim 1 + lim − 3 lim 2
n→∞ n→∞ n n→∞ n
=
1
lim 2 + lim 3
n→∞ n n→∞
1 + 0 − 3(0)
=
0+3
1
= .
3
2.2. LIMIT THEOREMS 53

Theorem 2.2.12 (Comparison Theorem) Suppose that {xn } and {yn } are convergent se-
quences. If there is an N0 ∈ N such that

xn ≤ yn for all n ≥ N0 ,

then
lim xn ≤ lim yn .
n→∞ n→∞

In particular, if xn ∈ [a, b] converges to some point c, then c must belong to [a, b].

Proof. Let xn → a and yn → b as n → ∞. Assume that there is an N0 ∈ N such that

xn ≤ yn for all n ≥ N0 .

Suppose that lim xn > lim yn , i.e., a > b. Then a − b > 0. By assumption, there is an N1 , N2 ∈ N
n→∞ n→∞
such that
a−b
n ≥ N1 implies |xn − a| <
2
and
a−b
n ≥ N2 implies |yn − b| < .
2
For each n ≥ max{N0 , N1 , N2 }, it follows that

a−b a−b
yn < b + =a− < xn
2 2

which contradics the assumption. Thus, a ≤ b.

We conclude by previous proof that if a ≤ xn ≤ b, a < c < b.
54 CHAPTER 2. SEQUENCES IN R

DIVERGENT.

Definition 2.2.13 Let {xn } be a sequence of real numbers.

1. {xn } is said to be diverge to +∞, written xn → +∞ as n → ∞ or lim xn = +∞

n→∞
if and only if for each M ∈ R there is an N ∈ N such that

n≥N implies xn > M .

2. {xn } is said to be diverge to −∞, written xn → −∞ as n → ∞ or lim xn = −∞

n→∞
if and only if for each M ∈ R there is an N ∈ N such that

n≥N implies xn < M .

Example 2.2.14 Show that lim n = +∞

n→∞

Proof. Let M ∈ R. By AP, there is an N ∈ N such that M < N .

Let n ∈ N such that n ≥ N . We obtain

n ≥ N > M.

Thus, lim n = +∞.

n→∞

n2
Example 2.2.15 Prove that lim = +∞.
n→∞ 1 + n

Proof. Let M ∈ R. By AP, there is an N ∈ N such that M + 1 < N.

Let n ∈ N such that n ≥ N . Then n − 1 > N − 1. Since 0 > −1, n2 > n2 − 1. We obtain

n2 n2 − 1 (n − 1)(n + 1)
> = = n − 1 > N − 1 > M.
1+n 1+n 1+n
n2
Hence, lim = +∞.
n→∞ 1 + n
2.2. LIMIT THEOREMS 55

4n2
Example 2.2.16 Prove that lim = −∞.
n→∞ 1 − 2n

Proof. Let M ∈ R. By AP, there is an N ∈ N such that − 12 M − 1

2
< N. It is equivalent to

−1 − 2N < M.

Let n ∈ N such that n ≥ N . It is clear that 2n − 1 > 0 and −2n < −2N . Since 0 < 1,

−4n2 < −4n2 + 1.

We obtain

4n2 −4n2 −4n2 + 1 (1 − 2n)(1 + 2n)

= < =
1 − 2n 2n − 1 2n − 1 2n − 1
= −1 − 2n < −1 − 2N < M

4n2
Therefore, lim = −∞.
n→∞ 1 − 2n

Example 2.2.17 Suppose that {xn } is a real sequence such that xn → +∞ as n → ∞.

If xn ̸= 0, prove that
1
lim = 0.
n→∞ xn
Proof. Assume that xn ̸= 0 and xn → +∞ as n → ∞.
Let ε > 0. By assumption, there is an N ∈ N such that
1
n≥N implies xn > .
ε
1
From > 0, for all n ≥ N it follow that
ε
1 1 1
= = < ε.
xn |xn | xn
1
Hence, lim = 0.
n→∞ xn
56 CHAPTER 2. SEQUENCES IN R

Theorem 2.2.18 Let {xn } and {yn } be a real sequence and xn ̸= 0. If {yn } is bounded and
xn → +∞ or xn → −∞ as n → ∞, then
yn
lim = 0.
n→∞ xn

Proof. Let {yn } be bounded and xn ̸= 0. There is a K > 0 such that

|yn | ≤ K for all n ∈ N.

Case 1. Assume that xn → +∞ as n → ∞. Let ε > 0. By assumption, there is an N ∈ N such

that
K
n≥N implies xn > .
ε
K 1 1 ε
Then xn > > 0. It follows that = < .
ε |xn | xn K
Let n ∈ N such that n ≥ N . We obtain
yn 1 ε
= |yn | · <K· = ε.
xn |xn | K
Case 2. Assume that xn → −∞ as n → ∞. Let ε > 0. By assumption, there is an N ∈ N such
that
K
n≥N implies xn < − .
ε
K K 1 ε
Since − < 0, |xn | > > 0. It follows that < .
ε ε |xn | K
Let n ∈ N such that n ≥ N . We obtain
yn 1 ε
= |yn | · <K· = ε.
xn |xn | K
yn
By two cases, we conclude that lim = 0.
n→∞ xn

sin n
Example 2.2.19 Show that → 0 as n → ∞.
n
Solution. By property of sine, we have

| sin n| ≤ 1 for all n ∈ N.

Since n → as n → ∞, we obtain by Theorem 2.2.18

sin n
lim = 0.
n→∞ n
2.2. LIMIT THEOREMS 57

Theorem 2.2.20 Let {xn } be a real sequence and α > 0.

1. If xn → +∞ as n → ∞, then lim (αxn ) = +∞.

n→∞

2. If xn → −∞ as n → ∞, then lim (αxn ) = −∞.

n→∞

Proof. 1. Assume that xn → +∞ as n → ∞.

Let M ∈ R and α > 0. By assumption, there is an N ∈ N such that
M
n≥N implies xn > .
α
Let n ∈ N such that n ≥ N . We obtain

M
αxn > α · = M.
α

Thus, lim αxn = +∞.

n→∞
2. Exercise.

Theorem 2.2.21 Let {xn } and {yn } be real sequences. Suppose that {yn } is bounded below and
xn → +∞ as n → ∞. Then
lim (xn + yn ) = +∞.
n→∞

Proof. Suppose that {yn } be bounded below and xn → +∞ as n → ∞. There is an m ∈ R such

that

m ≤ yn for all n ∈ N.

Let M ∈ R. By assumption, there is an N ∈ N such that

n≥N implies xn > M − m.

Let n ∈ N such that n ≥ N . We obtain

xn + yn > (M − m) + m = M.

Thus, lim (xn + yn ) = +∞.

n→∞
58 CHAPTER 2. SEQUENCES IN R

Theorem 2.2.22 Let {xn } and {yn } be real sequences such that

yn > K for some K > 0 and all n ∈ N.

It follows that

1. if xn → +∞ as n → ∞, then lim (xn yn ) = +∞

n→∞

2. if xn → −∞ as n → ∞, then lim (xn yn ) = −∞

n→∞

Proof. Let {xn } and {yn } be real sequences such that

yn > K for some K > 0 and all n ∈ N.

1. Exercise.
2. Assume that xn → −∞ as n → ∞. Let M ∈ R.
Case M = 0. There is an N ∈ N such that

n≥N implies xn < 0.

Let n ∈ N such that n ≥ N . Since yn > K > 0, we obtain

xn · yn < 0 = M.

Case M > 0. There is an N ∈ N such that

M
n≥N implies xn < − < 0.
K
Let n ∈ N such that n ≥ N . Since yn > K > 0, −yn < −K < 0. We obtain

M M M
xn · yn < − · yn = · (−yn ) < · (−K) = −M < 0 < M.
K K K
Case M < 0. There is an N ∈ N such that
M
n≥N implies xn < < 0.
K
Let n ∈ N such that n ≥ N . Since yn > K > 0, −yn < −K < 0. We obtain

M −M −M
xn · yn < · yn = · (−yn ) < · (−K) = M.
K K K
Thus, lim xn yn = −∞.
n→∞
2.2. LIMIT THEOREMS 59

Exercises 2.2

1. Prove that each of the following sequences coverges to zero.

sin(n4 + n + 1) n
1.1 xn = 1.4 xn =
n 2n
n (−1)n
1.2 xn = 2 1.5 xn =
n +1 n
√
n+1 1 + (−1)n
1.3 xn = 1.6 xn =
n+1 2n

2. Find the limit (if it exists) of each of the following sequences.

√
2n(n + 1) 2n2 − 1
2.1 xn = 2.4 xn =
n2 + 1 n+1
1 + n − 3n2 √ √
2.2 xn =
3 − 2n + n2 2.5 xn = n + 2 − n
n3 + n + 5 √
2.3 xn = 3 2.6 xn = n2 + n − n
5n + n − 1

3. Prove that each of the following sequences coverges to −∞ or +∞.

3.1 xn = n2 n2 + 1
3.4 xn =
n+1
3.2 xn = −n 1 − n2
3.5 xn =
n
n 2n
3.3 xn = √ 3.6 xn =
1+ n n

4. Let A ⊆ R. If A has a finite supremum, then there is a sequence xn ∈ A such that

xn → sup A as n → ∞.

5. Prove that given x ∈ R there is a sequence rn ∈ Q such that rn → x as n → ∞.

6. Use the result Excercise 1.2, show that the following

√
6.1 Suppose that 0 ≤ x1 ≤ 1 and xn+1 = 1 − 1 − xn for n ∈ N.
If xn → x as n → ∞, prove that x = 0 or 1.
60 CHAPTER 2. SEQUENCES IN R
√
6.2 Suppose that x1 > 0 and xn+1 = 2 + xn for n ∈ N.
If xn → x as n → ∞, prove that x = 2.

7. Let {xn } be a real sequence and α > 0. If xn → −∞ as n → ∞, then lim (αxn ) = −∞.
n→∞

8. Let {xn } and {yn } be real sequences such that yn > K for some K > 0 and all n ∈ N.
Prove that if xn → −∞ as n → ∞, then lim (xn yn ) = −∞.
n→∞

9. Let {xn } and {yn } are real sequences. Suuppose that {yn } is bounded above and xn → −∞
as n → ∞. Prove that
lim (xn + yn ) = −∞.
n→∞

10. Interpret a decimal expansion 0.a1 a2 a3 ... as

∑∞
ak
0.a1 a2 a3 ... = lim k
.
n→∞
k=1
10

Prove that

10.1 0.5 = 0.4999... 10.2 1 = 0.999...

2.3. BOLZANO-WEIERSTRASS THEOREM 61

2.3 Bolzano-Weierstrass Theorem

MONOTONE.

Definition 2.3.1 Let {xn }n∈N be a sequence of real numbers.

1. {xn } is said to be increasing if and only if x1 ≤ x2 ≤ x3 ≤ ... or

xn ≤ xn+1 for all n ∈ N.

2. {xn } is said to be decreasing if and only if x1 ≥ x2 ≥ x3 ≥ ... or

xn ≥ xn+1 for all n ∈ N.

3. {xn } is said to be monotone if and only if it is either increasing or decreasing.

If {xn } is increasing and converges to a, we shall write xn ↑ a as n → ∞.

If {xn } is decreasing and converges to a, we shall write xn ↓ a as n → ∞.

Example 2.3.2 Determine whether {xn }n∈N is increasing or decreasing or NOT both.

Sequences Decreasing Increasing Monotone

{n}n∈N Yes No No

1 ≤ 2 ≤ 3 ≤ ...

No Yes No
{ }
1
1≥ 1
2
≥ 1
3
≥ ...
n n∈N

{1}n∈N Yes Yes Yes

1 ≤ 1 ≤ 1 ≤ ... 1 ≥ 1 ≥ 1 ≥ ...

{(−1)n }n∈N No No No

−1 ≤ 1 ≥ −1 ≤ 1 ≥ ... −1 ≤ 1 ≥ −1 ≤ 1 ≥ ...
62 CHAPTER 2. SEQUENCES IN R

Theorem 2.3.3 (Monotone Converegence Theorem (MCT)) If {xn } is increasing and

bounded above, or if it is decreasing and bounded below, then {xn } has a finite limit.

Proof. Assume that {xn } is increasing and bounded above. By the Completeness Axiom, the
supremum

a := sup{xn : n ∈ N} exists and is finite.

Let ε > 0. By APS, there is an N ∈ N such that a − ε < xN ≤ a.

Since {xn } is increasing, xN ≤ xn for all n ≥ N . From xn ≤ a for all n ∈ N. It follows that

a − ε < xn ≤ a for all n ≥ N .

So, −ε < xn − a ≤ 0 < ε. We obtain |xn − a| < ε. We conclude that xn → a as n → ∞.

Exercise for the case that {xn } is decreasing and bounded below.

Theorem 2.3.4 If |a| < 1, then an → 0 as n → ∞.

Proof. Let |a| < 1.

Case 1 a = 0. Then an = 0 for all n ∈ N, and it follows that an → 0 as n → ∞.
Case 2 a ̸= 0. Then |a| > 0. We obtain

0 < |a|n+1 < |a|n < 1 for all n ∈ N.

So, {|a|n } is decreasing and bounded below by 0 . By MCT, |a|n → L as n → ∞.

Suppose that L ̸= 0. Then

L = lim |a|n+1 = lim |a|n |a| = |a| lim |a|n = |a|L.

n→∞ n→∞ n→∞

We have |a| = 1 which contradics |a| < 1. Thus, L = 0.

{ n+1 }
3 +1
Example 2.3.5 Find the limit of .
3n + 2n
Solution.
3n+1 + 1 3n (3 + ( 13 )n ) 3 + ( 13 )2 3+0
lim n n
= lim n 2 n
= lim 2 n
= = 3.
n→∞ 3 + 2 n→∞ 3 (1 + ( ) ) n→∞ 1 + ( ) 1+0
3 3
2.3. BOLZANO-WEIERSTRASS THEOREM 63

Definition 2.3.6 A sequence of sets {In }n∈N is said to be nested if and only if

I1 ⊇ I2 ⊇ I3 ⊇ ... or In+1 ⊆ In for all n ∈ N.

Example 2.3.7 Show that In = [ n1 , 1] is nested.

Proof. Let n ∈ N and x ∈ In+1 . Then 1 ≤ x ≤ 1

n+1
. Since n + 1 > n,

1 1
1≤x≤ < .
n+1 n

Then x ∈ In . Thus, In+1 ⊆ In . We conclude that {In }n∈N is nested.

Theorem 2.3.8 (Nested Interval Property) If {In }n∈N is a nested sequence of nonempty closed
bounded intervals, then
∩
E= In := {x : x ∈ In for all n ∈ N}
n∈N

contains at least one number. Moreover, if the lengths of these intervals satisfy |In | → 0 as n → ∞,
then E contains exactly one number.

Proof. Let In = [an , bn ] be nested. Then

[an+1 , bn+1 ] ⊆ [an , bn ] for all n ∈ N.

We obtain a1 ≤ a2 ≤ a3 ≤ ... and b1 ≥ b2 ≥ b3 ≥ ... So, {an } is increasing and bounded above by
a1 and {bn } is decreasing bounded below by b1 . By MCT, there are a and b such that an → a and
bn → b as n → ∞.

a1 a2 a3 ... a b ... b3 b2 b1

Since an ≤ bn for all n ∈ N, it also follows from the Comparison Theorem that

an ≤ a ≤ b ≤ bn .

Hence, a number x belongs to In for all n ∈ N if and only if a ≤ x ≤ b. We obtain E = [a, b].
Suppose that |In | → 0 as n → ∞. Then bn − an → 0 as n → ∞, and we have by Addition
Property that a − b = 0. In particular, E = [a, a] = {a} contain exactly one number.
64 CHAPTER 2. SEQUENCES IN R

Theorem 2.3.9 (Bolzano-Weierstrass Theorem) Every bounded sequence of real numbers has
a convergence subsequence.

Proof. Let {xn } be a bounded sequence. Choose a, b ∈ R such that

xn ∈ [a, b] for all n ∈ N.

Set I0 = [a, b]. Divide I0 into two halves, I0 = [a, a+b

2
] ∪ [ a+b
2
, b]. So, at least one of these half
intervals contains xn for infinitely many n. Call it I1 , and choose n1 > 1 such that xn1 ∈ I1 . Notice
that
|I0 | b−a
|I1 | = = .
2 2
Suppose that I0 ⊇ I1 ⊇ I2 ⊇ ... ⊇ Im and natural numbers n1 < n2 < ... < nm have been chosen
such that for each 0 ≤ k ≤ m,

b−a
|Ik | = , xnk ∈ Ik , and xn ∈ Ik for infinitely many n. (2.2)
2k

To choose Im+1 , divide Im = [am , bm ] into two halves, Im = [am , am +b

2
m
] ∪ [ am +b
2
m
, bm ]. So, at least
one of these half intervals contains xn for infinitely many n. Call it Im+1 , and choose nm+1 > nm
such that xnm+1 ∈ Im+1 . Since
|Im | bm − am
|Im+1 | = = ,
2 2m+1
it follows by induction that there is a nested sequence {Ik }k∈N of nonempty closed bounded intervals
that satisfy (2.2) for all k ∈ N. By Nested Interval Property, there is an x ∈ R that belongs to Ik
for all k ∈ N. Since x ∈ Ik , we have by (2.2) that
b−a
0 ≤ |xnk − x| ≤ |Ik | ≤ for all k ∈ N.
2k
Thus by the Squeeze Theorem, xnk → x as k → ∞.
2.3. BOLZANO-WEIERSTRASS THEOREM 65

Exercises 2.3

1. Prove that
(n2 + 22n + 65) sin(n3 )
xn =
n2 + n + 1
has a convergence sunsequence.

2. If {xn } is decreasing and bounded below, then {xn } has a finite limit.

3. Suppose that E ⊂ R is nonempty bpunded set and sup E ∈

/ E. Prove that there exist a
strictly increasing sequence {xn } (x1 < x2 < x3 < ...) that converges to sup E such that
xn ∈ E for all n ∈ N.

4. Suppose that {xn } is a monotone increasing in R (not necessarily bounded above). Prove
that there is extended real number x such that xn → x as n → ∞.
√
5. Suppose that 0 < x1 < 1 and xn+1 = 1 − 1 − xn for n ∈ N. Prove that

xn+1 1
xn ↓ 0 as n → ∞ and → , as n → ∞
xn 2
1 n+1
6. If a > 0, prove that a n → 1 as n → ∞. Use the resulte to find the limit of {3 n }.
√
7. Let 0 ≤ x1 ≤ 3 and xn+1 = 2xn + 3 for n ∈ N. Prove that xn ↑ 3 as n → ∞.
√
8. Suppose that x1 ≥ 2 and xn+1 = 1 + xn − 1 for n ∈ N. Prove that xn ↓ 2 as n → ∞. What
happens when 1 ≤ x1 < 2 ?

9. Prove that 



1 if x > 0


1
lim x 2n−1 = 0 if x = 0
n→∞ 




−1 if x < 0

1 + xn−1
10. Suppose that x0 ∈ R and xn = for n ∈ N. Prove that xn → 1 as n → ∞.
2
11. Let {xn } be a sequence in R. Prove that

11.1 if xn ↓ 0, then xn > 0 for all n ∈ N.

66 CHAPTER 2. SEQUENCES IN R

11.2 if xn ↑ 0, then xn < 0 for all n ∈ N.

12. Let 0 < y1 < x1 and set

xn + yn √
xn+1 = and yn+1 = xn yn , for n ∈ N
2

12.1 Prove that 0 < yn < xn for all n ∈ N.

12.2 Prove that yn is increasing and bounded above, and xn is decreasing and bounded below.
x1 − y1
12.3 Prove that 0 < xn+1 − yn+1 < for n ∈ N
2n
12.4 Prove that lim xn = lim yn . (the common value is called the arithemetic-geometric
n→∞ n→∞
mean of x1 and y1 .)

13. Suppose that x0 = 1, y0 = 0

xn = xn−1 + 2yn−1 ,

and
yn = xn−1 + yn−1

for n ∈ N. Prove that x2n − 2yn2 = ±1 for n ∈ N and

xn √
→ 2 as n → ∞.
yn
√
14. (Archimedes) Suppose that x0 = 2 3, y0 = 3,

2xn−1 yn−1 √
xn = , and yn = xn yn−1 for n ∈ N.
xn−1 + yn−1

14.1 Prove that xn ↓ x and yn ↑ y, as n → ∞, for some x, y ∈ R.

14.2 Prove that x = y and

3.14155 < x < 3.14161.

(The actual value of x is π.)

2.4. CAUCHY SEQUENCES 67

2.4 Cauchy sequences

Definition 2.4.1 A sequence of points xn ∈ R is said to be Cauchy if and only if every ε > 0
there is an N ∈ N such that

n, m ≥ N imply |xn − xm | < ε.

{ }
1
Example 2.4.2 Show that is Cauchy.
n
ε
Proof. Let ε > 0. By AP, there is an N ∈ N such that 1
N
< .
2
Let m, n ∈ N such that n, m ≥ N . Then, 1
n
≤ 1
N
and 1
m
≤ N . We obtain
1

1 1 1 1 1 1 2
− ≤ + ≤ + = < ε.
n m n m N N N
{ }
1
Thus, is Cauchy.
n

Theorem 2.4.3 The sum of two Cauchy sequences is Cauchy.

Proof. Let {xn } and {yn } be Cauchy. Let ε > 0. There are N1 , N2 ∈ N such that

ε
m, n ≥ N1 imply |xn − xm | <
2
and
ε
m, n ≥ N2 imply |yn − ym | < .
2

Choose N = max{N1 , N2 }. For m, n ≥ N , we obtain

|(xn + yn ) − (xm + ym )| = |(xn − xm ) + (yn − ym )|

≤ |xn − xm | + |yn − ym |
ε ε
< + = ε.
2 2

Thus, {xn + yn } is Cauchy.

68 CHAPTER 2. SEQUENCES IN R

Theorem 2.4.4 If {xn } is convergent, then {xn } is Cauchy.

Proof. Assume that xn → a as n → ∞. There are an N ∈ N such that

ε
n≥N implies |xn − a| < .
2
Let n, m ∈ N such that n, m ≥ N . We obtain
ε ε
|xn − xm | = |(xn − a) − (xm − a)| ≤ |xn − a| + |xm − a| < + = ε.
2 2
Hence, {xn } is Cauchy.

Theorem 2.4.5 (Cauchy’s Theorem) Let {xn } be a sequence of real numbers. Then

{xn } is Cauchy if and only if {xn } converges to some point in R.

Proof. Assume that {xn } is Cauchy. Given ε = 1. There is an N0 ∈ N such that

|xm − xN0 | < 1 for all m ≥ N0 .

Then, |xm | < 1 + |xN0 | for m ≥ N0 . Thus, {xn } is bounded by

M = max{|x1 |, |x2 |, ..., |xN0 −1 |, 1 + |xN0 |}.

By Bolzano-Weierstrass Theorem, {xn } has a convergent subsequence {xnk } by xnk → a as n → ∞.

Let ε > 0. There is an N1 ∈ N such that
ε
k ≥ N1 implies |xnk − a| < .
2
Since {xn } is Caucy, thereis an N2 ∈ N such that
ε
m, n ≥ N2 implies |xm − xn | < .
2
Let n ∈ N. Choose N = max{N0 , N1 , N2 }. For each n ≥ N , we have nk ≥ N since nk ≥ n. Then,
we obtain
ε ε
|xn − a| = |(xn − xnk ) + (xnk − a)| ≤ |xn − xnk | + |xnk − a| < + = ε.
2 2
Thus, {xn } converges to a.
Coversely, it is clear by Theorem 2.4.4.
2.4. CAUCHY SEQUENCES 69

Example 2.4.6 Prove that any real sequence {xn } that satisfies

1
|xn − xn+1 | ≤ , n ∈ N,
2n

is convergent.
1
Proof. Let ε > 0. By AP, there is an N ∈ N such that < ε.
N
1 1
Let n, m ∈ N such that n, m ≥ N . Then ≤ . By the fact that n < 2n for all n ∈ N,
n N
1 1 1
we get n < . Suppose that m > n. Then m − n > 0. So, 1 − m−n ≤ 1. We obtain
2 n 2

|xn − xm | = |xn − xn+1 + xn+1 − xn+2 + xn+2 − · · · + xm−1 − xm |

≤ |xn − xn+1 | + |xn+1 − xn+2 | + · · · + |xm−1 − xm |

1 1 1
< n
+ n+1 + · · · + m−1
2 [2 2 ]
1 1 1 1
= n−1 + + · · · + m−n
2 2 22 2
1 ∑ 1
m−n
= n−1
2 2k
[ k=1
]
1 1
= n 1 − m−n
2 2
1
≤ n
2
1
<
n
1
< <ε
N

Thus, {xn } is Cauchy. Therefore, {xn } is convergent.

70 CHAPTER 2. SEQUENCES IN R

Exercises 2.4

1. Use definition to show that {xn } is Cauchy if

1 n
1.1 xn = 1.2 xn =
n2 n+1

2. Prove that the product of two Cauchy sequences is Cauchy.

3. Prove that if {xn } is a sequence that satisfies

1+n
|xn | ≤
1 + n + 2n2

for all n ∈ N, then {xn } is Cauchy.

4. Suppose that xn ∈ N for n ∈ N. If {xn } is Cauchy prove that there are numbers a and N
such that xn = a for all n ≥ N .

5. Let {an } be a sequence in R such that there is an N ∈ N satisfying the statement:

1
if n, m ≥ N , then |xn − xm | < for all k ∈ N.
k

Prove that {an } converges.

∑n
lim xk exists and is finite.
n→∞
k=1

6. Let {xn } be Cauchy. Prove that {xn } converges if and only if at least one of its subsequence
converges.
∑
n
(−1)k
7. Prove that lim exists and is finite.
n→∞
k=1
k

8. Let {xn } be a sequence. Suppose that there is an a > 1 such that

|xk+1 − xk | ≤ a−k

for all k ∈ N. Prove that xn → x for some x ∈ R.

9. Show that a sequence that satisfies xn+1 − xn → 0 is not necessarily Cauchy.

Chapter 3

Topology on R

3.1 Open sets

Open sets are among the most important subsets of R. A collection of open sets is called a
topology, and any property (such as convergence, compactness, or continuity) that can be dened
entirely in terms of open sets is called a topological property.

Definition 3.1.1 A set E ⊆ R is open if for every x ∈ E there exists a δ > 0 such that

(x − δ, x + δ) ⊆ E.

In other word,

E is open ↔ ∀x ∈ E ∃δ > 0, (x − δ, x + δ) ⊆ E

and

E is not open ↔ ∃x ∈ E ∀δ > 0, (x − δ, x + δ) * E.

Since the empty set has no element, by definition it imples that ∅ is open. For E = R, we
obatin

∀x ∈ R ∃δ > 0, (x − δ, x + δ) ⊆ R is true.

It follows that R is open.

72 CHAPTER 3. TOPOLOGY ON R

Example 3.1.2 Show that interval (0, 1) is open.

{ }
x 1−x
Proof. Let x ∈ (0, 1). Choose δ = min , .
2 2
1−x

0 x 1
x

We obtain (x − δ, x + δ) ⊆ (0, 1). Hence, (0, 1) is open.

Theorem 3.1.3 Intervals (a, b), (a, ∞) and (−∞, b) are open.

{ }
x−a b−x
Proof. 1. Let x ∈ (a, b). Choose δ = min , . We obtain (x − δ, x + δ) ⊆ (a, b).
2 2
Hence, (a, b) is open.
x−a
2. Let x ∈ (a, ∞). Choose δ = . We obtain (x − δ, x + δ) ⊆ (a, ∞). Hence, (a, ∞) is open.
2
b−x
3. Let x ∈ (−∞, b). Choose δ = . We obtain (x − δ, x + δ) ⊆ (−∞, b). Hence, (−∞, b) is
2
open.

Example 3.1.4 Show that [0, 1) is not open.

Proof. Suppose that [0, 1) is open. Given x = 0, there is a δ > 0 such that

(−δ, δ) ⊆ [0, 1).

Since −δ < − 2δ < 0, − 2δ ∈ (−δ, δ). It implies that − 2δ ∈ [0, 1) which is imposible.
3.1. OPEN SETS 73

Theorem 3.1.5 Let A and B be open. Prove that A ∪ B and A ∩ B are open.

Proof. Let A and B be open.

1. Let x ∈ A ∪ B. Then x ∈ A. There is a δ > 0 such that (x − δ, x + δ) ⊆ A.

Since A ⊆ A ∪ B, (x − δ, x + δ) ⊆ A ∪ B. Thus, A ∪ B is open.

2. Let x ∈ A ∩ B. Then x ∈ A and x ∈ B. There are δ1 , δ2 > 0 such that

(x − δ1 , x + δ1 ) ⊆ A and (x − δ2 , x + δ2 ) ⊆ B.

Choose δ = min{δ1 , δ2 }. We obtain (x − δ, x + δ) ⊆ A ∩ B. Thus, A ∩ B is open.

Theorem 3.1.6 Let A1 , A2 , .., An be open sets. Then

∪
n
1. Ak := A1 ∪ A2 ∪ ... ∪ An is open.
k=1

∩
n
2. Ak := A1 ∩ A2 ∩ ... ∩ An is open.
k=1

Proof. Excercise
74 CHAPTER 3. TOPOLOGY ON R

NEIGHBORHOOD.
Next, we introduce the notion of the neighborhood of a point, which often gives clearer, but
equivalent, descriptions of topological concepts than ones that use open intervals.

Definition 3.1.7 A set U ⊆ R is a neighborhood of a point x ∈ R if

(x − δ, x + δ) ⊆ U for some δ > 0.

For example x = 1, we have (0, 2), [0, 2] and [0, 2) to be neighborhoods of 1.

Theorem 3.1.8 A set E ⊆ R is open if every x ∈ E has a neighborhood U such that U ⊆ E.

Proof. If every x ∈ E has a neighborhood U such that U ⊆ E, then there is a δ > 0 such that

(x − δ, x + δ) ⊆ U ⊆ E.

Hence, E ⊆ R is open .

Theorem 3.1.9 A sequence {xn } of real numbers converges to a limit x ∈ R if and only if for
every neighborhood U of x there exists N ∈ N such that xn ∈ U for all n > N .

Proof. Assume that xn → x as n → ∞. Let U be a neighborhood of x. There is a ε > 0 such that

(x − ε, x + ε) ⊆ U.

By assumption, there is an N ∈ N such that n ≥ N implies |xn − x| < ε.

It follows that x − ε < xn < x + ε. Thus, xn ∈ (x − ε, x + ε) ⊆ U for all n ≥ N .
Conversely, assume that for every neighborhood U of x there exists N ∈ N such that xn ∈ U for
all n > N . Let ε > 0. Fixed x. Then (x − ε, x + ε) is a neighborhood of x.
By assumption, there exists N ∈ N such that xn ∈ (x − ε, x + ε) for all n > N . We have

|xn − a| < ε for all n ≥ N .

Therefore, xn → x as n → ∞.
3.1. OPEN SETS 75

Exercises 3.1

1. Show that interval [a, b], [a, b) and (a, b], are not open.

2. Show that interval [a, ∞) and (−∞, b] are not open.

3. Give two neighborhoods of x = 2.

4. Let A and B be subsets of R. Suppose that A and B are open.

Determine whether A\B is open.

5. Let U ⊆ R be a nonempty open set. Show that sup U ∈

/ U and inf U ∈
/ U.

6. Let A1 , A2 , .., An be open sets. Prove that

∪
n
6.1 Ak := A1 ∪ A2 ∪ ... ∪ An is open.
k=1
∩n
6.2 Ak := A1 ∩ A2 ∩ ... ∩ An is open.
k=1

7. Find a sequence In of bounded, and open interval that

∩
∞
In+1 ⊂ In for each n ∈ N and In = ∅.
n=1
76 CHAPTER 3. TOPOLOGY ON R

3.2 Closed sets

Definition 3.2.1 A set F ⊆ R is closed if

F c = R\F = {x ∈ R : x ∈
/ F } is open.

Since ∅c = R and Rc = ∅ ( ∅ and R are open), ∅ and R are closed sets.

Example 3.2.2 Show that interval [0, 1] is closed.

Solution. Consider [0, 1]c = (−∞, 0) ∪ (1, ∞). By Theorem 3.1.3 and 3.1.5, we obtain

(−∞, 0) ∪ (1, ∞) is open.

We conclude that [0, 1] is closed.

Example 3.2.3 Show that [0, 1) is neither open nor closed.

Solution. Consider [0, 1)c = (−∞, 0) ∪ [1, ∞). Choose x = 1. Then

(1 − δ, 1 + δ) * (−∞, 0) ∪ [1, ∞) for all δ > 0.

So, (−∞, 0) ∪ [1, ∞) is not open. We conclude that [0, 1) is neither open nor closed.

Theorem 3.2.4 Let A and B be closed. Prove that A ∪ B and A ∩ B are closed.

Proof. Let A and B be closed. Then Ac and B c are open. By Theorem 3.1.5, it implies that

Ac ∩ B c and Ac ∪ B c are open.

Since (A ∪ B)c = Ac ∩ B c and (A ∩ B)c = Ac ∪ B c ,

(A ∪ B)c and (A ∩ B)c are open.

We conclude that A ∪ B and A ∩ B are closed.

3.2. CLOSED SETS 77

Theorem 3.2.5 Let A1 , A2 , .., An be closed sets. Then

∪
n
1. Ak := A1 ∪ A2 ∪ ... ∪ An is closed.
k=1

∩
n
2. Ak := A1 ∩ A2 ∩ ... ∩ An is closed.
k=1

Proof. Let A1 , A2 , .., An be closed sets. Then Ac1 , Ac2 , .., Acn are open. We consider
( n )c
∪
Ak = (A1 ∪ A2 ∪ ... ∪ An )c = Ac1 ∩ Ac2 ∩ ... ∩ Acn
(k=1 )c
∩n
Ak = (A1 ∩ A2 ∩ ... ∩ An )c = Ac1 ∪ Ac2 ∪ ... ∪ Acn
k=1

By theorem 3.1.6, it follows that

( n )c ( )c
∪ ∩
n
Ak and Ak are open.
k=1 k=1

The proof of Theorem is complete.

78 CHAPTER 3. TOPOLOGY ON R

Exercises 3.2

1. Show that interval [a, b], [a, ∞) and (−∞, b] are closed.

2. The set of rational numbers Q ⊂ R is neither open nor closed.

3. Show that every closed interval I is a closed set.

∩∞ ( )
1 n+1
4. Is − , open or closed ?
n=1
n n

∪∞ [ ]
1 n−1
5. Is , open or closed ?
n=1
n n

6. Suppose, for n ∈ N, the intervals In = [an , bn ] are such that In+1 ⊂ In . If

a = sup{an : n ∈ N} and b = inf{bn : n ∈ N},

∩
∞
show that In = [a, b].
n=1

∩
∞
7. Find a sequence In of closed interval that In+1 ⊂ In for each n ∈ N and In = ∅.
n=1

8. Suppose that U ⊆ R is a nonempty open set. For each x ∈ U , let

(
Jx = x − ε, x + δ),

where the union is taken over all ε > 0 and δ > 0 such that (x − ε, x + δ) ⊂ U .

8.1 Show that for every x, y ∈ U , either Jx ∩ Jy = ∅, or Jx = Jy .

∪
8.2 Show that U = Jx , where B ⊆ U is either finite or countable.
x∈B
3.3. LIMIT POINTS 79

3.3 Limit points

Definition 3.3.1 A point x ∈ R is called a limit point of a set A ⊆ R if for every ε > 0 there
exists a ∈ A, a ̸= x, such that a ∈ (x − ε, x + ε) or

[(x − ε, x) ∪ (x, x + ε)] ∩ A ̸= ∅.

We denote the set of all limit points of a set A by A′ .

( )
x

Definition 3.3.2 Let A ⊆ R. Then x ∈ R is an interior point of A if there exists an δ > 0

such that
(x − δ, x + δ) ⊆ A.

The set of all interior points of A is called the interior of A, denoted A◦ .

( )
x

Definition 3.3.3 Suppose A ⊆ R. A point x ∈ A is called an isolated point of A if there exists

an δ > 0 such that
A ∩ (x − δ, x + δ) = {x}.

( )
x

A
80 CHAPTER 3. TOPOLOGY ON R

Example 3.3.4 Fill the blanks of the following table.

Set Set of limit points Set of interior points Set of isolated points

[0, 1] [0, 1] (0, 1) ∅

(0, 1) [0, 1] (0, 1) ∅

[0, 1) [0, 1] (0, 1) ∅

(0, 1] ∪ {3} [0, 1] (0, 1) {3}

{1} ∅ ∅ {1}

N ∅ ∅ N

Q R R ∅

Example 3.3.5 Show that 0 is a limit point of (0, 1).

1
Proof. Let ε > 0. By AP, there is an N ∈ N such that < ε.
N
1
Choose a = . We have,
N +1
1 1
< < ε.
N +1 N
1 1
It implies that ∈ (−ε, ε). Since N + 1 > 1, 0 < < 1. We obatin
N +1 N +1
1
∈ (0, 1).
N +1

We obtain
[(−ε, 0) ∪ (0, ε)] ∩ (0, 1) ̸= ∅.

Thus, 0 is a limit point of (0, 1).

3.3. LIMIT POINTS 81

Theorem 3.3.6 Let A and B be sets. If A ⊆ B, then A′ ⊆ B ′ .

Proof. Let A and B be sets such that A ⊆ B. Let x ∈ A′ . Then, for all ε > 0, we obtain

[(x − ε, x) ∪ (x, x + ε)] ∩ A ̸= ∅.

Since A ⊆ B,

[(x − ε, x) ∪ (x, x + ε)] ∩ B ̸= ∅.

So, x ∈ B ′ . We conclude that A′ ⊆ B ′ .

Theorem 3.3.7 Let A be a closed subset of R. Then A′ ⊆ A.

Proof. Assume that A is closed. Then Ac is open.

Let x ∈ A′ or x be a limit point of A.
Suppose that x ∈
/ A. Then x ∈ Ac . There is an ε > 0 such that

(x − ε, x + ε) ⊆ Ac .

It follows that (x − ε, x + ε) ∩ A = ∅. Since x ∈

/ A,

[(x − ε, x) ∪ (x, x + ε)] ∩ A = ∅.

So, x is not a limit point of A which is imposible. Thus, x ∈ A.

82 CHAPTER 3. TOPOLOGY ON R

CLOSURE.

Definition 3.3.8 Given a set A ⊆ R, the set Ā = A ∪ A′ is called the closure of A.

Example 3.3.9 Fill the blanks of the following table.

Set Set of limit points Closure

[0, 1] [0, 1] [0, 1]

(0, 1) [0, 1] [0, 1]

[0, 1) [0, 1] [0, 1]

(0, 1] ∪ {3} [0, 1] [0, 1] ∪ {3}

{1} ∅ {1}

N ∅ N

Q R R

Theorem 3.3.10 Let A and B be subsets of R. If A ⊆ B, then Ā ⊆ B̄.

Proof. Let A and B be sets such that A ⊆ B. By Theorem 3.3.6, it implies that A′ ⊆ B ′ .
We conclude that Ā = A ∪ A′ ⊆ B ∪ B ′ = B̄.

Theorem 3.3.11 Let A ⊆ R. Then Ā is closed.

Proof. Let x ∈ (Ā)c = (A ∪ A′ )c . Then x ∈ / A′ . There is an ε > 0 such that

/ A and x ∈

(x − ε, x + ε) ∩ A = [(x − ε, x) ∪ (x, x + ε)] ∩ A = ∅.

Since x ∈
/ A, (x − ε, x + ε) ∩ A = [(x − ε, x) ∪ (x, x + ε)] ∩ A. Use the fact that A ⊆ Ā, we obtain

(x − ε, x + ε) ∩ Ā = ∅.

So, (x − ε, x + ε) ⊆ (Ā)c . Thus, (Ā)c is open. We conclude that Ā is closed.

3.3. LIMIT POINTS 83

Theorem 3.3.12 Let A ⊆ R. Then A is closed if and only if A = Ā.

Proof. Assume that A is closed. By Theorem 3.3.7, A′ ⊆ A. It follows that

Ā = A ∪ A′ ⊆ A.

From definition of closer, A ⊆ A ∪ A′ = Ā. Thus, A = Ā.

Coversely, assume that A = Ā. By Theorem 3.3.11, Ā is closed. Hence, A is also closed.

Theorem 3.3.13 A set F ⊆ R is closed if and only if

the limit of every convergent sequence in F belongs to F .

Proof. Let F be a closed set. Assume that {xn } is a sequence in F . We will prove by contradiction.
Assume that xn → a as n → ∞ and a ∈
/ F . Then a ∈ F c .
Since F c is open, there δ > 0 such that (a − δ, a + δ) ⊆ F c . So,

(a − δ, a + δ) ∩ F = ∅ (3.1)

From xn → a as n → ∞, (ε = δ) there is an N ∈ N such that

n≥N implies |xn − a| < δ.

Then xn ∈ (a − δ, a + δ). But xn ∈ F , this is contradiction to (3.1). Thus, a ∈ F .

Coversely, we will prove in Excercise.
84 CHAPTER 3. TOPOLOGY ON R

Exercises 3.3
1. Identify the limit points, interior point and isolated points of the following sets:

1.1 A = (0, 1) ∪ {3} 1.4 A = (0, 1) ∪ [3, 4]

{ }
1
1.2 A = [0, 1]c 1.5 A = :n∈N
n
1.3 A = [1, ∞) 1.6 A = [0, 1] ∩ Q

2. Find A′ , A◦ and Ā where

2.1 A = (0, 1) 2.4 A = (0, 1) ∪ {2, 3}

{ }
1
2.2 A = [0, 1] 2.5 A = :n∈N
n2
2.3 A = [0, ∞) 2.6 A = Q

3. Let A and B be two subset of R. Show that (A ∪ B)′ = A′ ∪ B ′ .

4. Let A and B be two subset of R. Determine whether

4.1 (A ∩ B)′ = A′ ∩ B ′

4.2 A ∪ B = Ā ∪ B̄

4.3 A ∩ B = Ā ∩ B̄

4.4 (A ∪ B)◦ = A◦ ∪ B ◦

4.5 (A ∩ B)◦ = A◦ ∩ B ◦

4.6 if Ā ⊆ B̄, then A ⊆ B.

5. Prove that A◦ is open.

6. Prove that A is open if and only if A = A◦ .

7. Suppose x is a limit point of the set A. Show that for every ε > 0, the set

(x − ε, x + ε) ∩ A is infinite.
∪
∞ ∪
∞
8. Suppose that Ak ⊆ R for each k ∈ N, and let B = Ak . Show that B̄ = Āk .
k=1 k=1

9. If the limit of every convergent sequence in F belongs to F ⊆ R, prove that F is closed.

Chapter 4

Limit of Functions

4.1 Limit of Functions

Definition 4.1.1 Let E ⊆ R and f : E → R be a function and let a ∈ R be a limit point of E.

Then f (x) is said to converge to L, as x approaches a, if and only if
for every ε > 0 there is a δ > 0 such that for all x ∈ E,

0 < |x − a| < δ implies |f (x) − L| < ε.

In this case we write

lim f (x) = L or f (x) → L as x → a.

x→a

and call L the limit of f (x) as x approaches a.

Y
y = f (x)

L+ε

L
L−ε

X
a−δ a a+δ
86 CHAPTER 4. LIMIT OF FUNCTIONS

Example 4.1.2 Suppose that f (x) = 2x + 1. Prove that

lim f (x) = 3.
x→1

ε
Proof. Let ε > 0. Choose δ = > 0. Let x ∈ R such that 0 < |x − 1| < δ. We obtain
2

|f (x) − 3| = |(2x + 1) − 3| = |2(x − 1)| = 2|x − 1| < 2δ = ε.

Thus, f (x) → 3 as x → 1.
√
Example 4.1.3 Let f (x) = x2 where x ∈ R. Prove that f (x) → 0 as x → 0.

Proof. Let ε > 0. Choose δ = ε > 0. Let x ∈ R such that 0 < |x| < δ. We obtain
√
|f (x) − 0| = | x2 − 0| = |x| < ε.
√
Thus, x2 → 0 as x → 0.

Example 4.1.4 Prove that ( )

1
lim x cos = 0.
x→0 x
Proof. Let ε > 0. Choose δ = ε > 0. Let x ∈ R such that 0 < |x| < δ.
Use the property of cosine that
( )
1
cos ≤ 1 for all x ̸= 0.
x
We obtain
( ) ( ) ( )
1 1 1
x cos − 0 = x cos = |x| cos ≤ |x| · 1 = |x| < δ = ε.
x x x
( )
1
Thus, x cos → 0 as x → 0.
x
4.1. LIMIT OF FUNCTIONS 87

Example 4.1.5 Prove that

lim x2 = 9.
x→3
{ ε}
Proof. Let ε > 0. Choose δ = min 1, . Let x ∈ R such that 0 < |x − 3| < δ.
7
Then 0 < |x − 3| < 1. By Triangle inequality, |x| − 3 < |x − 3| < 1. So, |x| < 4. We obtain

ε
|x2 − 9| = |(x + 3)(x − 3)| = |x + 3||x − 3| ≤ (|x| + |3|)δ < (4 + 3) = ε.
7
√
Thus, x → 0 as x → 0.
1
Example 4.1.6 Prove that f (x) = → 1 as x → 1.
x
{ }
1 ε
Proof. Let ε > 0. Choose δ = min , . Let x ∈ R\{0} such that 0 < |x − 1| < δ.
2 2
1
Then 0 < |x − 1| < . By Triangle inequality,
2
1
1 = |1 − x + x| ≤ |1 − x| + |x| < + |x|.
2
1 1
So, |x| > . It follows that < 2. We obtain
2 |x|

1 1−x 1 ε
−1 = = · |x − 1| < 2δ < 2 · = ε.
x x |x| 2
1
Thus, f (x) → as x → 1.
x

Theorem 4.1.7 (Limit of Constant function) The limit of a constant function is equal to the
constant.

Proof. Let K be a constant. Define f (x) = K for all x ∈ R.

Let a ∈ R and ε > 0. Whatever a positive δ, we obtain for all x ∈ R,

0 < |x − a| < δ implies |K − K| = 0 < ε.

We conclude that lim K = K.

x→a
88 CHAPTER 4. LIMIT OF FUNCTIONS

Theorem 4.1.8 (Limit of Linear function) Let m and c be constant such that f (x) = mx + c
for all x ∈ R. Then
lim (mx + c) = ma + c.
x→a

ε
Proof. Let ε > 0. Choose δ = > 0. Let x ∈ R such that 0 < |x − a| < δ.
|m| + 1
|m|
We obtain by < 1 that
|m| + 1

|f (x) − (ma + c)| = |(mx + c) − (ma + c)| = |m(x − a)|

ε
= |m||x − a| < |m|δ = |m| · < 1 · ε = ε.
|m| + 1
Thus, f (x) → (ma + c) as x → a.

Theorem 4.1.9 Let E ⊆ R and f, g : E → R be functions and let a ∈ R be a limit point of E. If

f (x) = g(x) for all x ∈ E\{a} and f (x) → L as x → a,

then g(x) also has a limit as x → a, and

lim f (x) = lim g(x).

x→a x→a

Proof. Assume that f (x) = g(x) for all x ∈ E\{a} and f (x) → L as x → a.
Let ε > 0. There is a δ > 0.

∀x ∈ E, 0 < |x − a| < δ → |f (x) − L| < ε.

From 0 < |x − a| < δ, it implies that x ̸= a. So, f (x) = g(x) on the condition. We obtain

∀x ∈ E, 0 < |x − a| < δ → |g(x) − L| < ε.

Thus, g(x) → L as x → a.
x2 − 1
Example 4.1.10 Prove that f (x) = has a limit as x → 1.
x−1
Solution. We see that g(x) = x + 1. We have
x2 − 1 (x − 1)(x + 1)
f (x) = = = x + 1 = g(x) for all x ̸= 1
x−1 x−1
By Theorem 4.1.9, it follows that
x2 − 1
lim = lim f (x) = lim g(x) = lim (x + 1) = 2.
x→1 x − 1 x→1 x→1 x→1
4.1. LIMIT OF FUNCTIONS 89

Theorem 4.1.11 (Sequential Characterization of Limit (SCL)) Let E ⊆ R and f : E → R

be a function and let a ∈ R be a limit point of E. Then

lim f (x) = L exists

x→a

if and only if f (xn ) → L as n → ∞ for every sequence xn ∈ E\{a} that converges to a as n → ∞.

Proof. Assume that the limit of f (x) exists and equals to L and assume that a sequence xn ∈ E\{a}
that converges to a as n → ∞. Let ε > 0. There is a δ > 0 such that for all x ∈ E,

0 < |x − a| < δ implies |f (x) − L| < ε. (4.1)

There is an N ∈ N such that

n≥N implies |xn − a| < δ.

Since xn ̸= {a} and |xn − a| < δ for all n ≥ N , we obtain by (4.1)

|f (xn ) − L| < ε for all n ≥ N.

Coversely, assume that f (xn ) → L as n → ∞ for every sequence xn ∈ E\{a} that converges
to a as n → ∞. Suppose that f (x) does not converge to L as x approaches to a.
There is an ε0 > 0 such that

∀δ > 0, 0 < |x − a| < δ and |f (x) − L| ≥ ε0 . (4.2)

1 1
Choose δ = for all n ∈ N. Then 0 < |x − a| < . By Squeeze Theorem, xn → a as n → ∞.
n n
By assumption, f (xn ) → L as n → ∞, i.e., there N ∈ N

n≥N implies |f (x) − L| < ε0

which contradics (4.2). Therefore, f (x) converges to L as x approaches to a.

90 CHAPTER 4. LIMIT OF FUNCTIONS

Example 4.1.12 Prove that  ( )


 1
cos if x ̸= 0
f (x) = x


0 if x = 0

has no limit as x → 0.

Solution. Choose two sequence as follow

1
xn = →0 and f (xn ) = cos(2nπ) → 1,
2nπ
1
yn = →0 and f (yn ) = cos(2n − 1)π → −1.
(2n − 1)π

Then f (xn ) and f (yn ) converge to distinct limits. By SCL, we conclude that f has no limit as
x → 0.
Next, we will use the SCL together Theorems of limit for addition, mutiplication, scalar mul-
tiplication and quotient in order to proof Theorem 4.1.13.

Theorem 4.1.13 Let α ∈ R, E ⊆ R and f, g : E → R be functions and let a ∈ R be a limit point

of E. If f (x) and g(x) converge as x approaches a, then so do

(f + g)(x), (αf )(x), (f g)(x) and ( fg )(x).

In fact,

1. lim (f + g)(x) = lim f (x) + lim g(x)

x→a x→a x→a

2. lim (αf )(x) = α lim f (x)

x→a x→a

3. lim (f g)(x) = lim f (x) lim g(x)

x→a x→a x→a

( ) lim f (x)
f
4. lim (x) = x→a when the limit of g(x) is nonzero.
x→a g lim g(x)
x→a

Example 4.1.14 Show that lim x2 = a2 fo all a ∈ R.

x→a

Solution. Use Theorem 4.1.13 to give

lim x2 = lim x · x = lim x · lim x = a · a = a2 .

x→a x→a x→a x→a
4.1. LIMIT OF FUNCTIONS 91

Theorem 4.1.15 Suppose that E ⊆ R and f : E → R is a function. Let a ∈ R be a limit point of

E. Then,

lim |f (x)| = 0 if and only if lim f (x) = 0.

x→a x→a

Proof. Exercise.

Theorem 4.1.16 (Squeeze Theorem for Functions) Suppose that E ⊆ R and f, g, h : E → R

are functions. Let a ∈ R be a limit point of E. If

g(x) ≤ f (x) ≤ h(x) for all x ∈ E\{a},

and lim g(x) = lim h(x) = L, then the limit of f (x) exists, as x → a and
x→a x→a

lim f (x) = L.
x→a

Proof. Use SCL and the Squeeze Thorem (Theorem 2.2.1).

Corollary 4.1.17 Suppose that E ⊆ R and f, g : E → R are functions. Let a ∈ R be a limit point
of E and M > 0. If

|g(x)| ≤ M for all x ∈ E\{a} and lim f (x) = 0,

x→a

then
lim f (x)g(x) = 0.
x→a

Proof. Assume that |g(x)| ≤ M for all x ∈ E\{a} and lim f (x) = 0.
x→a
Case f (x) = 0. Then f (x)g(x) = 0. It follows that lim f (x)g(x) = 0.
x→a
Case f (x) ̸= 0. Then |f (x)| > 0. So, lim M |f (x)| = 0. We obtain
x→a

0 ≤ |g(x)f (x)| = |g(x)||f (x)| ≤ M |f (x)|.

By the Squeeze Theorem for Functions, it imlies that lim |g(x)f (x)| = 0.
x→a
From Theorem 4.1.15, we conclude that lim f (x)g(x) = 0.
x→a
92 CHAPTER 4. LIMIT OF FUNCTIONS
( )
1
Example 4.1.18 Show that lim x cos =0
x→0 x
Solution. By property of sine,
( )
1
cos ≤ 1 for all x ̸= 0.
x
( )
1
We have lim x = 0. By Corollary 4.1.17, lim x cos = 0.
x→0 x→0 x

Theorem 4.1.19 (Comparision Theorem for Functions) Suppose that E ⊆ R and

f, g : E → R are functions. Let a ∈ R be a limit point of E. If f and g have a limit as x approaches
a and
f (x) ≤ g(x), x ∈ E\{a},

then
lim f (x) ≤ lim g(x).
x→a x→a

Proof. Use SCL together the Comparison Theorem (Theorem 2.2.12), we will this theorem.
4.1. LIMIT OF FUNCTIONS 93

Exercises 4.1

1. Use Definition 4.1.1, prove that each of the following limit exists.

1.1 lim x2 = 1 1.3 lim x3 + 1 = 0.

x→1 x→−1

x−1
1.2 lim x2 − x + 1 = 3 1.4 lim = −1
x→2 x→0 x + 1

2. Decide which of the following limit exist and which do not.

( ) ( ) ( )
1 1 1
2.1 lim sin 2.2 lim x sin 2.3 lim tan
x→0 x x→0 x x→0 x

3. Evaluate the following limit using result from this section.

( )
x2 + x − 2 1
3.1 lim 3.3 lim x sin
x→1 x3 − x x→0 x2
√ ( )
3
π − x2 2 1
3.2 lim√ 3.4 lim x cos
x→ π x + π x→0 x
( )
1
n
4. Prove that lim x sin exists for all n ∈ N.
x→0 x
5. Show that lim xn = an fo all a ∈ R and n ∈ N.
x→a

6. Prove that lim |f (x)| = 0 if and only if lim f (x) = 0.

x→a x→a

7. Prove Squeeze Theorem for Functions.

8. Prove Comparision Theorem for Functions.

9. Suppose that f is a real function.

9.1 Prove that if

lim f (x) = L
x→a

exists, then |f (x)| → |L| as x → a.

9.2 Show that there is a function such that as x → a, |f (x)| → |L| but the limit of f (x)
does not exist.
94 CHAPTER 4. LIMIT OF FUNCTIONS

4.2 One-sided limit

√
What is the limit of f (x) := x − 1 as x → 1.

4
√
3 y= x−1
2
1

X
0 1 2 3 4 5 6 7 8 9

A reasonable answer is that the limit is zero. This function, however, does not satisfy Definition
4.1.1 because it is not defined on an OPEN interval containg a = 1. Indeed, f is defined only for
x ≥ 1. To handle such situations, we introduce one-sided limits.

Definition 4.2.1 Let a ∈ R.

1. A real function f said to converge to L as x approaches a from the right if and only if
f defined on some interval I with left endpoint a and every ε > 0 there is a δ > 0 such that
a + δ ∈ I and for all x ∈ I,

a < x < a + δ implies |f (x) − L| < ε.

In this case we call L the right-hand limit of f at a, and denote it by

f (a+ ) := L =: lim+ f (x).

x→a

Y
y = f (x)

L+ε

X
a a+δ
4.2. ONE-SIDED LIMIT 95

2. A real function f said to converge to L as x approaches a from the left if and only if f
defined on some interval I with right endpoint a and every ε > 0 there is a δ > 0 such that
a + δ ∈ I and for all x ∈ I,

a − δ < x < a implies |f (x) − L| < ε.

In this case we call L the left-hand limit of f at a, and denote it by

f (a− ) := L =: lim− f (x).

x→a

Y
y = f (x)

L
L−ε

X
a−δ a

√
Example 4.2.2 Prove that lim+ x − 1 = 0.
x→1

Proof. Let ε > 0. Choose δ = ε2 > 0. Let x > 1 such that 0 < x − 1 < δ. We obtain
√ √ √
|f (x) − 0| = | x − 1 − 0| = x − 1 < δ = ε.
√
Thus, x − 1 → 0 as x → 1+ .
|x|
Example 4.2.3 If f (x) = , prove that f has one-sided limit at a = 0 but lim f (x) = 0 DNE.
x x→0

Solution. Let ε > 0. We can choose any δ > 0. Let x ∈ R\{0} such that −δ < x < 0.
Then |x| = −x. We obtain
|x| −x
|f (x) − 0| = − (−1) = − (−1) = | − 1 + 1| = 0 < ε.
x x
Thus, lim− f (x) = −1. Similarly, lim− f (x) exists and equals 1.
x→0 x→0
Choose two sequence as follow
1
xn = →0 and f (xn ) = 1 → 1,
n
1
yn = − → 0 and f (yn ) = −1 → −1.
n
Then f (xn ) and f (yn ) converge to distinct limits. By SCL, we conclude that f has no limit as
x → 0.
96 CHAPTER 4. LIMIT OF FUNCTIONS

Theorem 4.2.4 Let f be a real function. Then the limit

lim f (x)
x→a

exists and equals to L if and only if

L = lim+ f (x) = lim− f (x).

x→a x→a

Proof. Assume that f (x) → L as x → a. Let ε > 0. There is a δ > 0 such that

0 < |x − a| < δ and |f (x) − L| < ε. (4.3)

If a < x < a + δ, it satisfies (4.3) which implies |f (x) − L| < ε. Thus, lim+ f (x) = L.
x→a
If a − δ < x < a, it satisfies (4.3) which implies |f (x) − L| < ε. Thus, lim− f (x) = L.
x→a
Conversely, assume that L = lim+ f (x) = lim− f (x). Let ε > 0. There are δ1 , δ2 > 0 such that
x→a x→a

a < x < a + δ1 → |f (x) − L| < ε (4.4)

and

a − δ2 < x < a → |f (x) − L| < ε. (4.5)

Choose δ = min{δ1 , δ1 }. If |x − a| < δ, it satisfies (4.4) and (4.5) which imply

|f (x) − L| < ε.

Therefore, lim f (x) = L.

x→a


x + 1 if x ≥ 0
Example 4.2.5 Use Theorem 4.2.4 to show that f (x) = has limit at a = 0.

2x + 1 if x < 0

Solution. We see that

lim f (x) = lim+ (x + 1) = 1 = lim− (2x + 1) = lim− f (x).

x→0+ x→0 x→0 x→0

By Theorem 4.2.4, we conclude that lim f (x) = 1

x→0
4.2. ONE-SIDED LIMIT 97

Exercises 4.2

1. Use definitons to prove that lim+ f (x) exists and equal to L in each of the following cases.
x→a

1.1 f (x) = 2x2 + 1, a = 1, and L = 3.

x−1
1.2 f (x) = , a = 1, and L = 1.
|1 − x|
√
1.3 f (x) = 3x − 5, a = 2, and L = 1.

2. Use definitons to rove that lim− f (x) exists and equal to L in each of the following cases.
x→a

2.1 f (x) = 1 + x2 , a = 1, and L = 2.

√
2.2 f (x) = 1 − x2 , a = 1, and L = 0.
1 − x2
2.3 f (x) = , a = 1, and L = 0.
1+x
3. Evauate the following limit when they exist.

x+1 3.3 lim+ (x2 + 1) sin x

3.1 lim+
x→0 x2 − 2 x→π

x3 − 3x + 2 cos x
3.2 lim− 3.4 lim
x→1 x3 − 1
π−
x→ 2 1 − sin x
√ √
1 − cos x 2
4. Prove that → as x → 0+ .
sin x 2
5. Determine whether the following functions are limit at a.


3x + 1 if x ≥ 1
5.1 f (x) = and a=1

x + 3 if x < 1


2 − 2x if x ≥ 0
5.2 f (x) = √ and a=0

 1−x if x < 0

6. Suppose that f : [0, 1] → R and f (a) = lim f (x) for all x ∈ [0, 1]. Prove that
x→a

f (q) = 0 for all q ∈ Q ∩ [0, 1] if and only if f (x) = 0 for all x ∈ [0, 1].
98 CHAPTER 4. LIMIT OF FUNCTIONS

4.3 Infinite limit

The definition of limit of real functions can be expanded to include extended real numbers.

Definition 4.3.1 Let E ⊆ R and f : E → R be a function.

1. We say that f (x) → L as x → ∞ if and only if there exists a c > 0 such that (c, ∞) ⊆ E
and for every ε > 0, there is an M ∈ R such that

x>M implies |f (x) − L| < ε.

In this case we shall write lim f (x) = L.

x→∞

2. We say that f (x) → L as x → −∞ if and only if there exists a c > 0 such that (−∞, −c) ⊆ E
and for every ε > 0, there is an M ∈ R such that

x<M implies |f (x) − L| < ε.

In this case we shall write lim f (x) = L.

x→−∞

1
Example 4.3.2 Prove that lim = 0.
x→∞ x

1
Proof. Let ε > 0. Choose M = > 0. If x > M > 0, it implies
ε
1 1 1
−0 = < = ε.
x x M
1
We conclude that lim = 0.
x→∞ x
x−1
Example 4.3.3 Prove that lim exists and equals to 1.
x→∞ x + 1

2
Proof. Let ε > 0. Choose M = > 0. If x > M > 0, it follows that x + 1 > x > M . So,
ε
1 1
< . We obtain
x+1 M
x−1 −2 1 2
−1 = =2· < = ε.
x+1 x+1 x+1 M
x−1
We conclude that lim = 1.
x→∞ x + 1
4.3. INFINITE LIMIT 99

1
Example 4.3.4 Prove that lim = 0.
x→∞ x2 +1
1
Proof. Let ε > 0. Choose M = √ > 0. If x > M > 0, it follows that x2 > M 2 > 0. So,
ε
1 1
< 2 . We obtain
x2 M
1 1 1 1
2
−0 = 2 < 2 < 2 = ε.
x +1 x +1 x M
1
We conclude that lim = 0.
x→∞ x2 + 1

1
Example 4.3.5 Prove that lim = 0.
x→−∞ x

1
Proof. Let ε > 0. Choose M = − < 0. If x < M < 0, it implies −x > −M > 0. We obtain
ε
1 1 1
−0 = < = ε.
x −x −M
1
We conclude that lim = 0.
x→−∞ x

x
Example 4.3.6 Prove that lim = 1.
x→−∞ x + 1

1 1
Proof. Let ε > 0. Choose M = −1 − . Then M + 1 = − < 0. If x < M , it implies
ε ε
1 1
1 + x < 1 + M < 0. So, 0 < − <− . We obtain
x+1 M +1
x 1 1 1
−1 = = < = ε.
x+1 |x + 1| −(x + 1) −(M + 1)
x
We conclude that lim = 1.
x→−∞ x + 1
100 CHAPTER 4. LIMIT OF FUNCTIONS

Definition 4.3.7 Let E ⊆ R and f : E → R be a function.

1. We say that f (x) → +∞ as x → a if and only if there is an open interval I containing a

such that I\{a} ⊂ E and for every M > 0 there is a δ > 0 such that

0 < |x − a| < δ implies f (x) > M .

In this case we shall write lim f (x) = +∞.

x→a

2. We say that f (x) → −∞ as x → a if and only if there is an open interval I containing a

such that I\{a} ⊂ E and for every M < 0 there is a δ > 0 such that

0 < |x − a| < δ implies f (x) < M .

In this case we shall write lim f (x) = −∞.

x→a

Obviousl modification define f (x) → ±∞ as x → a+ and x → a− , and f (x) → ±∞ as x → ±∞.

1
Example 4.3.8 Prove that lim = +∞.
x→0 |x|

1
Proof. Let M > 0. Choose δ = > 0. If 0 < |x| < δ, it follows
M
1 1
> = M.
|x| δ
1
Thus, lim = +∞.
x→0 |x|

x
Example 4.3.9 Prove that lim+ = −∞.
x→1 1−x
1 1 1 1 1
Proof. Let M < 0. Choose δ = − > 0. If 0 < x − 1 < δ, it follows < . So, <− .
M δ x−1 1−x δ
We obatin
x 1 1 1
= −1 + <0+ < − = M.
1−x 1−x 1−x δ
x
Thus, lim+ = −∞.
x→1 1−x
4.3. INFINITE LIMIT 101

Exercises 4.3

1. Use definitons to prove that lim+ f (x) exists and equal to L in each of the following cases.
x→a

1
1.1 f (x) = , a = 3, and L = +∞.
x−3
1
1.2 f (x) = − , a = 0, and L = −∞.
x

2. Use definitons to prove that lim− f (x) exists and equal to L in each of the following cases.
x→a

x
2.1 f (x) = , a = 2, and L = −∞.
−4
x2
1
2.2 f (x) = , a = 1, and L = +∞.
1 − x2

3. Use definition to prove that the follwing limits

2x + 1 x
3.1 lim =2 3.4 lim = +∞
x→∞ x + 1 x→2 |x − 2|

1−x 1 x+1
3.2 lim =− 3.5 lim+ = +∞
x→−∞ 2x + 1 2 x→2 x − 2
2
2x + 1 x+1
3.3 lim = −2 3.6 lim− = −∞
x→∞ 1 − x2 x→2 x − 2

4. Evauate the following limit when they exist.

3x2 − 13x + 4 4.4 lim arctan x

4.1 lim x→∞
x→∞ 1 − x − x2
x2 + x + 2 sin x
4.2 lim 3 4.5 lim
x→∞ x − x − 2 x→∞ x2

x3 − 1
4.3 lim 2 4.6 lim x2 sin x
x→−∞ x + 2 x→−∞

sin(x + 3) − sin 3
5. Prove that converges to 0 as x → ∞.
x

6. Prove the following comparision theorems for real functions.

6.1 If f (x) ≥ g(x) and g(x) → ∞ as x → a, then f (x) → ∞ as x → a.

6.2 If f (x) ≤ g(x) ≤ h(x) and L = lim f (x) = lim h(x), then g(x) → L as x → ∞.
x→∞ x→∞
102 CHAPTER 4. LIMIT OF FUNCTIONS

7. Recall that a polynomial of degree n is a functon of the form

P (x) = an xn + an−1 xn−1 + · · · + a1 x + a0

where aj ∈ R for j = 0, 1, ..., n and an ̸= 0.

7.1 Prove that lim xn = an for n = 0, 1, 2, ...

x→a

7.2 Prove that if P is a polynomial, then

lim P (x) = P (a)

x→a

for every a ∈ R.
P (x)
7.3 Suppose that P is a polynomial and P (a) > 0. Prove that → ∞ as x → a+ ,
x−a
P (x)
→ −∞ as x → a− , but
x−a
P (x)
lim
x→a x−a
does not exist.

8. Cauchy. Suppose that f : N → R. If

lim f (n + 1) − f (n) = L,
n→∞

f (n)
prove that lim exists and equals L.
n→∞ n
Chapter 5

Continuity on R

5.1 Continuity

Definition 5.1.1 Let E be a nonempty subset of R and f : E → R.

f is said to be continuous at a point a ∈ E if and only if given ε > 0 there is a δ > 0 such that

|x − a| < δ and x ∈ E imply |f (x) − f (a)| < ε.

Example 5.1.2 Let f (x) = 2x − 1 where x ∈ R. Prove that f is continuous at x = 1.

ε
Proof. Let ε > 0. Choose δ = > 0. Let x ∈ R such that |x − 1| < δ. We obtain
2

|f (x) − f (1)| = |(2x − 1) − 1| = |2(x − 1)| = 2|x − 1| < 2δ = ε.

Thus, f is continuous at x = 1.

Example 5.1.3 Let f (x) = x2 where x ∈ R. Prove that f is continuous at x = 2.

{ ε}
Proof. Let ε > 0. Choose δ = min 1, . Let x ∈ R such that |x − 2| < δ.
5
We obtain |x| − 2 < |x − 2| < 1. It follows |x| < 3. So,

ε
|f (x) − f (2)| = |x2 − 4| = |x + 2||x − 2| < (|x| + 2)δ < (3 + 2) = ε.
5

Thus, f is continuous at x = 2.
104 CHAPTER 5. CONTINUITY ON R
√
Example 5.1.4 Let f (x) = x where x ∈ (0, ∞). Prove that f is continuous at 1.

Proof. Let ε > 0. Choose δ = ε. Let x ∈ (0, ∞) such that |x − 1| < δ.

√ 1
Since x + 1 > 1, √ < 1. We obtain
x+1
√
|f (x) − f (1)| = | x − 1|
√
√ x+1 x−1 1
= ( x − 1) · √ = √ = |x − 1| · √ < δ · 1 = ε.
x+1 x+1 x+1

Thus, f is continuous at x = 2.

Example 5.1.5 Let f (x) = 3 − x2 where x ∈ [−1, 2] ∪ {3}. Prove that f is continuous at x = 3

Proof. Let ε > 0. Choose δ = 0.5. Let x ∈ [−1, 2] ∪ {3} such that |x − 3| < δ = 0.5. It follows
x = 3. We obtain

|f (x) − f (3)| = |f (3) − f (3)| = 0 < ε.

Thus, f is continuous at x = 3.

Example 5.1.6 Prove that the function


 |x|

if x ̸= 0
f (x) = x

0 if x = 0

is discontinuous at 0.

Proof. Suppose that f is continuous at 0. Given ε = 1. There is a δ > 0 such that

|x| < δ and x ∈ R imply |f (x)| = |f (x) − f (0)| < 1. (5.1)

For 0 < x < δ, we obtain by (5.1) such that

x |x|
1= = = |f (x)| < 1
x x

It is imposible. Thus, f is discontinuous at 0.

5.1. CONTINUITY 105

Theorem 5.1.7 Let I be an open interval that contain a point a and f : I → R. Then

f is continuous at a ∈ I if and only if f (a) = lim f (x).

x→a

Proof. Let I = (c, d) such that contain a point a.

d−a

( )
c a d
a−c

Set δ0 = min{a − c, d − a}. Choose δ < δ0 . Then |x − a| < δ implies x ∈ I.

Therefore, conditions

|x − a| < δ and x ∈ I imply |f (x) − f (a)| < ε

is identical to

0 < |x − a| < δ implies |f (x) − f (a)| < ε.

We conclude that f is continuous at a ∈ I if and only if f (a) = lim f (x).

x→a
( )
1
Example 5.1.8 Let f (x) = x cos where x ̸= 0. If f is continuous at 0, what is f (0) defined?
x
Solution. Use Example 4.1.18 and Theorem 5.1.7 in order to define
( )
1
f (0) = lim x cos = 0.
x→0 x
Thus, we define f (0) = 0 that makes f be continuous at 0.


ax + 1 if x ≥ 1
Example 5.1.9 Find a such that the function f (x) = is continuous at 1.

2x + 3 if x < 1
Solution. From f is continuous at 1, we obtain

f (1) = lim f (x) = lim f (x)

x→1+ x→1−

a+1 = lim (ax + 1) = lim (2x + 3)

x→1+ x→1−

a+1 = a+1 = 5
Hence, a = 4.
106 CHAPTER 5. CONTINUITY ON R

Theorem 5.1.10 Suppose that E is a nonempty subset of R, a ∈ E, and f : E → R. Then the

following statements are equivalent:

1. f is continuous at a ∈ E.

2. If xn converges to a and xn ∈ E, then f (xn ) → f (a) as n → ∞.

Proof. The proof Theorem is complete by Theorem 5.1.7 and SCL.

√
n
Example 5.1.11 Use Theorem 5.1.10 to find lim .
n→∞ n+1
√
Solution. Let f (x) = x where x ∈ (0, ∞). By Example 5.1.4, f is continuos at 1. Set

n
xn = .
n+1

Then lim xn = 1 by Example 2.1.6. By Theorem 5.1.7, it implies that

n→∞
√
n
f (xn ) = → f (1) = 1.
n+1

Next, we will use Theorem 5.1.10 together Theorems of limit for addition, mutiplication, scalar
multiplication and quotient in order to proof Theorem 5.1.12.

Theorem 5.1.12 Let E be a nonempty subset of R and f, g : E → R and α ∈ R. If f, g are

continuous at a point a ∈ E, then so are

f + g, f g and αf

Moreover, f /g is continuous at a ∈ E when g(a) ̸= 0.

5.1. CONTINUITY 107

CONTINUITY OF COMPOSITION.

Definition 5.1.13 Suppose that A and B are subsets of R and that f : A → R and g : B → R.
If {f (x) : x ∈ A} ⊆ B, then the composition of g with f is the function

(g ◦ f )(x) := g(f (x)), x ∈ A.

R R
A
B

f g
x f (x) g(f (x))

g◦f

Theorem 5.1.14 Suppose that A and B are subsets of R and that f : A → R and g : B → R with
{f (x) : x ∈ A} ⊆ B. If f is continuous at a ∈ A and g is continuous at f (a) ∈ B, then

g ◦ f is continuous at a ∈ A

and moreover,
( )
lim (g ◦ f )(x) = g lim f (x) .
x→a x→a

Proof. Assume that f is continuous at a ∈ A and g is continuous at f (a) ∈ B.

Let ε > 0. There is a δ1 > 0 such that

|y − f (a)| < δ1 and y ∈ B imply |g(y) − g(f (a))| < ε. (5.2)

There is a δ2 > 0 such that

|x − a| < δ2 and x ∈ A imply |f (x) − f (a)| < δ1 . (5.3)

For each x ∈ A such that |x − a| < δ2 , it implies |f (x) − f (a)| < δ1 . Set y = f (x).
We obtain by (5.2) that |g(f (x)) − g(f (a))| < ε. We conclude that g ◦ f is continuous at a ∈ A.
108 CHAPTER 5. CONTINUITY ON R
√
Example 5.1.15 Show that lim 2x − 1 exists and equals to 1.
x→1
√
Solution. Let g(x) = x and f (x) = 2x − 1. Then f is continous at 1 and g is continuous at
f (1) = 1. By Theorem 5.1.14,
( ) ( )
lim (g ◦ f )(x) = g lim f (x) = g lim (2x − 1) = g(1) = 1.
x→1 x→1 x→1

CONTINUITY ON A SET.

Definition 5.1.16 Let E be a nonempty subset of R and f : E → R.

f is said to be continuous on E if and only if f is continuous at every a ∈ E.

Note that if f is continuous on E, then f is continuous on nonempty subset of E.

Example 5.1.17 Show that f (x) = x2 is continuous on R.

{ }
ε
Proof. Let a ∈ R and ε > 0. Choose δ = min 1, .
2|a| + 1
Let x ∈ R such that |x − a| < δ. We obtain |x| − |a| < |x − a| < 1. It follows

|x| < 1 + |a|.

We obtain

|f (x) − f (a)| = |x2 − a2 | = |x + a||x − a|

ε
< (|x| + |a|)δ < (|a| + 1 + |a|) = ε.
2|a| + 1
Thus, f is continuous on R.

Theorem 5.1.18 (Continuity of linear function) Let m and c be constants and let

f (x) = mx + c where x ∈ R.

Prove that f is continuous on R

ε
Proof. Let a ∈ R and ε > 0. Choose δ = > 0. Let x ∈ R such that |x − a| < δ. We obtain
|m| + 1

|f (x) − f (a)| = |(mx + c) − (ma + c)| = |m||x − a|

ε
< |m|δ ≤ |m| · < 1 · ε = ε.
|m| + 1
Thus, f is continuous at R.
5.1. CONTINUITY 109

Example 5.1.19 Show that h(x) = (3x + 1)2 is continuous on R.

Solution. Let f (x) = x2 and g(x) = 3x + 1. By Example 5.1.17 and Theorem 5.1.18, f and g are
continuous on R. We conclude by Theorem 5.1.14 that

h(x) = f ◦ g(x) = (3x + 1)2 is continuous on R.

Example 5.1.20 Prove that 


2x + 4 if x > −1
f (x) =

3x + 5 if x ≤ −1
is continuous on R.

Solution. We see that f is a linear function on (−1, ∞) ∪ (−1, ∞). By Continuity of Linear
function, f is continuous on (−1, ∞) ∪ (−1, ∞). From

f (−1) = 2 = lim + (3x + 5) = lim − (2x + 4),

x→−1 x→−1

it follows that f is continuous at −1. We conclude that f is is continuous on R.



ax + 1 if x ≥ 2
Example 5.1.21 Find a such that the function f (x) = is continuous on R.

x + a if x < 2

Solution. From f is continuous at 2, we obtain

f (2) = lim f (x) = lim f (x)

x→2+ x→2−

2a + 1 = lim (ax + 1) = lim (x + a)

x→2+ x→2−

2a + 1 = 2a + 1 = 2 + a.

Hence, a = 1.
110 CHAPTER 5. CONTINUITY ON R

Exercises 5.1

1. Use definition to prove that f is continuous at a.

1.1 f (x) = x2 + 1 and a = 1. 1

1.3 f (x) = and a = 1.
x
x
1.2 f (x) = x3 and a = −1. 1.4 f (x) = and a = 2.
x2 +1

2. Determine whether the following functions are continuous at a.



1 − 2x if x ≥ 1
2.1 f (x) = and a=1

2 − 3x if x < 1


x2 − 1 if x ≥ 0
2.2 f (x) = √ and a=0

 1−x if x < 0

3. Use definition to prove that f is continuous at E.

3.1 f (x) = x3 and E = R.

√
3.2 f (x) = 1 − x and E = (−∞, 1).
1
3.3 f (x) = and E = R.
x2 +1

4. Use limit theorem to show that the following function are continuous on [0, 1].

√
4.1 f (x) = 3x2 + 1 4.3 f (x) = 2−x
1−x 1
4.2 f (x) = 4.4 f (x) =
1+x x2 +x−6




 ax + 3 if x ≤ 1


5. Find a and b such that the function f (x) = x + b if 1 < x ≤ 2 is continuous on R.





2ax − 2 if x > 2

6. If f : [a, b] → R is continuous, prove that sup |f (x)| is finite.

x∈[a,b]

7. Show that there exist nowhere continuous functions f and g whose sum f + g is continuous
on R. Show that the same is ture for product of functions.
5.1. CONTINUITY 111

8. Let 

cos( 1 )
x
if x ̸= 0
f (x) =

1 if x = 0

is continuous on (−∞, 0) and (0, ∞), discontinuous at 0, and neither f (0+ ) nor f (0− ) exists.

8.1 Prove that f is continuous on (−∞, 0) and (0, ∞) discontinuous at 0.

8.2 Suppose that g : [0, π2 ] → R is continuous on (0, π2 ) and that there is a positive constant
C > 0 such that
√
|g(x)| ≤ C x for all x ∈ (0, π2 ),

Prove that f (x)g(x) is continuous on [0, π2 ].

9. Suppose that a ∈ R, that I is an open interval containing a, that, f, g : I → R, and that f

is continuous at a.

9.1 Prove that g is continuous at a if and only if f + g is continuous at a.

9.2 Make and prove an analogous atstement for the product f g. Show by example that
hypothesis about f added cannot be dropped.

10. Let f : A → R be a continuous function. Suppose that E ⊆ A and is open. Determine

whether {f (x) : x ∈ E} is open.

11. Let f (x) = xn where n ∈ N. Prove that f is continuous on R

12. Suppose that f : R → R satisfies f (x + y) = f (x) + f (y) for each x, y ∈ R.

12.1 Show that f (nx) = nf (x) for all x ∈ R and n ∈ Z.

12.2 Prove that f (qx) = qf (x) for all x ∈ R and q ∈ Q.

12.3 Prove that f is continuous at 0 if and only if f is continuous on R.

12.4 Prove that f is continuous at 0, then there is an m ∈ R such that f (x) = mx for all
x ∈ R.
ln(x + 1) 1
13. Assume that lim = 1 and f (x) = ex is continuous on R. Show that lim (1+x) x = e.
n→0 x x→0
112 CHAPTER 5. CONTINUITY ON R

5.2 Intermediate Value Theorem

Definition 5.2.1 Let E be a nonempty subsets of R. A function f : E → R is said to be bounded

on E if and only if there is an M > 0 such that

|f (x)| ≤ M for all x ∈ E

For a example f (x) = sin x, by sine property that

| sin x| ≤ 1 for all x ∈ R.

So, f is bounded by 1 on R.
Next, let f : I → R be a function. We define

sup f (x) := sup{f (x) : x ∈ I}

x∈I

inf f (x) := inf{f (x) : x ∈ I}

x∈I

For example sup x2 = 1 and inf x2 = 0.

x∈[0,1) x∈[0,1)

Theorem 5.2.2 (Extreme Value Theorem (EVT)) If I is a closed, bounded interval and
f : I → R is continuous on I, then f is bounded on I. Moreover, if

M = sup f (x) and m = inf f (x),

x∈I x∈I

then there exist point xm , xM ∈ I such that

f (xM ) = M and f (xm ) = m.

Proof. Suppose that f is not bounded in I. Then there exist xn ∈ I such that

|f (xn )| > n for n ∈ N (5.4)

Since I is bounded, we know by the Bolzano-Weierstrass Theorem that {xn } has a convergent
subsequence, say xnk → a as k → ∞. Since I is closed, we also know by the Comparison Theorem
that a ∈ I and f (a) ∈ R. By (5.4), we obtain

f (a) = lim |f (xnk )| > lim nk ≥ lim k = ∞

k→∞ k→∞ k→∞

which contradics f (a) ∈ R. Thus, f is bounded in I.

We will prove that M and m are finite real numbers. Suppose that
5.2. INTERMEDIATE VALUE THEOREM 113

f (x) < M = sup f (x) for all x ∈ I.

x∈I

Then the function

1
g(x) = is continuous on I.
M − f (x)
So, g is bounded on I. There i a C > 0 such that |g(x)| = g(x) ≤ C for all x ∈ I. It follows that

1
f (x) ≤ M − .
C

We obtain
1
M = sup f (x) ≤ M − < M.
x∈I C
It is imposible. Thus, there is an xM ∈ I such that f (xM ) = M . A similar argument proves that
there is an xm ∈ I such that f (xm ) = m.

Lemma 5.2.3 (Sign-Preserving Property) Let f : I → R where I is open. If f is continuous

at a point x0 ∈ I and f (x0 ) > 0, then there are positive numbers ε and δ such that

|x − x0 | < δ implies f (x) > ε.

Proof. Assume that f is continuous at a point x0 ∈ I and f (x0 ) > 0.

f (x0 )
Given ε = . There is a δ > 0 such that
2
f (x0 )
|x − x0 | < δ and x ∈ I imply |f (x) − f (x0 )| < .
2
It follows that
f (x0 ) f (x0 )
− < f (x) − f (x0 ) <
2 2
f (x0 ) 3f (x0 )
< f (x) <
2 2
f (x0 )
Thus, f (x) > = ε.
2
114 CHAPTER 5. CONTINUITY ON R

Theorem 5.2.4 (Intermediate Value Theorem (IVT)) Let f : [a, b] → R be continuous.

If y0 lies between f (a) and f (b), then

there is an x0 ∈ (a, b) such that f (x0 ) = y0 .

f (b) y = f (x)
y0

f (a)
X
a x0 b

Proof. We may suppose that f (a) < y0 < f (b). Consider

E = {x ∈ [a, b] : f (x) < y0 }.

Since a ∈ E and E ⊆ [a, b], E is a nonempty bounded subset of R. Thus, by the Completeness
Axiom, x0 = sup E is a finite real number. Since y0 is equals neither f (a) nor f (b), x0 cannot
equal to a or b. Hence, x0 ∈ (a, b).
It remains to show that f (x0 ) = y0 . By Theorem 2.2.5, there is a sequence xn ∈ E such that

xn → sup E = x0 as n → ∞.

Since f is continuous and the definition of E, by the Comparison Theorem and Theorem 5.1.10
we obtain
f (x0 ) = lim f (xn ) ≤ y0 .
n→∞

Finally, we will prove that f (x0 ) = y0 , suppose to the contrary that f (x0 ) < y0 . Set

g(x) = y0 − f (x) where x ∈ E.

Then g is continuous and g(x0 ) > 0. Hence, by Lemma 5.2.3, we can choose positive numbers ε
and δ such that

|x − x0 | < δ implies g(x) > ε > 0.

For any x, it satisfies x0 < x < x0 + δ also satisfies y0 − f (x) = g(x) > 0 or f (x) < y0 which
contradics the fact that x0 = sup E.
5.2. INTERMEDIATE VALUE THEOREM 115

Corollary 5.2.5 Let f : [a, b] → R be continuous.

1. If f (a) > 0 and f (b) < 0, then there is an c ∈ (a, b) such that f (c) = 0.

2. If f (a) < 0 and f (b) > 0, then there is an c ∈ (a, b) such that f (c) = 0.

Proof. It is obviously by the IVT.

Example 5.2.6 Show that there is a real number such that x2 = x + 1.

Solution. Let f (x) = x2 − x − 1. Then f (1) = −1 < 0 and f (2) = 2 > 0.

Since f is continuous on (1, 2), we obatin by Corollary 5.2.5 that there is an c ∈ (1, 2) such that

c2 − c − 1 = f (c) = 0.

Thus, there exists a real number c such that c2 = c + 1.

Example 5.2.7 Prove that ln x = 3 − 2x has at least one real root and find the approximate root
to be the midpont of an interval [a, b] of length 0.01 that contain a root.

Solution. Let f (x) = ln x + 2x − 3. Consider each values of f (x) by calculator

x f (x) Interval Length of Interval

2 1.6931
1 −1 [1, 2] 1
1.4 0.1365
1.3 −0.1376 [1.3, 1.4] 0.1
1.35 0.00010
1.34 −0.02733 [1.34, 1.35] 0.01

Since f is continuous on (1.34, 1.35), we obatin by Corollary 5.2.5 that there is an c ∈ (1.34, 1.35)
such that
ln c + 2c − 3 = f (c) = 0.

Thus, there exists a real number c such that ln c = 3 − 2c.

We may approximate the root by choosing midpoint c = 1.345 of (1.34, 1.35). It follows that
f (c) = −0.0136 which has error 0.01.
116 CHAPTER 5. CONTINUITY ON R

Exercises 5.2

For these exercise, assume that sin x, cos x and ex are continuous on R and ln x is continuous on
R+ .
1. For each of the following, prove that there is at least one x ∈ R that satisfies the given
equation.

1.1 x3 + x = 3 1.6 ex = x2

1.2 x3 + 2 = 2x 1.7 x ln x = 1

1.3 x4 + x3 − 2 = 0 1.8 sin x = ex

1.4 x5 + x + 1 = 0 1.9 cos x = x2

1.5 2x = 2 − x 1.10 ex = cos x + 1

2. Prove that the follwing equations have at least one real root and find the approximate root
to be the midpont of an interval [a, b] of length ℓ that contain a root.

2.1 x3 + x = 1 and ℓ = 0.001 2.4 cos x = x and ℓ = 0.01

2.2 2x = x3 and ℓ = 0.01 2.5 sin x + x = 1 and ℓ = 0.001

2.3 ln x + x = 2 and ℓ = 0.001 2.6 xex = cos x and ℓ = 0.01

3. Suppose that f is a real-value function of a real variable. If f is continuous at a with

f (a) < M for some M ∈ R, prove that there is an open interval I containing a such that

f (x) < M for all x ∈ I.

4. If f : R → R is continuous and

lim f (x) = lim f (x) = ∞,

x→∞ x→−∞

prove that f has a minimum on R; i.e., there is an xm ∈ R such that

f (xm ) = inf f (x) < ∞.

x∈R
5.3. UNIFORM CONTINUITY 117

5.3 Uniform continuity

Definition 5.3.1 Let E be a nonempty subset of R and f : E → R. Then f is said to be

uniformly continuous on E if and only if for every ε > 0 there is a δ > 0 such that

|x − a| < δ and x, a ∈ E imply |f (x) − f (a)| < ε.

Example 5.3.2 Prove that f (x) = x is uniformly continuous on (0, 1).

Solution. Let ε > 0. Choose δ = ε. Let x, a ∈ (0, 1) such that |x − a| < δ. We obtain

|f (x) − f (a)| = |x − a| < δ = ε.

Thus, f is uniformly continuous on (0, 1).

Example 5.3.3 Prove that f (x) = x2 is uniformly continuous on (0, 1).

ε
Solution. Let ε > 0. Choose δ = . Let x, a ∈ (0, 1) such that |x − a| < δ.
2
Then |x + a| ≤ |x| + |a| < 1 + 1 = 2. We obtain

|f (x) − f (a)| = |x2 − a2 | = |x + a||x − a| < 2δ = ε.

Thus, f is uniformly continuous on (0, 1).

Theorem 5.3.4 (Uniform continuity of linear function) A Linear function is uniformly

continuous on R.

Proof. Let m, c be contants and f (x) = mx + c where x ∈ R.

ε
Let ε > 0. Then |m| + 1 > 0. Choose δ = > 0. Let x ∈ R such that 0 < |x − a| < δ.
|m| + 1
|m|
We obtain by < 1 that
|m| + 1

|f (x) − f (a)| = |(mx + c) − (ma + c)| = |m(x − a)| = |m||x − a|

ε
< |m|δ = |m| · < 1 · ε = ε.
|m| + 1

Thus, f is uniformly continuous on R.

118 CHAPTER 5. CONTINUITY ON R

Example 5.3.5 Prove that f (x) = x2 is not uniformly continuous on R.

Solution. Suppose that f is uniformly continuous on R.

Given ε = 1. There is a δ > 0 such that

|x − a| < δ and x, a ∈ R imply |f (x) − f (a)| < 1. (5.5)

( )
1 1 δ 1 1 δ δ
Choose x = and a = + . Then |x − a| = − + = < δ which satisfies (5.5).
δ δ 2 δ δ 2 2
We have |f (x) − f (a)| < 1 but
( )
δ 2 δ δ2
|f (x) − f (a)| = x − a
2 2
= |(x − a)(x + a)| = + =1+ > 1.
2 δ 2 4

It is contradiction. Hence, f (x) = x2 is not uniformly continuous on R.

Theorem 5.3.6 Suppose that I is a closed, bounded interval. If f : I → R is continuous on I,

then f is uniformly continuous on I.

Proof. Suppose to the contrary that f is continuous but not uniformly continuos on I.
Then there is an ε0 > 0 such that

for all δ > 0, |x − a| < δ and x, a ∈ I and |f (x) − f (a)| ≥ ε0 .

1 1
Set δ = . Then xn , yn ∈ I such that |xn − yn | < and
n n

|f (xn ) − f (yn )| ≥ ε0 , for n ∈ N. (5.6)

Then sequence {xn } and {yn } are bounded. By The Bolzano-Weierstrass Theorem, {xn } has a
subsequence, say xnk , that converges, as k → ∞, to some x ∈ I. Similarly, {yn } has a subsequence,
say ynj , that converges, as j → ∞, to some y ∈ I. Since xnj → x as j → ∞ and f is continuous,
it follows by the Comparison Theorem from (5.6) that

lim |f (xnj ) − f (ynj )| ≥ ε0

j→∞

|f (x) − f (y)| ≥ ε0 > 0

1
So, f (x) ̸= f (y). But |xn − yn | < for all n ∈ R, so Theorem 1.3.10 implies that x = y. Thus,
n
f (x) = f (y), a contradiction.
5.3. UNIFORM CONTINUITY 119

Theorem 5.3.7 Suppose that E ⊆ R and f : E → R is uniformly continuous. If xn ∈ E is

Cauchy, then f (xn ) is Cauchy.

Proof. Assume that f : E → R is uniformly continuous and xn is a Cauchy in E.

Let ε > 0. There is a δ > 0 such that

|x − a| < δ and x, a ∈ E imply |f (x) − f (a)| < ε. (5.7)

There is an N such that

n, m ≥ N implies |xn − xm | < δ.

For each n, m ≥ N such that |xn − xm | < δ it satisfies (5.7) that we have

|f (xn ) − f (xm )| < ε.

Therefore, f (xn ) is Cauchy.

120 CHAPTER 5. CONTINUITY ON R

Exercises 5.3

1. Use Definition to prove that each of the following functions is uniformly continuous on (0, 1).

1.1 f (x) = x3 1.2 f (x) = x2 − x 1

1.3 f (x) =
x+1

2. Prove that each of the following functions is uniformly continuous on (0, 1).

2.1 f (x) = (x + 1)2 2.4 f (x) is any polynomial

x −1
3
sin x
2.2 f (x) = 2.5 f (x) =
x−1 x
2.3 f (x) = x sin( x1 ) 2.6 f (x) = x2 ln x

1
3. Prove that f (x) = is uniformly continuous on R.
x2 +1
4. Find all real α such that xα sin( x1 ) is uniformly continuous on the open interval (0, 1).

5. Suppose that f : [0, ∞) → R is continuous and there is an L ∈ R such that f (x) → L as

x → ∞. Prove that f is uniformly continuous on [0, ∞).

6. Let I be a bounded interval. Prove that if f : I → R is is uniformly continuous on I, then

f is bounded on I.

7. Prove that (6) may be false if I is unbounded or if f is merely continuous.

8. Suppose that α ∈ R, E is nonempty subset of R, and f, g : E → R are uniformly continuous

on E.

8.1 Prove that f + g and αf are uniformly continuous on E.

8.2 Suppose that f, g are bounded on E. Prove that f g is uniformly continuous on E.

8.3 Show that there exist functions f, g uniformly continuous on R such that f g is not
uniformly continuous on R.

9. Prove that a polynomial of degree n is uniformly continuous on R if and only if n = 0 or

n = 1.
Chapter 6

Differentiability on R

6.1 The Derivative

Definition 6.1.1 A real function f is siad to be differentiable at a point a ∈ R if and only if f

is defined on some open interval I containing a and
f (a + h) − f (a)
f ′ (a) := lim
h→0 h
exists. In this case f ′ (a) is called the derivative of f at a.

You may recall that the graph of y = f (x) has a tangent line at the point (a, f (a)) if and
only if f has a derivative at a, in which case the slope of that tangent line is f ′ (a). Suppose that
f is differentiable at a. A secant line of the graph y = f (x) is a line passing through at least
two points on the graph, an a chord is a line segment that runs from one point on the graph to
another.
Y
y = f (x)

Chord

Tangent

a x2 x1 X
122 CHAPTER 6. DIFFERENTIABILITY ON R

Let x = a + h and observe that the slope of the chord (chord function : F (x)) passing through
the points (x, f (x)) and (a, f (a)) is given by

f (x) − f (a)
F (x) := , x ̸= a.
x−a
Now, since x = a + h, f ′ (a) becomes

f (x) − f (a)
f ′ (a) = lim .
x→a x−a
Example 6.1.2 Let f (x) = x2 where x ∈ R. Find f ′ (1)

Solution. We consider

f (x) − f (1) x2 − 1 (x − 1)(x + 1)

lim = lim lim = lim (x + 1) = 2.
x→1 x−1 x→1 x − 1 x→1 x−1 x→1

Thus, f is differentiable at 1 and f ′ (1) = 2.

Example 6.1.3 Show that the function



x2 cos( 1 )
x
if x ̸= 0
f (x) =

0 if x = 0

is differentiable at the origin.

Solution. Consider
( )
f (x) − f (0) x2 cos( x1 ) 1
lim = lim = lim x cos = 0.
x→0 x−0 x→0 x x→0 x
By Example 4.1.18, f ′ (0) = 0. Thus, f is differentiable at the origin.

Example 6.1.4 Show that the function



x cos( 1 )
x
if x ̸= 0
f (x) =

0 if x = 0

is not differentiable at the origin.

Solution. We consider
( )
f (x) − f (0) x cos( x1 ) 1
lim = lim = lim cos .
x→0 x−0 x→0 x x→0 x
By Example 4.1.12, the limit does not exist. Thus, f is not differentiable at the origin.
6.1. THE DERIVATIVE 123

Theorem 6.1.5 Let f : R → R. Then f is differentiable at a if and only if there is a function T

of the form T (x) := mx such that

f (a + h) − f (a) − T (h)
lim = 0.
h→0 h

Proof. Assume that f is differentiable at a. Then f ′ (a) exists. Choose m := f ′ (a).

We obtain

f (a + h) − f (a) − T (h) f (a + h) − f (a) − mh

lim = lim
h→0 h h→0 h
f (a + h) − f (a)
= lim −m
h→0 h
= f ′ (a) − f ′ (a) = 0

f (a + h) − f (a) − T (h)
Conversely, assume that lim = 0. Then
h→0 h

f (a + h) − f (a) f (a + h) − f (a) − mh
lim − m = lim
h→0 h h→0 h
f (a + h) − f (a) − T (h)
= lim = 0.
h→0 h

So, f ′ (a) = m. Thus, f is differentiable at a.

Theorem 6.1.6 If f is differentiable at a, then f is continuous at a.

Proof. Assume that f is differentiable at a. Then f ′ (a) exists. For x ̸= a, we have

f (x) − f (a)
f (x) − f (a) = · (x − a).
x−a

Taking limit x → a, we obtain

f (x) − f (a)
lim f (x) − f (a) = lim · (x − a) = f ′ (a) · 0 = 0
x→a x→a x−a

So, f (x) → f (a) as x → a. Hence, f is continuous at a.

124 CHAPTER 6. DIFFERENTIABILITY ON R

Example 6.1.7 Show that f (x) = |x| is continuous at 0 but not differentiable there.

Solution. We see that

f (x) − f (0) |x|
lim = lim
x→0 x−0 x→0 x

does not exist by Example 4.2.3. Thus, f is not differentiable at 0 but it easy to prove that f
continuous at 0.

DIFFERENTIABLE ON INTERVAL.

Definition 6.1.8 Let I be an interval and f : I → R be a function. f is said to be differentiable

on I if and only if f is differentiable at a for every a ∈ I

Example 6.1.9 Show that the function f (x) = x2 is differentiable on R.

Solution. Let a ∈ R. Then

f (x) − f (a) x 2 − a2 (x − a)(x + a)
lim = lim = lim = lim (x + a) = 2a.
x→a x−a x→a x − a x→a x−a x→a

Thus, f is is differentiable at a and f ′ (a) = 2a, i.e., f ′ (x) = 2x for all x ∈ R.

Theorem 6.1.10 Let n ∈ N. If f (x) = xn , then f is differentiable on R and

f ′ (x) = nxn−1 .

Proof. Use Binomial formula, we have

f (x + h) − f (x) (x + h)n − xn
f ′ (x) = lim = lim
h→0
 h    h
h→0
  
n n n
xn +   xn−1 h +   xn−2 h2 + · · · +   xhn−1 + hn  − xn
1 2 n−1
= lim
h→0
    h  
n n n
h   xn−1 +   xn−2 h + · · · +   xhn−2 + hn−1 
1 2 n−1
= lim
h→0
    h     
n n n n
= lim   xn−1 +   xn−2 h + · · · +   xhn−2 + hn−1  =   xn−1 = nxn−1 .
h→0
1 2 n−1 1
6.1. THE DERIVATIVE 125

Theorem 6.1.11 Every constant function is differentiable on R and its value equals to zero.

Proof. Let f (x) = c where c is a constant. Then

f (x + h) − f (x) c−c
f ′ (x) = lim = lim = lim 0 = 0
h→0 h h→0 h h→0

Thus, f is differentiable on R and f ′ (x) = 0.

√
Example 6.1.12 Show that f (x) = x is differentiable on (0, ∞) and f ′ (x).

Solution. Let a > 0

√ √ √ √
f (x) − f (a) x− a x+ a
lim = lim ·√ √
x→a x−a x→a x−a x+ a
x−a
= lim √ √
x→a (x − a)( x + a)
1 1
= lim √ √ = √
x→a x+ a 2 a
1
Thus f is is differentiable on (0, ∞) and f ′ (x) = √ for all x > 0.
2 x
Example 6.1.13 Show that f (x) = |x| is differentiable on [0, 1] and [−1, 0] but not on [−1, 1].


x if x ≥ 0
Solution. Consider f (x) = |x| = . Then f is differentiable on (−∞, 0) ∪ (0, ∞)

−x if x < 0
and 

1
′
if x > 0
f (x) = .

−1 if x < 0

Since f is not differentiable at 0, f is not differentiable on [−1, 1]. We see that

|x| |x|
lim+ =1 and lim− = −1.
x→0 x x→0 x
We conclude that f is not differentiable on [−1, 0] and [0, 1].
126 CHAPTER 6. DIFFERENTIABILITY ON R

Exercises 6.1

1. For each of the following real functions, use definition directly to prove that f ′ (a) exists.

1.1 f (x) = x3 , a ∈ R 1.3 f (x) = x2 + x + 2, a ∈ R

1 1
1.2 f (x) = , a ̸= 0 1.4 f (x) = √ , a > 0
x x

2. Prove that f (x) = x|x| is differentiable on R.

3. Let I be an open interval that contains 0 and f : I → R. If there exists an α > 1 such that

|f (x)| ≤ |x|α for all x ∈ I,

prove that f is differentiable at 0. What happens when α = 1 ?

( )
x
4. Suppose that f : (0, ∞) → R satisfies f (x)−f (y) = f for all x, y ∈ (0, ∞) and f (1) = 0.
y
4.1 Prove that f is continuous on (0, ∞) if and only if f is continuous at 1.

4.2 Prove that f is differentiable on (0, ∞) if and only if f is differentiable at 1.

f ′ (1)
4.3 Prove that if f is differentiable at 1, then f ′ (x) = for all x ∈ (0, ∞).
x


|x|α sin( 1 )
x
if x ̸= 0
5. Suppose that fα (x) = . Show that fα (x) is continuous at x = 0 when

0 if x = 0
α > 0 and differentiable at x = 0 when α > 1. Graph these functions for α = 1 and α = 2
and give a geometric interpretation of your results.

6. Prove that if f (x) = xα where α = 1

n
for somw n ∈ N, then y = f (x) is differentiable on
f ′ (x) = αxα−1 for every x ∈ (0, ∞).
sin x
7. Given lim = 1. Show that
x→0 x

7.1 (sin x)′ = cos x 7.2 (cos x)′ = − sin x

8. f is a constant function on I if and only if f ′ (x) = 0 for every x ∈ I.

6.2. DIFFERENTIABILITY THEOREM 127

6.2 Differentiability theorem

Theorem 6.2.1 (Additive Rule) Let f and g be real functions. If f and g are differentiable at
a, then f + g is differentiable at a. In fact,

(f + g)′ (a) = f ′ (a) + g ′ (a).

Proof. Assume that f and g are differentiable at a. Then

(f + g)(a + h) − (f + g)(a)
(f + g)′ (a) = lim
h→0 h
[f (a + h) − f (a)] + [g(a + h) − g(a)]
= lim
h→0 h
f (a + h) − f (a) g(a + h) − g(a)
= lim + lim
h→0 h h→0 h
= f ′ (a) + g ′ (a)

Thus, (f + g)′ (a) = f ′ (a) + g ′ (a).

Theorem 6.2.2 (Scalar Multiplicative Rule) Let f be a real function and α ∈ R. If f is

differentiable at a, then αf is differentiable at a. In fact,

(αf )′ (a) = αf ′ (a).

Proof. Assume that f is differentiable at a. Then

αf (a + h) − αf (a)
(αf (a))′ = lim
h→0 h
f (a + h) − f (a)
= α · lim
h→0 h
= αf ′ (a).

Thus, (αf )′ (a) = αf ′ (a).

128 CHAPTER 6. DIFFERENTIABILITY ON R

Theorem 6.2.3 (Product Rule) Let f and g be real functions. If f and g are differentiable at
a, then f g is differentiable at a. In fact,

(f g)′ (a) = g(a)f ′ (a) + f (a)g ′ (a).

Proof. Assume that f and g are differentiable at a. Then

(f g)(a + h) − (f g)(a)
(f g)′ (a) = lim
h→0 h
f (a + h)g(a + h) − f (a)g(a) + f (a + h)g(a) − f (a + h)g(a)
= lim
h→0 h
f (a + h)[g(a + h) − g(a)] + g(a)[f (a + h) − f (a)]
= lim
h→0 h
g(a + h) − g(a) f (a + h) − f (a)
= lim f (a + h) · + lim g(a) ·
h→0 h h→0 h
= f (a)g ′ (a) + g(a)f ′ (a).

Thus, (f g)′ (a) = g(a)f ′ (a) + f (a)g ′ (a).

Theorem 6.2.4 (Quotient Rule) Let f and g be real functions. If f and g are differentiable at
f
a, then is differentiable at a when g(a) ̸= 0. In fact,
g
( )′
f g(a)f ′ (a) − f (a)g ′ (a)
(a) = .
g [g(a)]2

Proof. Assume that f and g are differentiable at a when g(a) ̸= 0. Then

( )′
f
f
g
(a + h) − fg (a)
f (a+h)
g(a+h)
− fg(a)
(a)

(a) = lim = lim

g h→0 h h→0 h
f (a+h)
g(a+h)
− f (a)
g(a+h)
+ f (a)
g(a+h)
− f (a)
g(a)
= lim
h→0 h [ ]
1
g(a+h)
[f (a + h) − f (a)] + f (a) 1
g(a+h)
− 1
g(a)
= lim
h→0 h [ ]
g(a)
g(a)g(a+h)
[f (a + h) − f (a)] − f (a) g(a+h)−g(a)
g(a+h)g(a)
= lim
h→0
[ ] h [ ]
g(a) f (a+h)−f (a)
h
− f (a) g(a+h)−g(a)
h g(a)f ′ (a) − f (a)g ′ (a)
= lim = .
h→0 g(a)g(a + h) [g(a)]2
6.2. DIFFERENTIABILITY THEOREM 129

Example 6.2.5 Let f and g be differentiable at 1 with f (1) = 1, g(1) = 2 and f ′ (1) = 3, g ′ (1) = 4.
Evaluate the following derivatives.

1. (f + g)′ (1) = f ′ (1) + g ′ (1) = 3 + 4 = 7.

2. (2f )′ (1) = 2f ′ (1) = 2 · 3 = 6.

3. (f g)′ (1) = f (1)g ′ (1) + f ′ (1)g(1) = 1 · 4 + 3 · 2 = 10.

( )′
f g(1)f ′ (1) − f (1)g ′ (1) 2·3−1·4 1
4. (1) = 2
= 2
= .
g [g(1)] 2 2

Theorem 6.2.6 (Chain Rule) Let f and g be real functions. If f is differentiable at a and g is
differentiable at f (a), then g ◦ f is differentiable at a with

(g ◦ f )′ (a) = g ′ (f (a))f ′ (a).

Proof. Assume that f is differentiable at a and g is differentiable at f (a).

Then f ′ (a) and g ′ (f (a)) exist. We consider

f (x) − f (a)
f (x) = · (x − a) + f (a), x ̸= a
x−a
g(y) − g(f (a))
g(y) = · (y − f (a)) + g(f (a)), y ̸= f (a) (6.1)
y − f (a)

Since f is continuous at a, substitue y = f (x) in (6.1) to write

g(f (x)) − g(f (a))

g(f (x)) = · (f (x) − f (a)) + g(f (a))
f (x) − f (a)
g(f (x)) − g(f (a)) f (x) − f (a)
g(f (x)) = · · (x − a) + g(f (a))
f (x) − f (a) x−a
g(f (x)) − g(f (a)) g(f (x)) − g(f (a)) f (x) − f (a)
= ·
x−a f (x) − f (a) x−a
g(f (x)) − g(f (a)) g(f (x)) − g(f (a)) f (x) − f (a)
lim = lim ·
x→a x−a x→a f (x) − f (a) x−a
(g ◦ f )′ (a) = g ′ (f (a)) · f ′ (a)
130 CHAPTER 6. DIFFERENTIABILITY ON R

Example 6.2.7 Let f and g be differentiable on R with f (0) = 1, g(0) = −1 and f ′ (0) = 2,
g ′ (0) = −2 ,f ′ (−1) = 3, g ′ (1) = 4. Evaluate each of the following derivatives.

1. (f ◦ g)′ (0) = f ′ (g(0))g ′ (0) = f ′ (−1) · g ′ (0) = 3(−2) = −6.

2. (g ◦ f )′ (0) = g ′ (f (0))f ′ (0) = g ′ (1) · f ′ (0) = 4(2) = 8.

√ x
Example 6.2.8 Let f (x) = x2 + 1. Use the Chain Rule to show that f ′ (x) = √ .
2
x +1
√
Solution. Let g(x) = x and h(x) = x2 + 1. We have
1
g ′ (x) = √ and h′ (x) = 2x.
2 x
By Chain Rule,

f ′ (x) = (g ◦ h)′ (x) = g ′ (h(x))h′ (x)

1
= √ · h′ (x)
2 h(x)
x
=√ .
2
x +1
6.2. DIFFERENTIABILITY THEOREM 131

Exercises 6.2

1. For each of the following functions, find all x for which f ′ (x) exists and find a formula for f ′ .

x3 − 2x2 + 3x 1.3 f (x) = x|x|

1.1 f (x) = √
x
1
1.2 f (x) = 2 1.4 f (x) = |x3 + 2x2 − x − 2|
x +x−1

2. Let f and g be differentiable at 2 and 3 with f ′ (2) = a, f ′ (3) = b, g ′ (2) = c and g ′ (3) = d, If
f (2) = 1, f (3) = 2 ,g(2) = 3 and g(3) = 4. Evaluate each of the following derivatives.
( )′
2.1 (f g)′ (2) 2.2 f
(3) 2.3 (g ◦ f )′ (3) 2.4 (f ◦ g)′ (2)
g

3. If f, g and h is differentiable at a, prove that f gh is differentiable at a and

(f gh)′ (a) = f ′ (a)g(a)h(a) + f (a)g ′ (a)h(a) + f (a)g(a)h′ (a).

4. Let f (x) = (x − 1)(x − 2)(x − 3) · · · (x − 2565). Find f ′ (2565)

m
5. Prove that if f (x) = x n for some n, m ∈ N, then y = f (x) is differentiable
and satisfies ny n−1 y ′ = mxm−1 for every x ∈ (0, ∞).

6. (Power Rule) Prove that f (x) = xq for some q ∈ Q, then f is differentiable

and f ′ (x) = qxq−1 for every x ∈ (0, ∞).

7. (Reciprocal Rule) Suppose that f is differentiable at a and f (a) ̸= 0.

7.1 Show that for h sufficiently small, f (a + h) ̸= 0.

1
7.2 Use Definition 6.1.1 directly, prove that is differentiable at x = a and
f (x)
( )′
1 f ′ (a)
(a) = − 2 .
f f (a)
8. Suppose hat n ∈ N and f, g are real functions of a real variable whose nth derivatives f (n) , g (n)
exist at a point a. Prove Leibniz’s generalization of the Product Rule:
 
∑n
n
(f g)(n) (a) =   f (k) (a)g (n−k) (a).
k=0 k
132 CHAPTER 6. DIFFERENTIABILITY ON R

6.3 Mean Value Theorem

Lemma 6.3.1 (Rolle’s Theorem) Suppose that a, b ∈ R with a ̸= b. If f is continuous on [a, b],
differentiable on (a, b), and if f (a) = f (b), then f ′ (c) = 0 for some c ∈ (a, b).

Y
f ′ (c) = 0

y = f (x)
f (a) = f (b)

a c X
b

Proof. Let a ̸= b such that f is continuous on [a, b] and differentiable on (a, b).
Assume that f (a) = f (b). By EVT, f has a finite maximum M and a finite minimum m on [a, b].
Case M = m. Then f is a constant function. Thus, f ′ (x) = 0 for all x ∈ (a, b).
Case M ̸= m. Since f (a) = f (b), there is a c ∈ (a, b) such that f (c) = M . We have

f (c + h) ≤ f (c) for all h that satisfy c + h ∈ (a, b).

In the case h > 0 this implies that

f (c + h) − f (c)
f ′ (c) = lim+ ≤ 0,
h→0 h

and in this case h < 0 this implies that

f (c + h) − f (c)
f ′ (c) = lim− ≥ 0.
h→0 h

It follows that f ′ (c) = 0.

6.3. MEAN VALUE THEOREM 133

Y
Tangent
Chord
y = f (x)

a c X
b

Theorem 6.3.2 (Mean Value Theorem (MVT)) Suppose that a, b ∈ R with a ̸= b.

If f is continuous on [a, b] and differentiable on (a, b), then there is an c ∈ (a, b) such that

f (b) − f (a) = f ′ (c)(b − a).

Proof. Let a ̸= b such that f is continuous on [a, b] and differentiable on (a, b). We set

h(x) = f (x)(b − a) − x[f (b) − f (a)] for x ∈ [a, b].

Then h is continuous on [a, b] and differentiable (a, b),

h′ (x) = f ′ (x)(b − a) − [f (b) − f (a)].

We obtain

h(a) = f (a)(b − a) − a[f (b) − f (a)] = bf (a) − af (b)

= bf (a) − af (b) + bf (b) − bf (b) = f (b)(b − a) − b[f (b) − f (a)] = h(b).

By the Rolle’s Theorem, there is a c ∈ (a, b) such that h′ (c) = 0, i.e.,

f ′ (c)(b − a) − [f (b) − f (a)] = 0.

Hence, f (b) − f (a) = f ′ (c)(b − a).

134 CHAPTER 6. DIFFERENTIABILITY ON R

Example 6.3.3 Prove that

sin x ≤ x for all x > 0.

Solution. Let a > 0 and define f (x) = sin x where x ∈ [0, a] . Then f is continuous on [0, a] and
f (x) is differentiable and f ′ (x) = cos x for every x ∈ (0, a).
By the MVT, there is a c ∈ (0, a) such that

f (a) − f (0) = f ′ (c)(a − 0)

sin a − 0 cos c · a

sin a = a cos c

From cos c ≤ 1 and a > 0, a cos c ≤ a, it implies that sin a < a. Therefore,

sin x ≤ x for all x > 0.

Example 6.3.4 Prove that

1 + x ≤ ex for all x > 0.

Solution. Let a > 0 and define f (x) = ex − x − 1 where x ∈ [0, a] . Then f is continuous on [0, a]
and f (x) is differentiable and f ′ (x) = ex − 1 for every x ∈ (0, a).
By the MVT, there is a c ∈ (0, a) such that

f (a) − f (0) = f ′ (c)(a − 0)

(ea − a − 1) − 0 = (ec − 1)a

ea − a − 1 = (ec − 1)a

Since c ≥ 0, ec ≥ 1 or ec − 1 ≥ 0. From a > 0, it implies that (ec − 1)a ≥ 0 which leads to

ea − a − 1 ≥ 0 Therefore,

1 + x ≤ ex for all x > 0.

6.3. MEAN VALUE THEOREM 135

Example 6.3.5 (Bernoulli’s Inequality) Let 0 < α ≤ 1 and δ ≥ −1. Prove that

(1 + δ)α ≤ 1 + αδ.

Proof. Let 0 < α ≤ 1 and δ ≥ −1. Define f (x) = xα where x ∈ R. Then f is continuous on R and
f (x) is differentiable and

f ′ (x) = αxα−1 for every x ∈ R.

Case −1 ≤ δ ≤ 0. By the MVT, there is a c ∈ (1 + δ, 1) such that

f (1) − f (1 + δ) = f ′ (c)[1 − (1 + δ)]

1 − (1 + δ)α = −δαcα−1

(1 + δ)α − 1 = δαcα−1

Since 0 < α ≤ 1, −1 < α − 1 ≤ 0. From 0 ≤ 1 + δ < c < 1, it implies that cα−1 ≥ c0 = 1. Since
δ ≤ 0 and α > 0, δα ≤ 0 which leads to δαcα−1 ≤ αδ. Thus,

(1 + δ)α ≤ 1 + αδ.

Case δ > 0. By the MVT, there is a c ∈ (1, 1 + δ) such that

f (1 + δ) − f (1) = f ′ (c)[(1 + δ) − 1]

(1 + δ)α − 1 = δαcα−1

Since 0 < α ≤ 1, −1 < α − 1 ≤ 0. From c > 1, it implies that cα−1 ≤ c0 = 1. Since δ > 0 and
α > 0, δα > 0 which leads to δαcα−1 ≤ αδ. Thus,

(1 + δ)α ≤ 1 + αδ.

We conclude that (1 + δ)α ≤ 1 + αδ for 0 < α ≤ 1 and δ ≥ −1.

136 CHAPTER 6. DIFFERENTIABILITY ON R

Theorem 6.3.6 (Generalized Mean Value Theorem) Suppose that a, b ∈ R with a ̸= b.

If f and g are continuous on [a, b] and differentiable on (a, b), then there is an c ∈ (a, b) such that

g ′ (c)[f (b) − f (a)] = f ′ (c)[g(b) − g(a)].

Proof. Let a ̸= b such that f and g are continuous on [a, b] and differentiable on (a, b). We set

h(x) = f (x)[g(b) − g(a)] − g(x)[f (b) − f (a)] for x ∈ [a, b].

Then h is continuous on [a, b] and differentiable (a, b),

h′ (x) = f ′ (x)[g(b) − g(a)] − g ′ (x)[f (b) − f (a)].

We obtain

h(a) = f (a)[g(b) − g(a)] − g(a)[f (b) − f (a)]

= f (a)g(b) − f (a)g(a) − g(a)f (b) + g(a)f (a)

= f (a)g(b) − g(a)f (b)

= f (a)g(b) − g(a)f (b) + g(b)f (b) − g(b)f (b)

= [f (b)g(b) − f (b)g(a)] + [g(b)f (a) − g(b)f (b)]

= f (b)[g(b) − g(a)] − g(b)[f (b) − f (a)]

= h(b).

By the Rolle’s Theorem, there is a c ∈ (a, b) such that h′ (c) = 0, i.e.,

f ′ (c)[g(b) − g(a)] − g ′ (c)[f (b) − f (a)] = 0.

Hence, g ′ (c)[f (b) − f (a)] = f ′ (c)[g(b) − g(a)].

6.3. MEAN VALUE THEOREM 137

Theorem 6.3.7 (L’Hôspital’s Rule) Let a be an extended real number and I be an open interval
that either contains a or has a as an endpoint. Suppose that f and g are differentiable on I\{a},
and g(x) ̸= 0 ̸= g ′ (x) for all x ∈ I\{a}. Suppose further that

A := lim f (x) = lim g(x)

x→a x→a

is either 0 or ∞. If
f ′ (x)
B := lim
x→a g ′ (x)

exists as an extended real number, then

f (x) f ′ (x)
lim = lim ′ .
x→a g(x) x→a g (x)

Proof. We will use the SCL to prove that

f (x)
lim = B.
x→a g(x)

Let xk ∈ I\{a} such that xk → a as k → ∞. Note that if g ′ is never zero on I\{a}. By the MVT,
for x, y < a or x, y > a there is a c ∈ (x, y) such that

g(x) − g(y) = g ′ (c)(y − x) ̸= 0 for all x ̸= y.

We suppose for simplicity that B ∈ R. (For case B = ±∞, see Exercise.)

Case 1. A = 0 and a ∈ R. Extend f and g to I ∪ {a} by f (a) = 0 = g(a). By hypothesis,
f and g are continuous on I ∪ {a} and differentiable on I\{a}. By the Generalized Mean Value
Theorem, there is a ck between xk and a such that

g ′ (ck )[f (xk ) − f (a)] = f ′ (ck )[g(xk ) − g(a)]

g ′ (ck )[f (xk ) − 0] = f ′ (ck )[g(xk ) − 0]

f (xk ) f ′ (ck )
= ′
g(xk ) g (ck )

From xk < ck < a or a < ck < xk , it implies ck → a as k → ∞ by the Squeeze Theorem.

We conclude that

f (xk ) f ′ (ck ) f ′ (a) f ′ (x)

lim = lim ′ = ′ = lim ′ = B.
k→∞ g(xk ) k→∞ g (ck ) g (a) x→a g (x)
138 CHAPTER 6. DIFFERENTIABILITY ON R

Case 2. A = ±∞ and a ∈ R. We suppose by symmetry that A = ∞. For each k, n ∈ N, apply

the Generalized Mean Value Theorem, there is a ck,n between xk and xn such that
f ′ (ck,n )
f (xn ) − f (xk ) = · [g(xn ) − g(xk )].
g ′ (ck,n )
We obtain
f (xn ) f (xk ) f (xn ) − f (xk ) 1 f ′ (ck,n )
− = = · ′ · [g(xn ) − g(xk )]
g(xn ) g(xn ) g(xn ) g(xn ) g (ck,n )
f ′ (ck,n ) g(xk ) f ′ (ck,n )
= ′ − · .
g (ck,n ) g(xn ) g ′ (ck,n )
It leads to
f (xn ) f (xk ) f ′ (ck,n ) g(xk ) f ′ (ck,n )
= + − · . (6.2)
g(xn ) g(xn ) g ′ (ck,n ) g(xn ) g ′ (ck,n )
1
Since A = ∞, it is clear that → 0 as n → ∞, and since cn,k lies between xk and xn , it also
g(xn )
f ′ (ck,n )
clear that ck,n → a as k, n → ∞ by the Squeeze Theorem. Thus, the limit of ′ exists as
g (ck,n )
n → ∞ and fixed k ∈ N, we obtain
f (xk ) g(xk )
lim =0 and lim = 0.
n→∞ g(xn ) n→∞ g(xn )

Hence, (6.2) becomes to

[ ]
f (xn ) f (xk ) f ′ (ck,n ) g(xk ) f ′ (ck,n )
lim = lim + ′ − ·
n→∞ g(xn ) n→∞ g(xn ) g (ck,n ) g(xn ) g ′ (ck,n )
f ′ (ck,n ) f ′ (ck,n )
= 0 + lim ′ − 0 · lim ′
n→∞ g (ck,n ) n→∞ g (ck,n )
′ ′
f (ck,n ) f (a) f ′ (x)
= lim ′ = ′ = lim ′ = B.
n→∞ g (ck,n ) g (a) x→a g (x)
Case 3. a = ±∞. We suppose by symmetry that a = ∞. Choose c > 0 such that (c, ∞) ⊂ I.
For each y ∈ (0, 1c ), set
( ) ( )
1 1
ϕ(y) = f and φ(y) = g .
y y
By the Chain Rule,
ϕ′ (y) f ′ ( y1 ) · (− y12 ) f ′ ( y1 )
= ′ 1 = ′ 1 .
φ′ (y) g ( y ) · (− y12 ) g (y)
ϕ′ (y) f ′ (x)
Thus, for x = 1
∈ (c, ∞), we have = . Since x → ∞ if and only if y = x1 → 0+ , it
y
φ′ (y) g ′ (x)
follows that ϕ and φ satisfy the hypothesis of Case 1 or 2 for a = 0 and I = (0, 1c ). In particular,
f ′ (x) ϕ′ (y) ϕ(y) f (x)
lim = lim = lim = lim .
x→∞ g ′ (x) y→0+ φ′ (y) y→0+ φ(y) x→∞ g(x)
6.3. MEAN VALUE THEOREM 139

1
Given (ln x)′ = for x > 0 and (ex )′ = ex for all x ∈ R.
x
x
Example 6.3.8 Use L’Hôspital’s Rule to prove that lim = 1.
x→0 ex −1
Solution. We see that
lim x = 0 = lim ex − 1.
x→0 x→0

By L’Hôspital’s Rule, it follows that

x (x)′ 1
lim = lim = lim x = 1.
x→0 e − 1
x x→0 (e − 1)
x ′ x→0 e

Example 6.3.9 Use L’Hôspital’s Rule to find lim+ x ln x.

x→0

Solution. We see that

1
lim+ ln x = ∞ = lim+ .
x→0 x→0 x

By L’Hôspital’s Rule, it follows that

ln x (ln x)′
lim+ x ln x = lim+ = lim
x→0 x→0 x−1 x→0+ (x−1 )′

x−1
= lim+ = lim+ (−x) = 0.
x→0 −x−2 x→0

Example 6.3.10 Use L’Hôspital’s Rule to find L = lim− (ln x)1−x .

x→1

Solution. We see that

1
lim− ln(ln x) = −∞ = lim− .
x→1 x→1 1−x
Since ln x is continuous on (0, ∞), by L’Hôspital’s Rule we have

(ln(ln x))′
ln L = ln lim− (ln x)1−x = lim− ln(ln x)1−x = lim− (1 − x) ln(ln x) = lim− 1 ′
x→1 x→1 x→1 x→1 ( 1−x )
1
ln x x
·1 (1 − x)2
= lim− 1 = lim−
x→1
(1−x)2
x→1 x ln x

Apply again L’Hôspital’s Rule, we obtain

[(1 − x)2 ]′ −2(1 − x)

ln L = lim− ′
= lim− =0
x→1 [x ln x] x→1 ln x + 1
L = e0 = 1.

Hence, L = lim− (ln x)1−x = 1.

x→1
140 CHAPTER 6. DIFFERENTIABILITY ON R

Exercises 6.3

1. Use the Mean Value Theorem to prove that each of the following inequalities.

√ x−1
1.1 2x + 1 < 1 + x for all x > 0 1.6 ≤ ln x for all x > 1
x
1.2 ln x ≤ x − 1 for all x > 1 √
1.7 x ≥ x for all x ∈ [0, 1]
1.3 7(x − 1) < ex √
for all x > 2 1.8 x≤x for all x > 1

1.4 cos x − 1 ≤ x for all x > 0 1.9 sin2 x ≤ 2|x| for all x ∈ R
2
x +1 √
1.5 ln x + 1 ≤ for all x > 1 1.10 ln x ≤ x for all x > 1
2

2. (Bernoulli’s Inequality) Let α ≥ 1 and δ ≥ −1. Prove that

(1 + δ)α ≤ 1 + αδ.

3. Use L’Hôspital’s Rule to evaluate the following limits.

( )x
sin(3x) 3.4 lim+ xx 3.7 lim− 1 + e−x
3.1 lim x→0 x→0
x→0 x
cos x − ex ln x 1
3.2 lim+ 3.5 lim 3.8 lim (1 + x) x
x→0 ln(1 + x2 ) x→1 sin(πx) x→0
( x ) 12 ( π) 1
3.6 lim x arctan x − 3.9 lim x(e x − 1)
x
3.3 lim
x→0 sin x x→∞ 2 x→∞

4. Show that the derivative of 


e− x12 if x ̸= 0
f (x) =

0 if x = 0

exists and continuous on R with f ′ (0) = 0.

5. Suppose that f is differentiable on R.

5.1 If f ′ (x) = 0 for all x ∈ R, prove that f (x) = f (0) for all x ∈ R

5.2 If f (0) = 1 and |f ′ (x)| ≤ 1 for all x ∈ R, prove that |f (x)| ≤ |x| + 1 for all x ∈ R

5.3 If ′ (x) ≥ 0 for all x ∈ R, prove that a < b imply that f (a) < f (b)
6.3. MEAN VALUE THEOREM 141

6. Let f be differentiable on a nonempty, open interval (a, b) with f ′ bounded on (a, b). Prove
that f is uniformly continuous on (a, b).

7. Let f be differentiable on (a, b), continuous on [a, b], with f (a) = f (b) = 0. Prove that if
f ′ (c) > 0 for some c ∈ (a, b), then there exist x1 , x2 ∈ (a, b) such that f ′ (x1 ) > 0 > f ′ (x2 ).

8. Let f be twice differentiable on (a, b) and let there be points x1 < x2 < x3 in (a, b) such that
f (x1 ) > f (x2 ) and f (x3 ) > f (x2 ). Prove that there is a point c ∈ (a, b) such that f ′′ (c) > 0.

9. Let f be differentiable on (0, ∞). If L = lim f ′ (x) and lim f (n) both exist and are finite,
x→∞ n→∞
prove that L = 0.

10. Prove L’Hôspital’s Rule for the case B = ±∞ by first proving that

g(x) f (x)
→ 0 when → ±∞, as x → a.
f (x) g(x)
( )n
1
11. Prove that the sequence 1 + is increasing, as n → ∞, and its limit e satisfies 2 < e ≤ 3
n
and ln e = 1.
142 CHAPTER 6. DIFFERENTIABILITY ON R

6.4 Monotone function

Definition 6.4.1 Let E be a nonempty subset of R and f : E → R.

1. f is said to be increasing on E if and only if

x1 , x2 ∈ E and x1 < x2 imply f (x1 ) ≤ f (x2 ).

f is said to be strictly increasing on E if and only if

x1 , x2 ∈ E and x1 < x2 imply f (x1 ) < f (x2 ).

2. f is said to be decreasing on E if and only if

x1 , x2 ∈ E and x1 < x2 imply f (x1 ) ≥ f (x2 ).

f is said to be strictly decreasing on E if and only if

x1 , x2 ∈ E and x1 < x2 imply f (x1 ) > f (x2 ).

3. f is said to be monotone on E if and only if f is either decreasing or increasing on E.

f is said to be strictly monotone on E if and only if f is either strictly decreasing or
strictly increasing on E.

Example 6.4.2 Show that f (x) = x2 is strictly monotone on [0, 1] and on [−1, 0] but not monotone
on [−1, 1].

Solution.
If 0 ≤ x < y ≤ 1, then x2 < y 2 , i.e., f (x) < f (y). Thus, f is strictly increasing on [0, 1].
If −1 ≤ x < y ≤ 0, then x2 > y 2 , i.e., f (x) > f (y). Thus, f is strictly decreasing on [−1, 0].
We conclude that f is strictly monotone on [0, 1] and on [−1, 0].
Since f is increasing and decreasing on [−1, 1], f is not monotone on [−1, 1].
6.4. MONOTONE FUNCTION 143

Theorem 6.4.3 Let f : I → R and (a, b) ⊆ I. Then

1. f is increasing on (a, b) if f ′ (x) > 0 for all x ∈ (a, b)

2. f is decreasing on (a, b) if f ′ (x) < 0 for all x ∈ (a, b)

3. If f ′ (x) = 0 for all x ∈ (a, b), then f is constant on [a, b].

Proof. Let x, y ∈ (a, b) such that x < y. Then y − x > 0.

By the MVT, there is a c ∈ (x, y) such that

f (y) − f (x) = f ′ (c)(y − x) > 0.

If f ′ (x) > 0 for all x ∈ (a, b), f ′ (c) > 0. It follows that f (y) > f (x). So, f is increasing on (a, b).
If f ′ (x) < 0 for all x ∈ (a, b), f ′ (c) < 0. It follows that f (y) < f (x). So, f is decreasing on (a, b).
Let x ∈ [a, b]. By the MVT, there is a c ∈ (a, x) such that

f (x) − f (a) = f ′ (c)(x − a) = 0.

So, f (x) = f (a) for all x ∈ [a, b]. We conclude that f is constant on [a, b].

Example 6.4.4 Find each intervals of f (x) = x2 − 4x + 3 that increasing and decreasing.

Solution. We have f ′ (x) = 2x − 4. Consider

2x − 4 = f ′ (x) > 0 implies x > 2.

Thus, f is increasing on (2, ∞).

2x − 4 = f ′ (x) < 0 implies x < 2.

Thus, f is increasing on (−∞, 2).

144 CHAPTER 6. DIFFERENTIABILITY ON R

Theorem 6.4.5 If f is 1-1 and continuous on an interval I, then f is strictly monotone on I and
f −1 is continuous and strictly monotone on f (I) := {f (x) : x ∈ I}.

Proof. Assume that f is 1-1 and continuous on an interval I. Let a, b ∈ I such that

a < b implies either f (a) < f (b) or f (a) > f (b).

Suppose that f is not strictly monotone on I. Then there exist points a, b, c ∈ I such that a < c < b
but f (c) does not lie between f (a) and f (b). It follows that either f (a) lie between f (b) and f (c)
or f (b) lie between f (a) and f (c). Hence by the IVT, there is an x1 ∈ (a, b) such that

f (x1 ) = f (a) or f (x1 ) = f (b).

Since f is 1-1, we conclude that either x1 = a or x2 = b, a contradiction. Therefore, f is strictly

monotone on I.
We may suppose that f is strictly increasing on I. Since f is 1-1 on I, apply Theorem 1.4.3 to
verify that f −1 takes f (I) onto I. We will show that f −1 is strictly increasing on f (I). Suppose
to the contrary that there exist y1 , y2 ∈ f (I) such that

y1 < y 2 but f −1 (y1 ) ≥ f −1 (y2 ).

Then x1 = f −1 (y1 ) and x2 = f −1 (y2 ) satisfy x1 ≥ x2 and x1 , x2 ∈ I. Since f is strictly increasing

on I, it follows that y1 = f (x1 ) ≥ f (x2 ) = y2 , a contracdiction.
Thus, f −1 is strictly increasing on f (I).
Since I is a interval, it easy to prove that f (I) is also inverval. Fix y0 ∈ f (I) and ε > 0.
Since f −1 is strictly increasing on f (I), if y0 is not right endpoint of f (I), then x0 = f −1 (y0 )
is not right endpoint of I. There is an ε0 > 0 so small that ε0 < ε and x0 + ε0 ∈ I. Choose
δ = f (x0 + ε0 ) − f (x0 ) and suppose that 0 < y − y0 < δ. The choice of δ implies that

y0 < y < y0 + δ = f (x0 ) + δ = f (x0 + ε0 ).

Set y = f −1 (x). Then f (x0 ) < f (x) < f (x0 + ε0 ). Since f is strictly increasing on I, it implies
x0 < x < x0 + ε0 , i.e., 0 < x − x0 < ε0 . We conclude that

0 < f −1 (x) − f −1 (y0 ) < ε.

So, f −1 (y0+ ) = f −1 (y0 ). A similar argument show that if y0 is not a left endpoint of f (I),
f −1 (y0− ) = f −1 (y0 ). Hence, f −1 is continuous on f (I).
6.4. MONOTONE FUNCTION 145

Theorem 6.4.6 (Inverse Function Theorem (IFT)) Let f be 1-1 and continuous on an open
interval I. If a ∈ f (I) and if f ′ (f −1 (a)) exists and is nonzero, then f −1 is differentiable at a and

1
(f −1 )′ (a) = .
f ′ (f −1 (a))

Proof. Let f be 1-1 and continuous on an open interval I. By Theorem 6.4.5, f is strictly monotone,
say strictly increasing on I and f −1 exists, is continuous and strictly increasing on f (I).
Assume that a ∈ f (I) and f ′ (f −1 (a)) exists and is nonzero. Set x0 = f −1 (a) ∈ I and I is open,
we can choose c, d ∈ R such that x0 ∈ (c, d) ⊂ I. Then a = f (x0 ) ∈ (f (c), f (d)) ⊂ f (I).
We can choose h ̸= 0 so small that a + h ∈ f (I). i.e., f −1 (a + h) exists. Set x = f −1 (a + h) and
observe that f (x) − f (x0 ) = a + h − a = h. Since f −1 is continuous, x → x0 if and only if h → 0.
Therefore,

f −1 (a + h) − f −1 (a) x − x0 1 1
(f −1 )′ (a) = lim = lim = ′ = ′ −1 .
h→0 h x→x0 f (x) − f (x0 ) f (x0 ) f (f (a))

Example 6.4.7 Use the Inverse Function Theorem to find derivative of f (x) = arcsin x

Solution. Let g(x) = sin x where x ∈ (− π2 , π2 ). Then g is 1-1 and continuous on (− π2 , π2 ).

We have g ′ (x) = cos x > 0 for all x ∈ (− π2 , π2 ) and g −1 (x) = arcsin x = f (x).
By the IFT, we obtain

1
f ′ (x) = (arcsin x)′ = (g −1 )′ (x) =
g ′ (g −1 (x))
1
=
g ′ (arcsin x)
1
=
cos(arcsin x)
1
=√
1 − sin2 (arcsin x)
1
=√ .
1 − x2
146 CHAPTER 6. DIFFERENTIABILITY ON R

Example 6.4.8 Let f (x) = x + ex where x ∈ R.

1. Show that f is 1-1 on x ∈ R.

2. Use the result from 1 and the IFT to explain that f −1 differentiable on R.

3. Compute (f −1 )′ (2 + ln 2).

Solution.

1. Proof. Let x, y ∈ R and x ̸= y. WLOG x > y. Then x − y > 0 and ex > ey . We obtain

ey − ex < 0 < x − y

y + ey < x + ex

f (y) < f (x).

So, f (x) ̸= f (y). Therefore, f is injective in R.

2. Since f is 1-1, f −1 exists. It is clear that f is continous on R. By the IFT, we conclude that
f −1 differentiable on R.

3. We see that f ′ (x) = 1 + ex and f (ln 2) = ln 2 + 2. So, f −1 (2 + ln 2) = ln 2.

By the IFT, we obtain

1 1 1 1
(f −1 )′ (2 + ln 2) = = = = .
f ′ (f −1 (2 + ln 2)) f ′ (ln 2) 1+2 3
6.4. MONOTONE FUNCTION 147

Exercises 6.4

1. Find each intervals of the following functions that increasing and decreasing.

1.1 f (x) = 2x − x2 1.4 g(x) = xex

1.2 f (x) = x3 − x2 − x + 3 1.5 g(x) = ex − x

1.3 f (x) = (x − 1)3 (x − 2)4

2
1.6 g(x) = x2 ex

2. Find all a ∈ R such that x3 + ax2 + 3x + 15 is strictly increasing near x = 1.

3. Find all a ∈ R such that ax2 + 3x + 5 is strictly increasing on the interval (1, 2).
√
4. Find where f (x) = 2|x − 1| + 5 x2 + 9 is strictly increasing and where f (x) is strictly
decreasing.

5. Let f and g be 1-1 and continuous on R. If f (0) = 2, g(1) = 2, f ′ (0) = π, and g ′ (1) = e,
compute the following derivatives.

5.1 (f −1 )′ (2) 5.2 (g −1 )′ (2) 5.3 (f −1 · g −1 )′ (2)

6. Let f (x) = x2 ex , x ∈ R.
2

6.1 Show that f −1 exists and its differentiable on (0, ∞).

6.2 Compute (f −1 )′ (e)

7. Let f (x) = x + e2x where x ∈ R.

7.1 Show that f is 1-1 on x ∈ R.

7.2 Use the result from 7.1 and the IFT to explain that f differentiable on R.

7.3 Compute (f −1 )′ (4 + ln 2).

8. Use the Inverse Function Theorem, prove that

1
8.1 (arccos x)′ = − √ where x ∈ (−1, 1)
1 − x2
1
8.2 (arctan x)′ = where x ∈ (−∞, ∞)
1 + x2
148 CHAPTER 6. DIFFERENTIABILITY ON R

√ 1
8.3 ( x)′ = √ where x ∈ (0, ∞)
2 x
9. Use the IFT to find derivative of invrese function f (x) = ex − e−x where x ∈ R.

10. Suppose that f ′ exists and continuous on a nonempty, open interval (a, b) with f ′ (x) ̸= 0 for
all x ∈ (a, b).

10.1 Prove that f is 1-1 on (a, b) and takes (a, b) onto some open interval (c, d)

10.2 Show that (f −1 )′ exists and continuous on (c, d)

10.3 Use the function f (x) = x3 , show that 7.2 is false if the assumption f ′ (x) ̸= 0 fails to
hold for some x ∈ (c, d)

11. Let [a, b] be a closed, bounded interval. Find all functions f that satisfy the following
conditions for some fixed α > 0 : f is continuous and 1-1 on [a, b],

f ′ (x) ̸= 0 and f ′ (x) = α(f −1 )′ (f (x)) for all x ∈ (a, b).

12. Let f be differentiable at every point in a closed, bounded interval [a, b]. Prove that if f ′ is
increasing on (a, b), then f ′ is continuous on (a, b).

13. Suppose that f is increasing on [a, b]. Prove that

13.1 if x0 ∈ [a, b), then f (x+

0 ) exists and f (x0 ) ≤ f (x0 ),
+

13.2 if x0 ∈ (a, b], then f (x− −

0 ) exists and f (x0 ) ≤ f (x0 ).
Chapter 7

Integrability on R

7.1 Riemann integral

PARTITION.

Definition 7.1.1 Let a, b ∈ R with a < b.

1. A partition of the interval [a, b] is a set of points P = {x0 , x1 , , ..., xn } such that

a = x0 < x1 < · · · < xn = b.

2. The norm of a partition P = {x0 , x1 , , ..., xn } is the number

∥P ∥ = max |xj − xj−1 |.

1≤j≤n

3. A refinement of a partition P = {x0 , x1 , , ..., xn } is a partition Q of [a, b] that satisfies

Q ⊇ P . In this case we say that Q is finer than P or Q is a refinement of P .

Example 7.1.2 Give example of partition and refinement of the interval [0, 1].

Partitions Norms of Partition

P = {0, 0.5, 1} ∥P ∥ = 0.5
Q = {0, 0.25, 0.5, 0.75, 1} ∥Q∥ = 0.25
R = {0, 0.2, 0.3, , 0.5, 0.6, 0.8, 1} ∥R∥ = 0.2

We see that Q and R are refinements of P but R is not a refinement of Q.

150 CHAPTER 7. INTEGRABILITY ON R

Example 7.1.3 Prove that for each n ∈ N,

{ }
j
Pn = : j = 0, 1, ..., n
n
is a partition of the interval [0, 1] and find a norm of Pn .

Solution. Let n ∈ N. It is easy to see that

0 1 2 n
0= < < < ... < = 1.
n n n n
Thus, Pn is a partition of [0, 1]. We have
j j−1 1
∥Pn ∥ = max − = .
1≤j≤n n n n
Example 7.1.4 (Dyadic Partition) Let n ∈ N and define
{ }
j n
Pn = : j = 0, 1, ..., 2 .
2n
1. Prove that Pn is a partition of the interval [0, 1].

2. Prove that Pm is finer than Pn when m > n.

3. Find a norm of Pn .

Solution. Let n ∈ N. It is easy to see that

0 1 2 2n
0= < < < ... < = 1.
2n 2n 2n 2n
Thus, Pn is a partition of [0, 1]. Next, we will show that Pn ⊆ Pm if m > n.
j
Let m > n and x ∈ Pn . Then there is a j ∈ {0, 1, 2, ..., 2n } such that x = n .
2
Since m > n, m − n > 0. Then 2 m−n
> 0. From 0 ≤ j ≤ 2 , it implies that
n

0 ≤ j · 2m−n ≤ 2n · 2m−n = 2m .

We obtain
j · 2m j · 2m−n
x= = ∈ Pm .
2n · 2m 2m
Thus, Pm is finer than Pn when m > n. We final have
j j−1 1
∥Pn ∥ = max − = .
1≤j≤n 2n 2n 2n
7.1. RIEMANN INTEGRAL 151

UPPER AND LOWER RIEMANN SUM.

Definition 7.1.5 Let a, b ∈ R with a < b, let P = {x0 , x1 , , ..., xn } be a partition of the interval
[a, b], and suppose that f : [a, b] → R is bounded.

1. The upper Riemann sum of f over P is the number

∑
n
U (f, P ) := Mj (f )(xj − xj−1 )
j=1

where
Mj (f ) := sup f (x).
x∈[xj−1 ,xj ]

2. The lower Riemann sum of f over P is the number

∑
n
L(f, P ) := mj (f )(xj − xj−1 )
j=1

where
mj (f ) := inf f (x).
x∈[xj−1 ,xj ]

Y
y = f (x)

Upper Riemann Sum

... ...

X
a = x0 x1 x2 x3 xk−1 xk xn−1 xn = b
Y
y = f (x)

Lower Riemann Sum

... ...

X
a = x0 x1 x2 x3 xk−1 xk xn−1 xn = b
152 CHAPTER 7. INTEGRABILITY ON R

Example 7.1.6 Let f (x) = x2 + 1 where x ∈ [0, 1]. Find L(f, P ) and U (f, P )
{ }
1 1 3
1. P = 0, , , , 1
4 2 4

Y f (x) = x2 + 1 Y f (x) = x2 + 1
2 2

1 1

X X
0 1 1 3 1 0 1 1 3 1
4 2 4 4 2 4

( ) ( ) ( )
1 1 1 1 1 1 3
L(P, f ) = f (0) + f + f + f
4 4 4 4 2 4 4
( )
1 17 5 25 79
= 1+ + + =
4 16 4 16 64
( ) ( ) ( )
1 1 1 1 1 3 1
U (P, f ) = f + f + f + f (1)
4 4 4 2 4 4 4
( )
1 17 5 25 47
= + + +2 =
4 16 4 16 32

2. P = {0, 0.2, 0.5, 0.6, 0.8, 1}

Y f (x) = x2 + 1 Y f (x) = x2 + 1
2 2

1 1

X X
0 0.2 0.5 0.6 0.8 1 0 0.2 0.5 0.6 0.8 1

L(P, f ) = 0.2f (0) + 0.3f (0.2) + 0.1f (0.5) + 0.2f (0.6) + 0.2f (0.8)

= 0.2(1) + 0.3(1.04) + 0.1(1.25) + 0.2(1.36) + 0.2(1.64)

= 1.237

U (P, f ) = 0.2f (0.2) + 0.3f (0.5) + 0.1f (0.6) + 0.2f (0.8) + 0.2f (1)

= 0.2(1.04) + 0.3(1.25) + 0.1(1.36) + 0.2(1.64) + 0.2(2)

= 1.447
7.1. RIEMANN INTEGRAL 153

Example 7.1.7 Let f (x) = x2 + 1 where x ∈ [0, 1]. Find L(Pn , f ) and U (Pn , f ) for n ∈ N if
{ }
j
Pn = : j = 0, 1, ..., n .
n

k 1
Solution. Let xk = and ∆xk = xk − xk−1 = for each k = 0, 1, 2, ..., n.
n n

2 f (x) = x2 + 1

1
... ...

X
0 1 2 3 k−1 k n−1 1
n n n n n n

For interval [xk−1 , xk ] and f is increasing on [0, 1], it follows that

() ( )2
k−1 k−1 1
mk = f (xk−1 ) = f = + 1 = 2 (k − 1)2 + 1
n n n
( ) ( )2
k k 1
mk = f (xk ) = f = + 1 = 2 · k2 + 1
n n n

Thus, we obtain
∑ ∑n [ ]
1 ∑ 1∑
n n n
1 1
L(Pn , f ) = mk ∆xk = (k − 1) + 1
2
= 3 (k − 1) +
2
1
k=1 k=1
n2 n n k=1 n k=1
1 2 1
= [0 + 1 2
+ 22
+ · · · + (n − 1) 2
] + ·n
n3 n
1 (n − 1)(n − 1 + 1)(2(n − 1) + 1)
= 3· +1
n 6
(n − 1)(n)(2n − 1) (n − 1)(2n − 1)
= 3
+1= +1
6n 6n2

and
∑ ∑n [ ]
1 ∑ 2 1∑
n n n
1 1
U (Pn , f ) = Mk ∆xk = ·k +1
2
= 3 k + 1
k=1 k=1
n2 n n k=1 n k=1
1 n(n + 1)(2n + 1) 1
= · + ·n
n3 6 n
(n + 1)(2n + 1)
= + 1.
6n2
154 CHAPTER 7. INTEGRABILITY ON R

Theorem 7.1.8 L(f, P ) ≤ U (f, P ) for all partition P and all bounded function f .

Proof. Let P = {x0 , x1 , ..., xn } be a partition and f be bounded on [a, b]. Then

mj (f ) = inf f (x) ≤ sup f (x) = Mj (f ) for all j = 1, 2, .., n.

x∈[xj−1 ,xj ] x∈[xj−1 ,xj ]

It follows that
∑
n ∑
n
L(f, P ) = mj (f )∆xj ≤ Mj (f )∆xj = U (f, P ).
j=1 j=1

Theorem 7.1.9 (Sum Telescopes) If g : N → R, then

∑
n
[g(k + 1) − g(k)] = g(n + 1) − g(m)
k=m

for all n ≥ m in N.

Proof. Fix m ∈ N. We will prove by induction on n. The Sum Telescopes is obvious for n = 1.
Assume that the Sum Telescopes is true for some n ∈ N. By inductive hypothesis,

∑
n+1 ∑
n
[g(k + 1) − g(k)] = [g(k + 1) − g(k)] + [g(n + 2) − g(n + 1)]
k=m k=m

= g(n + 1) − g(m) + [g(n + 2) − g(n + 1)]

= g(n + 2) − g(m).

The Sum Telescopes is true for some n+1. We conclude that by induction that the Sum Telescopes
holds for n ∈ N.
7.1. RIEMANN INTEGRAL 155

Theorem 7.1.10 If f (x) = α is constant on [a, b], then

U (f, P ) = L(f, P ) = α(b − a)

Proof. Let f (x) = α is constant on [a, b] and let P = {x0 , x1 , x2 , ..., xn } be a partition of [a, b] such
that x0 = a and xn = b.
Y

α
... ...

X
a = x0 x1 x2 x3 xk−1 xk xn−1 xn = b

For each j ∈ {1, 2, ..., n} and f (x) = α, we have

mj (f ) = inf f (x) = α and Mj (f ) = sup f (x) = α.

x∈[xj−1 ,xj ] x∈[xj−1 ,xj ]

Use the Sum Telescopes, we obtain

∑n ∑
n
L(P, f ) = mj (f )∆xj = α(xj − xj−1 ) = α(xn − x0 ) = α(b − a),
j=1 j=1
∑n ∑n
U (P, f ) = Mj (f )∆xj = α(xj − xj−1 ) = α(xn − x0 ) = α(b − a).
j=1 j=1

Theorem 7.1.11 If P is any partition of [a, b] and Q is a refinement of P , then

L(f, P ) ≤ L(f, Q) ≤ U (f, Q) ≤ U (f, P ).

Proof. It is clear that L(f, Q) ≤ U (f, Q) by Theorem 7.1.8.

Let P = {x0 , x1 , x2 , ..., xn } be a partition of [a, b] such that x0 = a and xn = b.
Assume that Q is a refinement of P . Special case Q = P ∪ {c} for some c ∈ (a, b).
If c ∈ P , then Q = P which implies that

L(f, P ) = L(f, Q) ≤ U (f, Q) = U (f, P ).

The proof is done for this case.

156 CHAPTER 7. INTEGRABILITY ON R

Suppose c ∈
/ P . Then there is an xk such that

xk−1 < c < xk for some k ∈ {1, 2, ..., n}.

Consider
∑
k−1 ∑
n
U (f, P ) = Mj (f )∆xj + sup f (x) · (xk − xk−1 ) + Mj (f )∆xj
j=1 x∈[xk−1 ,xk ] j=K+1

∑
k−1 ∑
n
U (f, Q) = Mj (f )∆xj + sup f (x) · (c − xk−1 ) + sup f (x) · (xk − c) + Mj (f )∆xj
j=1 x∈[xk−1 ,c] x∈[c,xk ] j=k+1

Set M = sup f (x). Then

x∈[xk−1 ,xk ]

sup f (x) ≤ M and sup f (x) ≤ M .

x∈[xk−1 ,c] x∈[c,xk ,c]

We obtain

U (f, P ) − U (f, Q) = sup f (x) · (xk − xk−1 ) − sup f (x) · (c − xk−1 ) − sup f (x) · (xk − c)
x∈[xk−1 ,xk ] x∈[xk−1 ,c] x∈[c,xk ]

≥ M (xk − xk−1 ) − M (c − xk−1 ) − M (xk − c)

= M (xk − xk−1 − c + xk−1 − xk + c) = 0.

Thus, U (f, P ) ≥ U (f, Q). A similar argument show that L(f, P ) ≤ L(f, Q).

Corollary 7.1.12 If P and Q are any partitions of [a, b], then

L(f, P ) ≤ U (f, Q).

Proof. Assume that P and Q are any partitions of [a, b]. Then

P ⊆P ∪Q and Q ⊆ P ∪ Q.

Thus, P ∪ Q is a refinement of P and Q. By Theorem 7.1.11, it implies that

L(f, P ) ≤ L(f, P ∪ Q) ≤ U (f, P ∪ Q) ≤ U (f, P ).

Hence, L(f, P ) ≤ U (f, Q).

7.1. RIEMANN INTEGRAL 157

RIEMANN INTEGRABLE.

Definition 7.1.13 Let a, b ∈ R with a < b.

A function f : [a, b] → R is said to be Riemann integrable or integrable on [a, b] if and only
if f is bounded on [a, b], and for every ε > 0 there is a partition of [a, b] such that

U (f, P ) − L(f, P ) < ε.

Theorem 7.1.14 Suppose that a, b ∈ R with a < b. If f is continuous on the interval [a, b], then
f is integrable on [a, b].

Proof. Let a, b ∈ R with a < b. Assume that f is continuous on the interval [a, b].
It follows that f is bounded on [a, b] by the EVT. Theorem 5.3.6 implies that f is uniformly
continuous on the interval [a, b]. Let ε > 0. There is a δ > 0 such that

ε
|x − y| < δ and x, y ∈ [a, b] imply |f (x) − f (y)| < . (7.1)
b−a

Let P = {x0 , x1 , ..., xn } be a partition of [a, b] such that ∥P ∥ < δ.

Fix j ∈ {1, 2, ..., n}. By agian the EVT, there are xn , xM ∈ [xj−1 , xj ] such that

f (xm ) = mj (f ) and f (xM ) = Mj (f ).

Since ∥P ∥ < δ, we have |xM − xm | ≤ |xj − xj−1 | < δ. Then xm , xM satisfy (7.1), it implies that

ε
|Mj (f ) − mj (f )| = |f (xM ) − f (xm )| < .
b−a

Use the Sum Telescopes, We obtain

∑
n
U (f, P ) − L(f, P ) = (Mj (f ) − mj (f ))(xj − xj−1 )
j=1
∑n
ε
< · (xj − xj−1 )
j=1
b−a
ε ε
= · (xn − x0 ) = · (b − a) = ε.
b−a b−a

Therefore, f is integrable on [a, b].

158 CHAPTER 7. INTEGRABILITY ON R

Example 7.1.15 Prove that the function



1 if 0 ≤ x < 1
2
f (x) =

0 if 1
2
≤x≤1

is integrable on [0, 1].

{ }
1 ε 1 ε
Solution. Let ε > 0. Case ε < 1. Choose P = 0, − , + , 1 .
2 4 2 4

Y
1

X
0 1
2
− ε
4
1
2
1
2
+ ε
4
1

We obtain
[( ) ] [( ) ( )] [ ( )]
1 ε 1 ε 1 ε 1 ε 1 ε
U (f, P ) = 1 − −0 +1 + − − +0 1− + = +
2 4 2 4 2 4 2 4 2 4
[( ) ] [( ) ( )] [ ( )]
1 ε 1 ε 1 ε 1 ε 1 ε
L(f, P ) = 1 − −0 +0 + − − +0 1− + = −
2 4 2 4 2 4 2 4 2 4
ε
U (f, P ) − L(f, P ) = < ε.
2
{ }
1
Case ε ≥ 1. Choose P = 0, , 1 . Then
2
( ) ( )
1 1 1
U (f, P ) = 1 −0 +0 1− =
2 2 2
( ) ( )
1 1
L(f, P ) = 0 −0 +0 1− =0
2 2
1
U (f, P ) − L(f, P ) = < 1 ≤ ε.
2

Thus, f is integrable on [0, 1].

7.1. RIEMANN INTEGRAL 159

Example 7.1.16 (Dirichlet function) Prove that the function



1 if x ∈ Q
f (x) =

0 if x ∈ /Q
is NOT Riemann integrable on [0, 1].

Solution. Suppose that f is Riemann integrable on [0, 1].

Given ε = 21 . There is a partition P = {x0 , x1 , ..., xn } of [0, 1] such that
1
U (f, P ) − L(f, P ) < .
2
Fix j ∈ {1, 2, ..., n}. By real property, it leads to that there are r ∈ Q and s ∈ Qc such that
r, s ∈ [xj−1 , xj ]. It implies that

mj (f ) = inf f (x) = 0 and Mj (f ) = sup f (x) = 1.

x∈[xj−1 ,xj ] x∈[xj−1 ,xj ]

Use the Sum Telescopes, we obtain

∑
n ∑
n
U (f, P ) = Mj (f )∆xj = 1(xj − xj−1 ) = xn − x0 = 1 − 0 = 1
j=1 j=1
∑n ∑n
L(f, P ) = mj (f )∆xj = 0(xj − xj−1 ) = 0
j=1 j=1
1
U (f, P ) − L(f, P ) = 1 − 0 = 1 > ,
2
a contradiction. We conclude that the Dirichlet function is not Riemann integrable on [0, 1].

UPPER AND LOWER INTEGRABLE.

Definition 7.1.17 Let a, b ∈ R with a < b, and f : [a, b] → R be bounded.

1. The upper integral of f on [a, b] is the number

∫ b
(U ) f (x) dx := inf{U (f, P ) : P is a partition of [a, b]}.
a

2. The lower integral of f on [a, b] is the number

∫ b
(L) f (x) dx := sup{L(f, P ) : P is a partition of [a, b]}.
a

3. If the upper and lower integrals of f on [a, b] are equal, we define the integral of f on [a, b]
to be the common value
∫ b ∫ b ∫ b
f (x) dx := (U ) f (x) dx = (L) f (x) dx.
a a a
160 CHAPTER 7. INTEGRABILITY ON R

Example 7.1.18 Let f (x) = α where x ∈ [a, b]. Show that

∫ b ∫ b
(U ) f (x) dx = (L) f (x) dx = α(b − a).
a a

Solution. By Theorem 7.1.10, for any partition of [a, b], we have U (f, P ) = L(f, P ) = α(b − a).
It follows that
∫ b
(U ) f (x) dx = inf U (f, P ) = α(b − a),
a P
∫ b
(L) f (x) dx = sup L(f, P ) = α(b − a).
a P

Example 7.1.19 The Dirichlet function is defined



1 if x ∈ Q
f (x) = .

0 if x ∈
/Q
Find the upper integral and lower integral of the Dirichlet function on [0, 1].

Solution. By Example 7.1.16, for any partition of [a, b], we have U (f, P ) = 1 and L(f, P ) = 0. It
follows that
∫ b
(U ) f (x) dx = inf U (f, P ) = 1,
a P
∫ b
(L) f (x) dx = sup L(f, P ) = 0.
a P

Theorem 7.1.20 If f : [a, b] → R is bounded, then its upper and lower integrals exist and are
finite, and satisfy ∫ ∫
b b
(L) f (x) dx ≤ (U ) f (x) dx.
a a

Proof. By Corollary 7.1.12, we have

L(f, P ) ≤ U (f, Q) for partitions P, Q of [a, b].

We obtain by taking supremum over all partitions P of [a, b],

∫ b
(L) f (x) dx = sup L(f, P ) ≤ sup U (f, Q) = U (f, Q).
a P P

Taking infimum over all partitions Q of [a, b], we have

∫ b ∫ b
(L) f (x) dx ≤ inf U (f, Q) = (U ) f (x) dx.
a Q a
∫ b ∫ b
Hence, (L) f (x) dx ≤ (U ) f (x) dx.
a a
7.1. RIEMANN INTEGRAL 161

Theorem 7.1.21 Let a, b ∈ R with a < b, and f : [a, b] → R be bounded. Then f is integrable on
[a, b] if and only if ∫ ∫
b b
(L) f (x) dx = (U ) f (x) dx.
a a

Proof. Let a, b ∈ R with a < b, and f : [a, b] → R be bounded.

Assume that f is integrable on [a, b]. Let ε > 0. There is a partition P of [a, b] such that

U (f, P ) − L(f, P ) < ε.

By definition,
∫ b ∫ b
L(f, P ) ≤ (L) f (x) dx and (U ) f (x) dx ≤ U (f, P ).
a a

By Theorem 7.1.20, it follows that

∫ b ∫ b ∫ b ∫ b
(U ) f (x) dx − (L) f (x) dx = (U ) f (x) dx − (L) f (x) dx
a a a a

≤ U (f, P ) − L(f, P ) < ε.

∫ b ∫ b
Thus, (L) f (x) dx = (U )
f (x) dx.
a ∫ b a ∫ b
Conversely, we assume that (L) f (x) dx = (U ) f (x) dx.
a a
Let ε > 0. Choose, by the API and APS, partitions P1 , P2 of [a, b] such that
∫ b ∫ b
ε ε
(L) f (x) dx − < L(f, P1 ) and U (f, P2 ) < (U ) f (x) dx + .
a 2 a 2
Set P = P1 ∪ P2 . Then P is a refinement of P1 and P2 . By Theorem 7.1.11, it follows that

U (P, f ) − L(f, P ) ≤ U (f, P2 ) − L(f, P1 )

( ∫ b ) ( ∫ b )
ε ε
< (U ) f (x) dx + − (L) f (x) dx − = ε.
a 2 a 2

Therefore, f is integrable on [a, b].

162 CHAPTER 7. INTEGRABILITY ON R

Theorem 7.1.22 For a constant α,

∫ b
α dx = α(b − a).
a

Proof. It is easy to prove by Example 7.1.18 and Theorem 7.1.21.

Example 7.1.23 Let f : [0, 2] → R defined by



0 if x ̸= 1
f (x) =

3 if x = 1
∫ 2
Show that f is integrable and find f (x)dx.
0
ε
Solution. Let ε > 0. Let P = {x0 , x1 , ..., xn } be a partition of [0, 2] such that ∥P ∥ < . Then
6
mj (f ) = inf f (x) = 0 for all j = 1, 2, ..., n.
x∈[xj−1 ,xj ]

∑
n
We obtain L(f, P ) = mj (f )∆xj = 0 which is not depend on ε. So,
j=1
∫ 2
(L) f (x) dx = sup L(f, P ) = 0.
0 P

Case 1 ∈ P . Then xk = 1 for some k ∈ {1, 2, ..., n − 1}. We have

Mj (f ) = inf f (x) = 0 for all j ̸= k, k + 1 and Mk (f ) = 3, Mk+1 (f ) = 3.

x∈[xj−1 ,xj ]
ε ε
From ∥P ∥ < , it follows that |xj − xj−1 | < for all j = 1, 2, ..., n. We obtain
6 6
U (f, P ) − L(f, P ) = U (f, P ) − 0
∑
n
ε ε
= Mj (f )∆xj = 3(xk − xk−1 ) + 3(xk+1 − xk ) < 3 · + 3 · = ε.
j=1
6 6

Case 1 ∈
/ P . Then 1 ∈ [xk−1 , xk ] for some k ∈ {1, 2, ..., n}. We have

Mj (f ) = inf f (x) = 0 for all j ̸= k and Mk (f ) = 3.

x∈[xj−1 ,xj ]

We obtain
∑
n
ε ε
U (f, P ) − L(f, P ) = U (f, P ) − 0 = Mj (f )∆xj = 3(xk − xk−1 ) < 3 · = < ε.
j=1
6 2
Thus, f is integrable on [0, 2] and
∫ 2 ∫ 2
f (x) dx = (L) f (x) dx = 0.
0 0
7.1. RIEMANN INTEGRAL 163

Example 7.1.24 Let f : [0, 1] → R defined by



2 if x ̸= 1
f (x) =

1 if x = 1
∫ 1
Show that f is integrable and find f (x)dx.
0

Solution. Let ε > 0. Let P = {x0 , x1 , ..., xn } be a partition of [0, 1] such that ∥P ∥ < ε.
Y
2
1
X
0 1

Then, Mj (f ) = sup f (x) = 2 for all j = 1, 2, ..., n.

x∈[xj−1 ,xj ]
We obtain
∑
n ∑
n
U (f, P ) = Mj (f )∆xj = 2(xj − xj−1 ) = 2(xn − x0 ) = 2(1 − 0) = 2
j=1 j=1

which is not depend on ε. So,

∫ 2
(U ) f (x) dx = inf U (f, P ) = 2.
0 P

We see that

mj (f ) = inf f (x) = 2 for all j ̸= n and Mn (f ) = 1.

x∈[xj−1 ,xj ]

From ∥P ∥ < ε, it follows that |xj − xj−1 | < ε for all j = 1, 2, ..., n. We obtain

U (f, P ) − L(f, P ) = 2 − L(f, P )

∑
n ∑
n−1
=2− mj (f )∆xj = 2 − 2(xj − xj−1 ) − 1(xn − xn−1 )
j=1 j=1

= 2 − 2(xn−1 − x0 ) − 1(xn − xn−1 )

= 2 − 2(xn−1 − 0) − 1(1 − xn−1 ) = 1 − xn−1 = xn − xn−1 < ε.

Thus, f is integrable on [0, 1] and

∫ 1 ∫ 2
f (x) dx = (U ) f (x) dx = 2.
0 0
164 CHAPTER 7. INTEGRABILITY ON R

Exercises 7.1
∫ 1
1. For each of the following, compute U (f, P ), L(f, P ), and f (x) dx, where
0
{ }
2 1 3
P = 0, , , , 1 .
5 2 5
Find out whether the lower sum or the upper sum is better approximation to the integral.
Graph f and explain why this is so.

1.1 f (x) = 1 − x2 1.2 f (x) = 2x2 + 1 1.3 f (x) = x2 − x

{ }
j
2. Let Pn = : n = 0, 1, ..., n for each n ∈ N. Prove that a bounded function f is integrable
n
on [0, 1] if
I0 := lim L(f, Pn ) = lim U (f, Pn ),
n→∞ n→∞
∫ 1
in which case f (x) dx equals I0 .
0

3. For each of the following functions, use Pn in 2. to find∫ formulas for the upper and lower
1
sums of f on Pn , and use them to compute the value of f (x) dx.
0

3.1 f (x) = x 
1 if 0 ≤ x < 1
2
3.3 f (x) =

0
3.2 f (x) = x 2 if 1
2
≤x≤1

{ } 
1
1 if x ∈ E
4. Let E = : n ∈ N . Prove that the function f (x) = is integrable on
n 
0 if otherwise
∫ 1
[0, 1]. What is the value of f (x) dx ?
0
∫ c
5. Suppose that f is continuous on an interval [a, b]. Show that f (x) dx = 0 for all c ∈ [a, b]
a
if and only if f (x) = 0 for all x ∈ [a, b].

6. Let f be bounded on a nondegenerate interval [a, b]. Prove that f is integrable on [a, b] if
and only if given ε > 0 there is a partition Pε of [a, b] such that

P ⊇ Pε imples |U (f, P ) − L(f, P )| < ε.

7.2. RIEMANN SUMS 165

7.2 Riemann sums

Definition 7.2.1 Let f : [a, b] → R.

1. A Riemann sum of f with respect to a partition P = {x0 , x1 , ..., xn } of [a, b] is a sum of

the form
∑
n
f (tj )∆xj ,
j=1

where the choice of tj ∈ [xj−1 , xj ] is arbitrary.

2. The Riemann sums of f are converge to I(f ) as ∥P ∥ → 0 if and only if given ε > 0 there
is a partition Pε of [a, b] such that

∑
n
P = {x0 , x1 , ..., xn } ⊇ Pε implies f (tj )∆xj − I(f ) < ε
j=1

for all choice of tj ∈ [xj−1 , xj ], j = 1, 2, ..., n. In this case we shall use the notation

∑
n
I(f ) = lim f (tj )∆xj .
∥P ∥→0
j=1

{ }
j
Example 7.2.2 Let f (x) = x2 where x ∈ [0, 1] and P = : j = 0, 1, ..., n be a partition of
n
[0, 1]. Show that if f (ti ) is choosen by the right end point and left end point in each subinterval,
then two I(f ), depend on two methods, are NOT different.

Solution. The Right End Point : Choose f (tj ) = f ( nj ) on the subinterval [xj−1 , xj ]
j j−1 1
and have ∆xj = − = for all j = 1, 2, 3, ..., n. We obtain
n n n
∑ ∑ ( ) n ( )2
1∑ j 1 ∑ 2
n n n
j 1
f (tj )∆xj = f = = 3 j
j=1 j=1
n n n j=1 n n j=1
1 n(n + 1)(2n + 1) (n + 1)(2n + 1)
= 3
· = .
n 6 6n2

Thus,

∑
n
(n + 1)(2n + 1) 1 1
I(f ) = lim f (tj )∆xj = lim 2
= = .
∥P ∥→0
j=1
n→∞ 6n 3 3
166 CHAPTER 7. INTEGRABILITY ON R

The Left End Point : Choose f (tj ) = f ( j−1

n
) on the subinterval [xj−1 , xj ]. We obtain

∑ ∑ ( ) n ( )2
1 ∑ j−1 1 ∑
n n n
j−1 1
f (tj )∆xj = f = = 3 (j − 1)2
j=1 j=1
n n n j=1
n n j=1
1 [ 2 ]
= 0 + 12 + 22 + · · · (n − 1)2
n3
1 (n − 1)(n)(2(n − 1) + 1) (n − 1)(2n − 1)
= 3 · = .
n 6 6n2

Thus,
∑
n
(n − 1)(2n − 1) 1
I(f ) = lim f (tj )∆xj = lim 2
= .
∥P ∥→0
j=1
n→∞ 6n 3

Theorem 7.2.3 Let a, b ∈ R with a < b, and suppose that f : [a, b] → R is bounded. Then f is
Riemann integrable on [a, b] if and only if

∑
n
I(f ) = lim f (tj )∆xj
∥P ∥→0
j=1

exists, in which case ∫ b

I(f ) = f (x) dx.
a

Proof. Assume that f is Riemann integrable on [a, b].

Let ε > 0. By the API and APS, there is a partition Pε of [a, b] such that
∫ b ∫ b
f (x) dx + ε < L(f, Pε ) and U (f, Pε ) < (U ) f (x) dx + ε.
a a

Let P = {x0 , x1 , ..., xn } ⊇ Pε . From mj (f ) ≤ f (tj ) ≤ Mj (f ) for any choice of tj ∈ [xj−1 , xj ].

Hence,
∫ b ∑
n
f (x) dx − ε < L(f, Pε ) < L(f, P ) ≤ f (tj )∆xj
a j=1
∫ b
≤ U (f, P ) < U (f, Pε ) < f (x) dx + ε.
a

It implies that
∑
n ∫ b
f (tj )∆xj − f (x) dx < ε.
j=1 a

for all partitions P ⊇ Pε and all choices of tj ∈ [xj−1 , xj ], j = 1, 2, ..., n.

7.2. RIEMANN SUMS 167

Conversely, assume that the Riemann sums of converge to I(f ). Let ε > 0 and choose a
partition P = {x0 , x1 , ..., xn } of [a, b] such that

∑
n
ε
f (tj )∆xj − I(f ) < (7.2)
j=1
3

for all choices of tj ∈ [xj−1 , xj ]. By the API and APS, choose uj , vj ∈ [xj−1 , xj ] such that
ε ε
Mj (f ) − < f (uj ) and f (vj ) < mj (f ) +
6(b − a) 6(b − a)
It implies that

ε ε ε
f (uj ) − f (vj ) > Mj (f ) − − mj (f ) − = Mj (f ) − mj (f ) − .
6(b − a) 6(b − a) 3(b − a)

So,
ε
Mj (f ) − mj (f ) < f (uj ) − f (vj ) + .
3(b − a)
By (7.2) and telescoping, we have

∑
n
U (f, P ) − L(f, P ) = (Mj (f ) − mj (f ))∆xj
j=1
∑n ∑
n
ε ∑ n
< f (uj )∆xj − f (vj )∆xj + (xj − xj−1 )
j=1 j=1
3(b − a) j=1
∑
n ∑
n
ε
≤ f (uj )∆xj − f (vj )∆xj + (xn − x0 )
j=1 j=1
3(b − a)
∑
n ∑
n
ε
= f (uj )∆xj − I(f ) − f (vj )∆xj + I(f ) + (b − a)
j=1 j=1
3(b − a)
∑
n ∑
n
ε
≤ f (uj )∆xj − I(f ) + f (vj )∆xj − I(f ) +
j=1 j=1
3
ε ε ε
< + + = ε.
3 3 3

Thus, f is Riemann integrable on [a, b].

168 CHAPTER 7. INTEGRABILITY ON R

Theorem 7.2.4 (Linear Property) If f, g are integrable on [a, b] and α ∈ R, then f + g and αf
are integrable on [a, b]. In fact,
∫ b ∫ b ∫ b
1. (f (x) + g(x)) dx = f (x) dx + g(x) dx
a a a
∫ b ∫ b
2. αf (x) dx = α f (x) dx
a a

Proof. Assume that f and g are integrable on [a, b] and α ∈ R.

Let ε > 0 and choose Pε such that for any partition P = {x0 , x1 , ..., xn } ⊇ Pε of [a, b] and any
choice of tj ∈ [xj−1 , xj ], we have
∑n ∫ b ∑
n ∫ b
ε ε
f (tj )∆xj − f (x) dx < and g(tj )∆xj − g(x) dx < .
j=1 a 2 j=1 a 2

By triangle inequality, for any choice tj ∈ [xj−1 , xj ],

∑
n ∫ b ∫ b ∑
n ∑
n ∫ b ∫ b
(f + g)(tj )∆xj − f (x) dx − g(x) dx = f (tj )∆xj + g(tj )∆xj − f (x) dx − f (x) dx
j=1 a a j=1 j=1 a a

∑
n ∫ b ∑
n ∫ b
≤ f (tj )∆xj − f (x) dx + g(tj )∆xj − g(x) dx
j=1 a j=1 a

ε ε
< + = ε.
2 2
∫ b ∫ b ∫ b
We conclude that f + g is integrable on [a, b] and (f (x) + g(x)) dx = f (x) dx + g(x) dx.
a a a
Similarly, if Pε is chosen so that if P = {x0 , x1 , ..., xn } is finer than Pε , then
∑n ∫ b
ε
f (tj )∆xj − f (x) dx < .
j=1 a |α| + 1

It is easy to see that, for any choice tj ∈ [xj−1 , xj ],

∑
n ∫ b ∑
n ∫ b
αf (tj )∆xj − α f (x) dx = |α| f (tj )∆xj − f (x) dx
j=1 a j=1 a

ε
< |α| · < ε.
|α| + 1
∫ b ∫ b
Thus, αf is integrable on [a, b] and αf (x) dx = α f (x) dx.
a a
7.2. RIEMANN SUMS 169

Theorem 7.2.5 If f is integrable on [a, b], then f is integrable on each subinterval [c, d] of [a, b].
Moreover, ∫ ∫ ∫
b c b
f (x) dx = f (x) dx + f (x) dx
a a c

for all c ∈ (a, b).

Proof. We may suppose that a < b. Let ε > 0 and choose a partition P of [a, b] such that

U (f, P ) − L(f, P ) < ε.

Let P0 = P ∪ {c} and P1 = P0 ∩ [a, c]. Since P1 is a partition of [a, c] and P0 is a refinement of P ,
we have
U (f, P1 ) − L(f, P1 ) ≤ U (f, P0 ) − L(f, P0 ) ≤ U (f, P ) − L(f, P ) < ε.

Therefore, f is integrable on [a, c]. A similar argument proves that f is integrable on any subinterval
[c, d] of [a, b].
Let P2 = P0 ∩ [c, d]. Then P0 = P1 ∪ P2 and by definition

U (f, P ) ≥ U (f, P0 ) = U (f, P1 ) + U (f, P2 )

∫ c ∫ b ∫ c ∫ b
≥ (U ) f (x) dx + (U ) f (x) dx = f (x) dx + f (x) dx.
a c a c

Next, we will take infimum of the last inequality over all partitions P of [a, b], we obtain
∫ b ∫ b
f (x) dx = (U ) f (x) dx
a a
∫ c ∫ b
= inf U (f, P ) ≥ f (x) dx + f (x) dx.
P a c

A similar argument using lower integrals shows that

∫ b ∫ c ∫ b
f (x) dx ≤ f (x) dx + f (x) dx.
a a c
∫ b ∫ c ∫ b
We conclude that f (x) dx = f (x) dx + f (x) dx.
a a c
170 CHAPTER 7. INTEGRABILITY ON R

By Theorem 7.2.5, we obtain

∫ b ∫ a ∫ b
f (x) dx = f (x) dx + f (x) dx
a a a

Thus,
∫ a ∫ b ∫ a
f (x) dx = 0 and f (x) dx = − f (x) dx.
a a b
∫ 5
Example 7.2.6 Using the connection between integrals are area, evaluate |x − 2| dx.
0

Solution. Define f (x) = |x − 2| where x ∈ [0, 5].

Y
y = |x − 2|

X
0 1 2 3 4 5

∫ 5 ∫ 5
1 1 13
f (x) dx = ·2·2+ ·3·3= |x − 2| dx =
0 0 2 2 2
∫ 2 √
Example 7.2.7 Using the connection between integrals are area, evaluate 4 − x2 dx.
0
√
Solution. Define f (x) = 4 − x2 where x ∈ [0, 2].

2
√
y= 4 − x2
1

X
0 1 2

∫ 2 ∫ 2 √ 1
f (x) dx = 4 − x2 dx = π(2)2 = π
0 0 4
7.2. RIEMANN SUMS 171

Theorem 7.2.8 (Comparison Theorem) If f, g are integrable on [a, b] and f (x) ≤ g(x) for all
x ∈ [a, b], then ∫ ∫
b b
f (x) dx ≤ g(x) dx.
a a

In particular, if m ≤ f (x) ≤ M for x ∈ [a, b], then

∫ b
m(b − a) ≤ f (x) dx ≤ M (b − a).
a

Proof. Assume that f, g are integrable on [a, b] and f (x) ≤ g(x) for all x ∈ [a, b].
Let P be a partition of [a, b]. By hypothesis, Mj (f ) ≥ Mj (g) whence U (f, P ) ≤ U (g, P ).
It follows that ∫ ∫
b b
f (x) dx = (U ) f (x) dx ≤ U (f, P ) ≤ U (g, P )
a a

for all partition P of [a, b]. Taking the infimum of this inequality over all partition P of [a, b],
we have ∫ ∫ ∫
b b b
f (x) dx ≤ inf U (g, P ) = (U ) g(x) dx = g(x) dx.
a P a a

If m ≤ f (x) ≤ M , then by Theorem 7.1.22

∫ b ∫ b ∫ b
m(b − a) = m dx ≤ f (x) dx ≤ M dx = M (b − a).
a a a

Theorem 7.2.9 If f is Riemann integrable on [a, b], then |f | is integrable on [a, b] and
∫ b ∫ b
f (x) dx ≤ |f (x)| dx.
a a

Proof. Assume that f is Riemann integrable on [a, b]. Let P = {x0 , x1 , ..., xn } be a partition of
[a, b] and let x, y ∈ [xj−1 , xj ] for j = 1, 2, ..., n. If f (x), f (y) have the same sign, say both are
positive, then
|f (x)| − |f (y)| = f (x) − f (y) ≤ Mj (f ) − mj (f ).

If f (x), f (y) have opposite signs, f (x) ≥ 0 ≥ f (y), then mj (f ) ≤ 0, hence

|f (x)| − |f (y)| = f (x) + f (y) ≤ Mj (f ) + 0 ≤ Mj (f ) − mj (f ).

172 CHAPTER 7. INTEGRABILITY ON R

It implies that

Mj (|f |) − mj (|f |) ≤ Mj (f ) − mj (f ). (7.3)

Let ε > 0 and choose a partition P of [a, b] such that

U (f, P ) − L(f, P ) < ε.

Since (7.3) implies that U (|f |, P ) − L(|f |, P ) ≤ U (f, P ) − L(f, P ), it follows that

U (|f |, P ) − L(|f |, P ) < ε.

Thus, |f | is Riemann integrable on [a, b]. Since −|f (x)| ≤ f (x) ≤ |f (x)| holds for any x ∈ [a, b],
we conclude by Theorem 7.2.8 that
∫ b ∫ b ∫ b
− |f (x)| dx ≤ f (x) dx ≤ |f (x)| dx.
a a a
∫ b ∫ b
Hence, f (x) dx ≤ |f (x)| dx.
a a
7.2. RIEMANN SUMS 173

Exercises 7.2

1. Using the connection between integrals are area, evaluate each of the following integrals.
∫ 1 ∫ 2
1.1 |x − 0.5| dx 1.3 (|x + 1| + |x|) dx
0 −2
∫ a √ ∫ b
1.2 a2 − x2 dx, a>0 1.4 (3x + 1) dx, a<b
0 a

2. Prove that if f is integrable on [0, 1] and β > 0, then

∫ 1
nβ
α
lim n f (x) dx = 0 for all α < β.
n→∞ 0

3. If f, g are integrable on [a, b] and α ∈ R, prove that

∫ b ∫ b ∫ b
(f (x) + g(x)) dx ≤ |f (x)| dx + |g(x)| dx.
a a a

∫ b
4. Suppose that gn ≥ 0 is a sequence of integrable function that satisfies lim gn (x) dx = 0.
n→∞ a
∫ b
Show that if f : [a, b] → R is integrable on [a, b], then lim f (x)gn (x) dx = 0.
n→∞ a
∫ 1
5. Prove that if f is integrable on [0, 1], then lim xn f (x) dx = 0.
n→∞ 0

6. Prove that if f is integrable on [0, 1], then

∫ b n ∫
∑ 1
2k
f (x) dx = lim f (x) dx.
a n→∞ 1
k=0 2k+1

7. Let f be continuous on a closed, nondegenerate interval [a, b] and set M = sup |f (x)|.
x∈[a,b]

7.1 Prove that if M > 0 and p > 0, then for every ε > 0 there is a nondegenerate on interval
I ⊂ [a, b] such that
∫ b
(M − ε) |I| ≤ p
|f (x)|p dx ≤ M p (b − a).
a

(∫ b ) p1
7.2 Prove that lim |f (x)|p dx = M.
p→∞ a
174 CHAPTER 7. INTEGRABILITY ON R

7.3 Fundamental Theorem of Calculus

df
Define a set C 1 [a, b] = {f : [a, b] → R : f is differentiable and f ′ are continuous } and f ′ (x) = .
dx

Theorem 7.3.1 (Fundamental Theorem of Calculus) Suppose that f : [a, b] → R.

∫ x
1. If f is continuous on [a, b] and F (x) = f (t) dt, then F ∈ C 1 [a, b] and
a
∫ x
d
f (t) dt := F ′ (x) = f (x)
dx a

for each x ∈ [a, b].

2. If f is differentiable on [a, b] and f ′ is integrable on [a, b], then

∫ x
f ′ (t) dt = f (x) − f (a)
a

for each x ∈ [a, b].

∫ x
Proof. Assume that f is continuous on [a, b] and F (x) = f (t) dt where x ∈ [a, b].
a
Let x0 ∈ [a, b). Then f (x) → f (x0 ) as x → x+
0 . Let ε > 0. There is a δ > 0 such that

x0 < t < x0 + δ and t ∈ [a, b] imply |f (t) − f (x0 )| < ε. (7.4)

Fix h such that 0 < h < δ. Use Theorem 7.1.22, We have

F (x0 + h) − F (x0 ) 1 1 1
− f (x0 ) = F (x0 + h) − F (x0 ) − f (x0 ) · h
h h h h
∫ x0 +h ∫ x0 ∫
1 1 1 x0 +h
= f (t) dt − f (t) dt − f (x0 ) dt
h a h a h x0
∫ ∫ ∫ ∫
1 x0 1 x0 +h 1 x0 1 x0 +h
= f (t) dt + f (t) dt − f (t) dt − f (x0 ) dt
h a h x0 h a h x0
∫
1 x0 +h
= (f (t) − f (x0 )) dt
h x0

By (7.4) and Theorem 7.2.9 , it implies that

∫ ∫
F (x0 + h) − F (x0 ) 1 x0 +h 1 x0 +h
− f (x0 ) ≤ |f (t) − f (x0 )| dx < ε dx = ε.
h h x0 h x0

F (x0 + h) − F (x0 )
Thus, F ′ (x0 ) = lim = f (x0 ). The proof of part 1 is complete.
x→x0 h
7.3. FUNDAMENTAL THEOREM OF CALCULUS 175

2. Assume that f is differentiable on [a, b] and f ′ is integrable on [a, b]. Let ε > 0 and choose
a partition P = {x0 , x1 , ..., xn } of [a, b] such that
∑
n ∫ b
′
f (tj )∆xj − f ′ (x) dx < ε
j=1 a

for any choice of points tj ∈ [xj−1 , xj ]. Use the MVT to choose points tj ∈ [xj−1 , xj ] such that

f (xj ) − f (xj−1 ) = f ′ (tj )(xj − xj−1 ) = f ′ (tj )∆xj .

∑
n
It follows by telescoping that (f (xj ) − f (xj−1 )) = f (b) − f (a) and
j=1

∫ b ∑
n ∫ b
′
f (b) − f (a) − f (t) dt = (f (xj ) − f (xj−1 )) − f ′ (t) dt
a j=1 a

∑
n ∫ b
′
= f (tj )∆xj − f ′ (t) dt < ε.
j=1 a

∫ b
Thus, f ′ (t) dt = f (b) − f (a) for case x = b. It suffices to prove part 2.
a

Example 7.3.2 Assume that f is differentiable on (0, 1) and integrable on [0, 1]. Show that
∫ 1
xf ′ (x) + f (x) dx = f (1).
0

Solution. By the Product Rule, we have (xf (x))′ = xf ′ (x) + f (x).

Apply the Fundamental Theorem of Calculus,
∫ 1 ∫ 1
′
xf (x) + f (x) dx = (xf (x))′ dx = 1f (1) − 0f (0) = f (1).
0 0

Theorem 7.3.3 Let α ̸= −1. Then

∫ b
xα+1
xα dx = f (b) − f (a) where f (x) = .
a α+1

Proof. Let α ̸= −1. The f ′ (x) = xα . By part 2 of the Fundamental Theorem of Calculus, we
obtain this Theorem.
∫ 1
Example 7.3.4 Find integral x2 dx.
0

Solution. By the Power Rule, we have

∫ 1
13 03 1
x2 dx = − = .
0 3 3 3
176 CHAPTER 7. INTEGRABILITY ON R

Theorem 7.3.5 Suppose that f, u : [a, b] → R. If f is continuous on [a, b] and

∫ u(x)
F (x) = f (t) dt, and F ∈ C 1 [a, b] and
a
∫ u(x)
′ d
F (x) = f (t) dt = f (u(x)) · u′ (x)
dx a

for each x ∈ [a, b].

Proof. Apply the Chain Rule.

∫ sin x
et dt. Find F (0) and F ′ (0).
2
Example 7.3.6 Let F (x) =
0
∫ 0
2
Solution. We obtain F (0) = et dt = 0 and by Theorem 7.3.5, it implies that
0
∫ sin x
′ d
et dt = e(sin x) · (sin x)′ = esin
2 2 2
F (x) = x
· cos x.
dx 0

Thus, F ′ (0) = 1.

INTEGRATION BY PART.

Theorem 7.3.7 (Integration by Part) Suppose that f, g are differentiable on [a, b] with f ′ , g ′
integrable on [a, b], Then
∫ b ∫ b
′
f (x)g(x) dx = f (b)g(b) − f (a)g(a) − f (x)g ′ (x) dx.
a a

Proof. Assume that f, g are differentiable on [a, b] with f ′ , g ′ integrable on [a, b]. By the Product
Rule, (f g)′ (x) = f ′ (x)g(x) + f (x)g ′ (x) for x ∈ [a, b]. It implies that (f g)′ is integrable on [a, b].
Thus, by the part 2 of the Fundamental Theorem of Calculus, we obtain
∫ b ∫ b ∫ b
′ ′
f (x)g(x) dx = (f g) (x) dx − f (x)g ′ (x) dx
a a a
∫ b ∫ b
′
f (x)g(x) dx = f (b)g(b) − f (a)g(a) − f (x)g ′ (x) dx.
a a

The proof is complete.

7.3. FUNDAMENTAL THEOREM OF CALCULUS 177

Example 7.3.8 Use the Integration by Part to find integrals.

∫ π ∫ 2
2
1. x sin x dx 2. ln x dx
0 1

Solution. By the Integration by Part and The Fundamental Theorem of Calculus, we have
∫ π ∫ π ∫ π
2 2
′ π π 2
x sin x dx = x(− cos x) dx = (− cos ) − 0(− cos 0) − (x)′ (− cos x) dx
0 0 2 2 0
∫ π ∫ π
2 2 π
= cos x dx = (sin x)′ dx = sin − sin 0 = 1.
2
∫ 2 ∫0 2 0
∫ 2
ln x dx = (x)′ ln x dx = 2 ln 2 − 1 ln 1 − x(ln x)′ dx
1 1
∫ 2 ∫ 21
1
= 2 ln 2 − x · dx = 2 ln 2 − 1 dx
1 x 1
∫ 2
= 2 ln 2 − (x)′ dx = 2 ln 2 − (2 − 1) = 2 ln 2 − 1.
1

∫ x3
2
Example 7.3.9 Let f (x) = et dt. Use integration by part to show that
0
∫ 1 ∫ 1
2
6 x f (x)dx − 2
2
ex dx = 1 − e.
0 0

Solution. By the Theorem 7.3.5, f ′ (x) = e(x ) · (x3 )′ = 3x2 ex . We obtain

3 2 6

∫ 1 ∫ 1
2
6 x f (x)dx = 2 (3x2 )f (x)dx
0
∫0 1
=2 (x3 )′ f (x)dx
(0 ∫ 1 )
3 ′
= 2 1f (1) − 0f (0) − x f (x)dx
0
( ∫ 1 )
2 x6
= 2 f (1) − 3
x (3x e )dx
0
∫ 1
6
= 2f (1) − 6x5 ex dx
0
∫ 1 ∫ 1
x2
(ex )′ dx
6
=2 e dx −
∫0 1 0
2
=2 ex dx − [e − 1]
0
∫ 1 ∫ 1
2
We conclude that 6 x2 f (x)dx − 2 ex dx = 1 − e.
0 0
178 CHAPTER 7. INTEGRABILITY ON R

CHANGE OF VARIABLES.

Theorem 7.3.10 (Change of Variables) Let ϕ be continuously differentiable on a closed interval

[a, b]. If f is continuous on ϕ([a, b]), or if ϕ is strictly incresing on [a, b] and f is integrable on
[ϕ(a), ϕ(b)], then
∫ ϕ(b) ∫ b
f (t) dt = f (ϕ(x))ϕ′ (x) dx.
ϕ(a) a

Proof. Exercise.
∫ 3
√
e x+1
Example 7.3.11 Find √ dx
0 x+1
√ 1
Solution. Let f (x) = ex and ϕ(x) = x + 1 where x ∈ [0, 3]. Then ϕ′ (x) = √ such that
2 x+1
ϕ(0) = 1 and ϕ(3) = 2. It follows that
√
e x+1
f (ϕ(x)) · ϕ′ (x) = √ .
2 x+1

By the Change of Variables, we obtain

∫ 3 √x+1 ∫ 3 ∫ ϕ(3)
e ′
√ dx = 2 f (ϕ(x)) · ϕ (x) dx = 2 f (t) dt
0 x+1 0 ϕ(0)
∫ 2 ∫ 2
=2 t
e dt = 2 (et )′ dt = 2(e2 − e).
1 1

Example 7.3.12 Evaluate ∫ 1

xf (x2 ) dx
−1

for any f is continuous on [0, 1].

Solution. Let ϕ(x) = x2 where x ∈ [−1, 1]. Then ϕ′ (x) = 2x such that ϕ(−1) = 1 and ϕ(1) = 1.
It follows that
f (ϕ(x)) · ϕ′ (x) = f (x2 ) · 2x.

By the Change of Variables, we obtain

∫ 1 ∫ ∫ ∫
1 1 ′ 1 ϕ(1) 1 1
2
xf (x ) dx = f (ϕ(x)) · ϕ (x) dx = f (t) dt = f (t) dt = 0
−1 2 −1 2 ϕ(−1) 2 1
7.3. FUNDAMENTAL THEOREM OF CALCULUS 179

Example 7.3.13 Let f : [−a, a] → R where a > 0. Suppose f (−x) = −f (x) for all x ∈ [−a, a].
Show that ∫ a
f (x) dx = 0.
−a

Solution. Let ϕ(x) = −x where x ∈ [−a, a]. Then ϕ′ (x) = −1 such that ϕ(−a) = a and
ϕ(a) = −a. It follows by the Change of Variables that
∫ a ∫ a
f (x) dx = −f (x) · (−1) dx
−a −a
∫ a
= f (−x) · ϕ′ (x) dx
∫−aa
= f (ϕ(x)) · ϕ′ (x) dx
−a
∫ ϕ(a)
= f (t) dt
ϕ(−a)
∫ −a
= f (t) dt
a
∫ a
=− f (t) dt.
−a
∫ a ∫ a
Then, 2 f (x) dx = 0. We conclude that f (x) dx = 0.
−a −a
180 CHAPTER 7. INTEGRABILITY ON R

Exercises 7.3
1. Compute each of the following integrals.
∫ 3
∫ e
1.1 |x + x − 2| dx
2 1.4 x ln x dx
−3 1
∫ 4 √ ∫ π
x−1 2
1.2 √ dx 1.5 ex sin x dx
1 x 0
∫ ∫ √
1 1
4x2 − 4x + 1
1.3 (3x + 1)99 dx 1.6 dx
0 0 x2 − x + 3

2. Use First Mean Value Theorem for Integrals to prove the followingversion of the Mean Value
Theorem for Derivatives. If f ∈ C 1 [a, b], then there is an x0 ∈ [a, b] such that

f (b) − f (a) = (b − a)f ′ (x0 ).

∫ x2
d
3. If f : [0, ∞) → R is continuous, find f (t) dt.
dx 0
∫
d t
4. If g : R → R is continuous, find g(x) dx.
dt cos t
∫ x2 √
5. Let g be differentiable and integrable on R. Define f (x) = g(t) · t dt.
∫ 1 1

Show that xg(x) + f (x) dx = 0.

0
∫ x2 ∫ 1 ∫ 1
6. If f (x) = 2 2
sec (t )dt. show that 2 sec (x ) dx − 4
2 2
xf (x) dx = tan 1.
0 0 0
∫ 3
7. Suppose that g is integrable and nonnegative on [1, 3] with g(x) dt = 1. Prove that
1
∫
1 9 √
g( x) dx < 2.
π 1
∫ 11
8. Suppose that h is integrable and nonnegative on [1, 11] with h(x) dt = 3. Prove that
1
∫ 2
h(1 + 3x + 3x2 − x3 ) dx ≤ 1.
0

9. If f is continuous on [a, b] and there exist numbers α ̸= β such that

∫ c ∫ b
α f (x) dx + β f (x) dx = 0
a c

holds for all c ∈ (a, b), prove that f (x) = 0 for all x ∈ [a, b].
Chapter 8

Infinite Series of Real Numbers

8.1 Introduction

Let {ak }k∈N be a sequence of numbers. We shall call an expression of the form
∑
∞
ak
k=1

an infinite series with terms ak .

∑
∞
Definition 8.1.1 Let S = ak be an infinite series whose terms ak belong to R.
k=1

1. The partial sums of S of order n are the numbers defined, for each n ∈ N, by
∑
n
sn := ak .
k=1

2. S is said to converge if and only if its sequence of partial sums {sn } to some s ∈ R as
n → ∞; i.e., for every ε > 0 there is an N ∈ N such that

n≥N implies |sn − s| < ε.

In this case we shall write

∑
∞
ak = s
k=1

∑
∞
and call s the sum, or value, of the series ak .
k=1

3. S is said to diverge if and only if its sequence of partial sums {sn } does not converge.
182 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

∞ [
∑ ]
1 1
Example 8.1.2 Prove that − = 1.
k=1
k k+1

Solution. Use telescoping, we have

n [
∑ ]
1 1 1
sn = − =1− .
k=1
k k+1 n+1
( ) ∞ [
∑ ]
1 1 1
Then, lim sn = lim 1 − = 1. We conclude that − = 1.
n→∞ n→∞ n+1 k=1
k k+1

∑
∞
Example 8.1.3 Prove that (−1)k diverges.
k=1

Solution. We see that



−1
∑
n
if n is odd
sn = (−1)k =

0
k=1 if n is even.

∑
∞
It is easy to see that sn does not converge as n → ∞. Hence, (−1)k diverges.
k=1

1
Theorem 8.1.4 (Harmonic Series) Prove that the sequence k
converges but the series
∑
∞
1
diverges.
k=1
k

Proof. By Example 2.1.5, it implies that 1

k
→ 0 as k → ∞. Let x ∈ [k, k + 1] for each k ∈ N. Then

1 1 1
≤ ≤ .
k+1 x k

By Comparison Theorem for integral, We obtain

∫ k+1 ∫ k+1
1 1 1
dx ≤ dx =
k x k k k

It follows that
∑
n n ∫
∑ k+1 ∫ n+1
1 1 1
sn = ≥ dx = dx = ln(n + 1)
k=1
k k=1 k x 1 x

∑
∞
1
We conclude that sn → ∞ as n → ∞, i.e., diverges.
k=1
k
8.1. INTRODUCTION 183

Theorem 8.1.5 (Divergence Test) Let {ak }k∈N be a sequence of real numbers.
∑
∞
If ak does not converge to zero, then the series ak diverges.
k=1

∑
∞
Proof. Assume that ak converges and equals to s. Then
k=1

∑
n
sn = ak and sn → s as n → ∞.
k=1

Since ak = sk+1 − sk ,
lim ak = lim (sk+1 − sk ) = s − s = 0.
k→∞ k→∞

Thus, ak converges to zero.

∑
∞
n
Example 8.1.6 Show that the series diverges.
k=1
n+1
Solution. We see that
n
lim = 1 ̸= 0.
n→∞ n + 1

∑
∞
n
By the Divergence Test, it imlplies that diverges.
k=1
n+1

Theorem 8.1.7 (Telescopic Seires ) If {ak } is a convergent real sequence, then

∑
∞
(ak − ak+1 ) = am − lim ak .
k→∞
k=m

Proof. By telescoping, we have

∑
n
sn = (ak − ak+1 ) = am − an+1 .
k=m

Thus,
∑
∞
(ak − ak+1 ) = lim (am − an+1 )
n→∞
k=m

= am − lim an+1
n→∞

= am − lim ak .
k→∞
184 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

∑
∞
1
Example 8.1.8 Evaluate the series .
k=1
(k + 1)(k + 2)

Solution. By the Telescopic Series, we obtain

∑
∞ ∑ ∞ ( )
1 1 1
= −
k=1
(k + 1)(k + 2) k=1 k+1 k+2
1 1 1
= − lim = .
1 + 1 k→∞ k + 1 2
∑
∞
1
Example 8.1.9 Determine whether √ √ converges or not.
k=1
k+1+ k

Solution. Use telescoping, we have

∑
n
1 ∑ [√ √ ] √
n
sn = √ √ = k + 1 − k = n + 1 − 1.
k=1
k+1+ k k=1

∑
∞
1
Then, sn → ∞ as n → ∞. We conclude that √ √ diverges.
k=1
k+1+ k
∑
∞
Theorem 8.1.10 (Geometric Seires) The series xk converges if and only if |x| < 1, in which
k=1
case
∑
∞
x
xk = .
k=1
1−x

∑
∞
Proof. If |x| ≥ 1, then {x } diverges. By The Divergence Test, it implies that
k
xk diverges.
k=1
Case |x| < 1. Then xk → 0 as k → ∞. Since xk − xk+1 = xk (1 − x), we have

xk xk+1
xk = − .
1−x 1−x

By the Telescopic Series,

∑
∞ ∑∞ [ ]
xk xk+1
k
x = −
k=1 k=1
1 − x 1−x
x xk
= − lim
1 − x k→∞ 1 − x
x
=
1−x
∑
∞ ∑
∞
x
Thus, x converges if and only if |x| < 1 and
k
xk = .
k=1 k=1
1−x
8.1. INTRODUCTION 185

Example 8.1.11 Determine whether the following series converges or diverges.

∑
∞ ∑
∞
√
−k
1. 2 2. ( 2 − 1)−k
k=1 k=1

∑
∞
Solution. For 1. We have x = 1
2
such that |x| < 1. It implies that 2−k converges and
k=1

∑
∞ ∞ ( )k
∑ 1
−k 1 2
2 = = = 1.
k=1 k=1
2 1− 1
2

∑∞ ∑∞ ( )k
√ −k 1 1 √
Since ( 2 − 1) = √ and √ = 2 + 1 > 1,
k=1 k=1
2−1 2−1
∑∞
√
we conclude that ( 2 − 1)−k diverges.
k=1

∑
∞ ∑
∞
Theorem 8.1.12 Let {ak } and {bk } be a real sequences. If ak and bk are convergent series,
k=1 k=1
then
∑
∞ ∑
∞ ∑
∞ ∑
∞ ∑
∞
(ak + bk ) = ak + bk and (αak ) = α ak
k=1 k=1 k=1 k=1 k=1

for any α ∈ R.

∑
n ∑
n
Proof. Let sn = ak and tn = bk . Assume that sn → s and tn → t as n → ∞. Then
k=1 k=1

∑
n ∑
n
sn + tn = (ak + bk ) and αsn = αak .
k=1 k=1

By the Limit Theorem, it implies that sn + tn → s + t and αsn → αs as n → ∞.

The proof of this Theorem is complete.

∑
∞ ∑
∞
Theorem 8.1.13 If ak converges and bk diverges, then
k=1 k=1

∑
∞
(ak + bk ) diverges.
k=1

Proof. Exercise.
186 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

∑
∞
1 + 2k+1
Example 8.1.14 Evaluate .
k=1
3k

Solution. Use the Geometric Series and Theorem 8.1.12, it implies that
[( ) ( )k ]
∑∞
1 + 2k+1 ∑∞
1 + 2k+1 ∑∞
1
k
2
k
=2+ k
=2+ +2
k=0
3 k=1
3 k=1
3 3
∑∞ ( )k ∑∞ ( )k
1 2
=2+ +2
k=1
3 k=1
3
1 2
1 13
=2+ 3
+2· 3
=2+ +4= .
1− 1
3
1− 2
3
2 2

∑∞
k
Example 8.1.15 Evaluate k
.
k=1
2

Solution. Consider the difference of

k 1 k−1 2k − k − 1 2k k + 1 k k+1
k
− k = k
= k
= k − k = k−1 − k .
2 2 2 2 2 2 2 2

By the Telescopic and Geometric Series, we have

∑∞ ∑∞ [ ( )]
k 1 k k+1
k
= k
+ k−1
− k
k=1
2 k=1
2 2 2
∑∞ ∑∞ [ ]
1 k k+1
= + − k
k=1
2k k=1 2k−1 2
1
k+1
= 2
+ 1 − lim =1+1−0=2
1− 1
2
k→∞ 2k

Example 8.1.16 Let π be a Pi constant. Show that

[ ( k )k ]
∑∞
1 π 2k π
k 2 1− +
k=1
π π π

converges and find its value.

Solution. We rewrite the term of this series

[ ( k )k ] 2
1 π 2k π 1 1 1 πk
1 − + = − + ·
π k2 π π π k2 π k2 −2k+1 π k2 π k
( ) ( )k
1 1 1
= k 2 − (k−1) 2 +
π π π
8.1. INTRODUCTION 187

Then, the first term is telescoping series and the second term is geometric series. Thus,
[ ( k )k ] ∑ ∞ ( ) ∑ ∞ ( )k
∑∞
1 π 2k π 1 1 1
k 2 1− + = k 2 − (k−1) 2 +
k=1
π π π k=1
π π k=1
π
∑ ∞ ( ) ∑ ∞ ( )k
1 1 1
=− (k−1) 2 − k 2 +
k=1
π π k=1
π
1
1
= −1 + lim + π
1 − π1
2
k→∞ π k

1 2−π
= −1 + 0 + = #
π−1 π−1
∑
∞
1
Example 8.1.17 Evaluate the series .
k=2
k2 −1

Solution. By the Telescopic Series, we obtain

∑
∞
1 ∑∞
1
=
k=2
k 2 − 1 k=2 (k − 1)(k + 1)
∞ ( )
1∑ 1 1
= −
2 k=2 k − 1 k + 1
∞ [( ) ( )]
1∑ 1 1 1 1
= − + −
2 k=2 k−1 k k k+1
[∞ ( ) ∑ ∞ ( )]
1 ∑ 1 1 1 1
= − + −
2 k=2 k − 1 k k k+1
[( ) ( k=2
)]
1 1 1 1
= 1 − lim + − lim
2 k→∞ k 2 k→∞ k + 1
[ ]
1 1 3
= 1−0+ −0 = .
2 2 4
∞ (
∑ )
1 2k
Example 8.1.18 Evaluate + .
k=2
n2 − 1 7 · 5k

Solution. Use Example 8.1.17, it implies that

∞ (
∑ ) ∑
∞ ∑
∞
1 2k 1 2k
+ = +
k=2
n2 − 1 7 · 5k k=2
n2 − 1 k=2 7 · 5k
∞ ( )k
3 1∑ 2
= +
4 7 k=2 5
4
3 1 3 4 331
= + · 25
= + =
4 7 1− 2
5
4 105 420
188 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

Exercises 8.1

1. Show that
∑
∞
xn
xk =
k=n
1−x

for |x| < 1 and n = 0, 1, 2, ....

2. Prove that each of the following series converges and find its value.

√ √
∑
∞
k+1− k ∑
∞
(−1)k+1 + 4 ∑
∞
2.1 √ 2.3 2.5 2k e−k
k=1
k(k + 1) k=1
5k k=0
∑∞ k+1
(−1) ∑∞
3k ∑∞
2k − 1
2.2 2.4 2.6
k=1
πk k=1
7k−1 k=1
2k

3. Represent each of the following series as a telescopic series and find its value.
∑
∞
1
3.1
k=1
(2k − 1)(2k + 1)
∑∞ ( )
k(k + 2)
3.2 ln
k=1
(k + 1)2
∑∞ √ ( ( π )jk )
π 1
3.3 k
1− , where jk = − for k ∈ N
k=1
4 4 k(k + 1)

4. Find all x ∈ R for which

∑
∞
3(xk − xk−1 )(xk + xk−1 )
k=1

converges. For each such x, find the value of this series.

5. Prove that each of the following series diverges.

∑
∞ ∑∞ ( )k ∑
∞
1 1 k+1
5.1 cos 2 5.2 1− 5.3
k=1
k k k=1
k2
k=1

∑
∞
6. Prove that if ak converges, then its partial sums sn are bounded.
k=1

7. Let {bk } be a real sequence and b ∈ R.

8.1. INTRODUCTION 189

7.1 Suppose that there is an N ∈ N such that |b − bk | ≤ M for all k ≥ N . Prove that

∑
n ∑
N
nb − bk ≤ |bk − b| + M (n − N )
k=1 k=1

for all n > N .

7.2 Prove that if bk → b as k → ∞, then

b1 + b2 + · · · + bn
→b as n → ∞.
n

7.3 Show that converse of 7.2 is false.

∑
∞
8. A series ak is said to be Cesàro summable to L ∈ R if and only if
k=0

n−1 (
∑ )
k
σn := 1− ak
k=0
n

converges to L as n → ∞.

∑
∞
s1 + s2 + · · · + sn
8.1 Let sn = ak . Prove that σn = for each n ∈ N.
k=0
n
∑
∞
8.2 Prove that if ak ∈ R and ak = L converges, then c is Cesàro summable to L.
k=0
∑
∞
8.3 Prove that (−1)k is Cesàro summable to 12 ; hence the converge of 8.2 is false.
k=0
∑
∞
8.4 TAUBER. Prove that if ak ≥ 0 for k ∈ N and ak is Cesàro summable to L, then
k=0
∑
∞
ak = L.
k=0

∑
∞
9. Suppose that {ak } is a decreasing sequence of real numbers. Prove that if ak converges,
k=1
then kak → 0 as k → ∞.
∑
∞
ak ∑∞
ak
10. Suppose that ak ≥ 0 for k large and converges. Prove that lim = 0.
k=0
k j→∞
k=1
j + k

∑
∞ ∑
∞ ∑
∞
11. If and ak converges and bk diverges, prove that (ak + bk ) diverges.
k=1 k=1 k=1
190 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

8.2 Series with nonnegative terms

INTEGRAL TEST.

Theorem 8.2.1 (Integral Test) Suppose that f : [1, ∞) → R is positive and decreasing on [1, ∞).
∑
∞
Then f (k) converges if and only if
k=1
∫ n
lim f (x) dx < ∞.
n→∞ 1

∑
n ∫ n
Proof. Let sn = f (k) and tn = f (x) dx for n ∈ N. Since f is positive and decreasing on
k=1 1
[1, ∞), f is locally integrable on [1, ∞). For each k ∈ N, we have

f (k + 1) ≤ f (x) ≤ f (k) for all x ∈ [k, k + 1].

Taking integrate on [k, k + 1], we obtain

∫ k+1 ∫ k+1 ∫ k+1
f (k + 1) = f (k + 1) dx ≤ f (x) dx ≤ f (k) dx = f (k).
k k k

Summing over k = 1, 2, ..., n − 1, it follows that

∑
n−1 n−1 ∫
∑ k+1 ∑
n−1
f (k + 1) ≤ f (k + 1) dx ≤ f (k)
k=1 k=1 k k=1
∫ n
sn − f (1) ≤ f (k + 1) dx ≤ sn − f (n)
1
sn − f (1) ≤ tn ≤ sn − f (n)

−f (1) ≤ tn − sn ≤ −f (n)

f (n) ≤ sn − tn ≤ f (1)

Thus, {sn } is bounded if and only if {tn } is. Since f is positir, it implies that both sn and tn are
incresing. It follows that from the Monotone Convergence Theorem that sn converges if and only
if tn converges.
8.2. SERIES WITH NONNEGATIVE TERMS 191

∑
∞
1
Example 8.2.2 Use the Integral Test to prove that diverges.
k=1
k

1
Solution. Let f (x) = . Then f is positive and decreasing on [1, ∞). We obtain
x
∫ n ∫ n
1
lim f (x) dx = lim dx
n→∞ 1 n→∞ 1 x
∫ n
= lim (ln x)′ dx
n→∞ 1

= lim (ln n − ln 1) = ∞.
n→∞

∑
∞
1
By the Integral Test, we conclude that diverges.
k=1
k

∑∞
1
Example 8.2.3 Show that converges.
k=1
k2

1
Solution. Let f (x) = . Then f is positive and decreasing on [1, ∞). We obtain
x2
∫ n ∫ n ∫ n
1
lim f (x) dx = lim dx = lim (−x−1 )′ dx
n→∞ 1 n→∞ 1 x2 n→∞ 1
( )
1
= lim − + 1 = 1 < ∞.
n→∞ n

∑∞
1
By the Integral Test, we conclude that 2
converges.
k=1
k

∑
∞
1
Example 8.2.4 Show that converges.
k=1
k2 + 1

1
Solution. Let f (x) = . Then f is positive and decreasing on [1, ∞). We obtain
x2+1
∫ n ∫ n
1
lim f (x) dx = lim dx
n→∞ 1 n→∞ 1 x2 + 1
∫ n
= lim (arctan x)′ dx
n→∞ 1
π π π
= lim (arctan n − arctan 1) = − = < ∞.
n→∞ 2 4 4
∑
∞
1
By the Integral Test, we conclude that converges.
k=1
k2 +1
192 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

p-SERIES TEST.

Theorem 8.2.5 (p-Series Test) The series

∑∞
1
k=1
kp

converges if and only if p > 1.

Proof. If p < 0 or p = 1, then the series diverges. Case p > 0 and p ̸= 1, set f (x) = x−p and
observe that

f ′ (x) = −px−p−1 < 0 for all x ∈ [1, ∞).

Thus, f is positive and decreasing on [1, ∞). Since

∫ n ∫ n ( 1−p )′
−p x n1−p − 1
lim x dx = lim dx = lim
n→∞ 1 n→∞ 1 1−p n→∞ 1 − p

has a finite limit if and only if 1 − p < 0. It follows from the Integral Test that p-series converges
if and only if p > 1.
∑
∞
2 −2
Example 8.2.6 Find p ∈ R such that kp converges.
k=1

∑
∞
1
Solution. Rewrite the sum which is a p-series. Then the series converges if and only if
k=1
k 2−p2
2 − p2 > 1. It follows that p − 1 < 0 is equivalent to p ∈ (−1, 1)
2

∑∞ ( )
k + 2k
Example 8.2.7 Determine whether converges or not.
k=1
k2k

Solution. Consider
k + 2k 1 1
k
= k+ .
k2 2 k
∑1
∞ ∑ 1
∞
Since diverges (the p-Series Test, p = 1) and k
converge (the geometric series, x = 12 ),
k=1
k k=1
2
we conclude that
∑∞ ( ) ∑ ∞ ( )
1 1 k + 2k
+ = diverges.
k=1
k 2k k=1
k2k
8.2. SERIES WITH NONNEGATIVE TERMS 193

COMPARISON TEST.

∑
∞
Theorem 8.2.8 Suppose that ak ≥ 0 for k ≥ N . Then ak converges if and only if its sequence
k=1
of partial sums {sn } is bounded, i.e., if and only if there exists a finite number M > 0 such that

∑
n
ak ≤ M for all n ∈ N.
k=1

∑
n ∑
∞
Proof. Let sn = ak for n ∈ N. If ak converges, then sn convergess as n → ∞. Since every
k=1 k=1
convergent sequence is bounded by the BCT, sn is bounded. The proof is complete.

Theorem 8.2.9 (Comparison Test ) Suppose that there is an M ∈ N such that

0 ≤ ak ≤ bk for all k ≥ M .

∑
∞ ∑
∞
1. If bk < ∞, then ak < ∞.
k=1 k=1

∑
∞ ∑
∞
2. If ak = ∞, then bk = ∞.
k=1 k=1

Proof. Assume that there is an M ∈ N such that 0 ≤ ak ≤ bk for all k ≥ M .

∑n ∑n
Let sn = ak and tn = bk . For each n ≥ M , we sum over k = M + 1, ..., n
k=1 k=1

∑
n ∑
n
0≤ ak ≤ bk
k=M +1 k=M +1

0 ≤ sn − sM ≤ tn − tM .

Since M is fixed, it follows that sn is bounded when tn is, tn is unbounded when sn is. Apply
Theorem 8.2.8, we obtain this Theorem.
194 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

Example 8.2.10 Determine whether the following series converges or diverges.

∑
∞
1 ∑
∞
1
1. 2.
k=1
k3 + 1 k=1
k 3 + 3k

Solution. Since k 3 + 1 > k 3 > 0 and 3k + k 3 > k 3 > 0 for all k ∈ N, we have
1 1 1 1
0< < 3 and 0< < 3.
k3 +1 k k3 +3 k k
∑∞
1
We see that 3
converges by the p-Series Test (p = 3 > 1). It implies by the Comparison Test
k=1
k
that
∑
∞
1 ∑
∞
1
3
and converge.
k=1
k +1 k=1
k3 + 3k

∑∞
1
Example 8.2.11 Determine whether converges or diverges.
k=2
ln k

Solution. Use the MVT to prove that (see 1.10 of Exercise 6.3)
√
ln x ≤ x for all x > 1.
√
It follows that 0 < ln k ≤ k for all k > 1. Then

1 1
0< √ < .
k ln k

∑∞
1
We see that √ diverges by the p-Series Test (p = 1
2
< 1). It implies by the Comparison Test
k=1
k
that
∑∞
1
diverges.
k=2
ln k
8.2. SERIES WITH NONNEGATIVE TERMS 195

LIMIT COMPARISON TEST.

Theorem 8.2.12 (Limit Comparison Test) Suppose that ak and bk are positive for lagre k and
an
L := lim
n→∞ bn

exists as an extended real number.

∑∞ ∑
∞
1. If 0 < L < ∞, then bk converges if and only if ak converges.
k=1 k=1

∑
∞ ∑
∞
2. If L = 0 and bk converges, then ak converges.
k=1 k=1

∑
∞ ∑
∞
3. If L = ∞ and bk diverges, then ak diverges.
k=1 k=1

ak
Proof. Assume that ak and bk are positive for lagre k and → L as k → ∞.
bk
L
1. Case 0 < L < ∞. Given ε = . There is an N ∈ N such that
2
ak L
k ≥ N implies −L < .
bk 2
L ak L
For each n ≥ N , we have − < − L < , i.e.,
2 bk 2
L 3L
0 < · bk < ak < · bk .
2 2
Hence, part 1 follows immediately from the Comparison Test and Theorem 8.1.12.
Similar arguments establish part 2 and 3.
∑
∞
1
Example 8.2.13 Use the Limit Comparison Test to prove that converge.
k=1
k2 +1
1 1
Solution. Let ak = and bk = 2 . Then
x2 +1 k
ak k2
lim = lim 2 = 1 < ∞.
k→∞ bk k→∞ k + 1

∑∞
1
We see that converges by the p-Series Test (p = 2 > 1). It implies by the Limit Comparison
k=1
k2
Test that
∑∞
1
2
converges.
k=1
k +1
196 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

∑
∞
k
Example 8.2.14 Determine whether converges or diverges.
k=1
2k 4 +k+3

k 1
Solution. Let ak = and bk = 3 . Then
2k 4 +k+3 k
ak k4 1
lim = lim 4
= < ∞.
k→∞ bk k→∞ 2k + k + 3 2
∑∞
1
We see that 3
converges by the p-Series Test (p = 3 > 1). It implies by the Limit Comparison
k=1
k
∑∞
k
Test that 4
converges.
k=1
2k + k + 3

∑
∞
1
Example 8.2.15 Determine whether √ converges or diverges.
k=1
k+1

1 1
Solution. Let ak = √ and bk = √ . Then
k+1 k
√
ak k
lim = lim √ = 1 < ∞.
k→∞ bk k→∞ k+1
∑∞
1
We see that √ diverges by the p-Series Test (p = 1
2
< 1). It implies by the Limit Comparison
k=1
k
∑∞
1
Test that √ diverges.
k=1
k+1

Theorem 8.2.16 Let ak → 0 as k → ∞. Prove that

∑
∞ ∑
∞
sin |ak | converges if and only if |ak | converges.
k=1 k=1

Proof. Assume that ak → 0 as k → ∞. We will see that

sin |ak | sin x

lim = lim+ = 1 < ∞.
k→∞ |ak | x→0 x

∑
∞ ∑
∞
By the Limit comparison Test, it implies that sin |ak | converges if and only if |ak | converges.
k=1 k=1
8.2. SERIES WITH NONNEGATIVE TERMS 197

Exercises 8.2

1. Prove that each of the following series converges.

∑
∞ ∑
∞ ∑∞ ( )
k−3 ln k 1
1.1 3
1.3 , p>1 1.5 10 + k −e
k=1
k +k+1 k=1
kp k=1
k
√
∑∞
k−1 ∑∞
1 ∑∞
3k 2 − k
1.2 1.4 √ 1.6
k=1
k2k k=1
k3k−1 k=1
k4 − k2 + 1

2. Prove that each of the following series diverges.

√
∑
∞ k
k ∑∞
k 2 + 2k + 3
2.1 2.3 √
k k 3 − 2k 2 + 2
k=1 k=1
∑∞
1 ∑
∞
1
2.2 p , p>0 2.4 , p≤1
k=1
ln (k + 1) k=1
k lnp k
√
∑
∞
3k ln k
3. Use the Comparison Test to determine whether converges or diverges.
k=1
k2 + k k

∑
∞
1
4. Find all p ≥ 0 such that the following series converges. p
k=1
k ln (k + 1)

∑
∞
ak
5. If ak ≥ 0 is a bounded sequence, prove that converges for all p > 1.
k=1
(k + 1)p

∑
∞ ∑
∞
|ak |
6. If |ak | converges, prove that converges for all p ≥ 0. What happen if p < 0 ?
k=1 k=1
kp

∑
∞ ∑
∞ ∑
∞
7. Prove that if ak and bk coverge, then ak bk also converges.
k=1 k=1 k=1

8. Suppose tha a, b ∈ R satisfy b

a
∈ R\Z. Find all q > 0 such that

∑
∞
1
converges.
k=1
(ak + b)q k

∑
∞ ∑
∞
9. Suppose that ak → 0. Prove that ak converges if and only if the series (a2k + a2k+1 )
k=1 k=1
converges.
198 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

8.3 Absolute convergence

∑
∞
Theorem 8.3.1 (Cauchy Criterion) Let {ak } be a real sequence. Then the infinite series ak
k=1
converges if and only if for every ε > 0, there is an N ∈ N such that

∑
m
m>n≥N imply ak < ε.
k=n

∑
∞
Proof. Let sn represent the sequence of partial sum of ak and set s0 = 0. By the Cauchy’s
k=1
Theorem (Theorem 2.4.5), sn converges if and only if for every ε > 0, there is an N ∈ N such that

m, n ≥ N imply |sm − sn−1 | < ε.

For all m > n ≥ 1, we obtain

∑
m
ak = |sm − sn−1 | < ε.
k=n

The proof is complete.

∑
∞
Corollary 8.3.2 Let {ak } be a real sequence. Then the infinite series ak converges if and only
k=1
if for every ε > 0, there is an N ∈ N such that

∑
∞
n≥N implies ak < ε.
k=n

Proof. Exercise.
8.3. ABSOLUTE CONVERGENCE 199

ABSOLUTE CONVERGENCE.
∑
∞
Definition 8.3.3 Let S = ak be an infinite series.
k=1

∑
∞
1. S is said to converge absolutely if and only if |ak | < ∞.
k=1

2. S is said to converge conditionally if and only if S converges but not absolutely.

∑
∞
Theorem 8.3.4 A series ak converges absolutely if and only if for every ε > 0 there is an
k=1
N ∈ N such that
∑
m
m>n≥N implies |ak | < ε.
k=n

Proof. The Cauchy Criterion gives us the Theorem 8.3.4.

∑
∞ ∑
∞
Theorem 8.3.5 If ak converges absolutely, then ak converges.
k=1 k=1

∑
∞ ∑
∞
Proof. Assume that ak converges absolutely. Then |ak | converges.
k=1 k=1
Let ε > 0. By Theorem 8.3.4, there is an N ∈ N such that
∑
m
m>n≥N implies |ak | < ε.
k=n

Apply the Triangle Inequality, we obtain

∑
m ∑
m
ak ≤ |ak | < ε.
k=n k=n

∑
∞
By the Cauchy Criterion, we conclude that ak converges.
k=1
200 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

∑
∞
(−1)k ∑
∞
(−1)k
Example 8.3.6 Prove that converges absolutely but is not.
k=1
k2 k=1
k

Solution. We consider
∑∞
(−1)k ∑∞
1 ∑∞
(−1)k ∑∞
1
2
= and = .
k=1
k k=1
k2 k=1
k k=1
k

Since the first and second series are a p-series such that p = 2 and p = 1, respectively, we obtain the
∑∞
(−1)k
first series converges but the second series is not. We conclude that 2
converges absolutely
k=1
k
∑ ∞
(−1)k
but is not.
k=1
k

LIMIT SUPREMUM.

Definition 8.3.7 The supremum s of the set of adherent points of a sequence {xk } is called the
limit supremum of {xk }, denoted by s := lim sup xk , i.e.,
k→∞

lim sup xk = lim sup{xk : k ≥ n}.

k→∞ n→∞

Example 8.3.8 Evaluate limit supremum of the following sequences.

1 (−1)k 3. zk = 1 + (−1)k
1. xk = 2. yk =
k k

Solution. By the Definition of limit supremum, we have

{ } { }
1 1 1 1 1
lim sup xk = lim sup : k ≥ n = lim sup , , , ... = lim = 0
k→∞ n→∞ k n→∞ n n+1 n+2 n→∞ n
{ }
(−1)k
lim sup yk = lim sup :k≥n
k→∞ n→∞ k


 1 , − 1 , 1 , ...
n n+1 n+2
if n is even
= lim sup
n→∞ 
− 1 , 1 , − 1 , ... if n is odd
n n+1 n+2
1
= lim =0
n→∞ n
{ }
lim sup zk = lim sup (−1)k + 1 : k ≥ n = lim sup {0, 2} = lim 2 = 2.
k→∞ n→∞ n→∞ n→∞
8.3. ABSOLUTE CONVERGENCE 201

Theorem 8.3.9 Let x ∈ R and {xk } be a real sequence.

1. If lim sup xk < x, then xk < x for large k.

k→∞

2. If lim sup xk > x, then xk > x for infinitely many k.

k→∞

Proof. Let x ∈ R and s := lim sup xk .

k→∞
1. Assume that s < x. Suppose to the contary that there exist natural numbers

k1 < k2 < k3 < · · · such that xkj ≥ x for j ∈ N.

If {xkj } is unbounded above, it implies that sup{xk : k ≥ n} is unbounded above so s = ∞, a

contradiction. If {xkj } is bounded above by C, then x ≤ xkj ≤ C for all j ∈ N. Thus, by the
Bolzano-Weierstrass Theorem and the fact that x ≤ xkj , {xkj } has a convergent subsequence. It
implies that s > x, another contradiction.
2. Assume that s > x. There is a c ∈ R such that x < c < s. By the Approximation Property in
the Theorem 2.2.5, there is a subsequence {xkj } that converges to c; i.e., xk > x for lagre j.

Theorem 8.3.10 Let x ∈ R and {xk } be a real sequence. If xk → x as k → ∞, then

lim sup xk = x.
k→∞

Proof. Assume that xk → x as k → ∞. By the Theorem 2.1.18, any subsequence {xkj } also
converges to x. It implies that lim sup xk = x.
k→∞
{ }
k
Example 8.3.11 Evaluate limit supremum of .
k+1
k
Solution. Since lim = 1, we obtain by Theorem 8.3.10 that
k→∞ k + 1

k k
lim sup = lim = 1.
k→∞ k + 1 k→∞ k + 1
202 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

ROOT TEST.

1
Theorem 8.3.12 (Root Test) Let ak ∈ R and r := lim sup |ak | k .
k→∞

∑
∞
1. If r < 1, then ak converges absolutely.
k=1

∑
∞
2. If r > 1, then ak diverges.
k=1

Proof. 1. Assume that r < 1. Then there is an x ∈ R such that r < x < 1.
∑
∞
We notice that the geometric series xk converges. By Theorem 8.3.9, we have
k=1
1
|ak | k < x for large k.
∑
∞
It follows that 0 < |ak | < x for large k. By the Comparison Test,
k
|ak | converges.
k=1
2. Assume that r > 1. By Theorem 8.3.9, we have
1
|ak | k > 1 for infinitely many k.

It follows that |ak | > 1 for infinitely many k. Then the limit of ak is not zero.
∑∞
By the Divergence Test, ak diverges.
k=1

∞ (
∑ )k
k
Example 8.3.13 Prove that converges absolutely.
k=1
1 + 2k

Solution. We notice that

( )k 1
k
k k k 1
lim sup = lim sup = lim = < 1.
k→∞ 1 + 2k k→∞ 1 + 2k k→∞ 1 + 2k 2

∞ (
∑ )k
k
By the Root Test, we conclude that converges absolutely.
k=1
1 + 2k
8.3. ABSOLUTE CONVERGENCE 203

∞ (
∑ )k
3 + (−1)k
Example 8.3.14 Prove that diverges.
k=1
2
Solution. We notice that
( )k 1
k
3 + (−1)k 3 + (−1)k
lim sup = lim sup
k→∞ 2 k→∞ 2

= lim sup{1, 2} = lim 2 = 2 > 1.

n→∞ n→∞

∞ (
∑ )k
3 + (−1)k
By the Root Test, we conclude that diverges.
k=1
2

RATIO TEST.

Theorem 8.3.15 (Ratio Test) Let ak ∈ R with ak ̸= 0 for large k and suppose that
ak+1
r := lim
k→∞ ak
exists as an extended real number.
∑
∞
1. If r < 1, then ak converges absolutely.
k=1

∑
∞
2. If r > 1, then ak diverges.
k=1

Proof. 1. Assume that r < 1. Then there is an x ∈ R such that r < x < 1.
∑
∞
We notice that the geometric series xk converges.
k=1
ak+1 ak+1
By Theorem 8.3.10, we have r = lim = lim sup . By Theorem 8.3.9, we obtain
k→∞ ak k→∞ ak
ak+1
<x for large k.
ak
ak+1 xk+1
It follows that < x = k for large k which is equivalent to
ak x
|ak+1 | |ak |
< for large k.
xk+1 xk
|ak |
Then is decreasing and bounded. So, there is an M > 0 such that |ak | ≤ M xk for all k ∈ N.
xk
∑∞ ∑∞
We see that M xk converges. By the Comparison Test, |ak | converges.
k=1 k=1
204 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

2. Assume that r > 1. By Theorem 8.3.9, we have

ak+1
>1 for infinitely many k.
ak
It follows that |ak+1 | > |ak | for infinitely many k. Thus, ak is incresasing which induces nonzero
∑ ∞
limit of ak . By the Divergence Test, ak diverges.
k=1

∑
∞
3k
Example 8.3.16 Prove that converges absolutely.
k=1
k!

Solution. We notice that

3k+1 k! 3
lim · k = lim = 0 < 1.
k→∞ (k + 1)! 3 k→∞ k + 1

∑
∞
3k
By the Ratio Test, we conclude that converges.
k=1
k!

∑∞
k!
Example 8.3.17 Prove that diverges.
k=1
kk

Solution. We notice that

(k + 1)k+1 k! (k + 1)k+1
lim · k = lim
k→∞ (k + 1)! k k→∞ (k + 1)k k

(k + 1)k
= lim
k→∞ kk
( )k
k+1
= lim
k→∞ k
( )k
1
= lim 1 + =e>1
k→∞ k

∑∞
k!
By the Ratio Test, we conclude that k
diverges.
k=1
k
8.3. ABSOLUTE CONVERGENCE 205

Exercises 8.3

1. Prove that each of the following series converges.

∑∞ ∑∞ ∑
∞
∑∞ ( )k2
1 1 2k k
1.1 1.2 1.3 1.4
k=1
k! k=1
kk k=1
k! k+1
k=1

2. Decide, using results convered so far in this chapter, which of the following series converge
and which diverge.

∑∞ ∑∞ ( )k ∞ ( )k 2
k2 k+1 ∑ k!
2.1 2.4 2.7
k=1
πk 2k + 3 (k + 2)!
k=1 k=1
∑
∞
∑∞ ( )k2 ∞ (
∑ )k
k! k 3 + (−1)k
2.2 2.5 2.8
k=1
2k k+1 3
k=1 k=1
∑
∞ ∑∞ ( ) ∑
∞
k! 1 (1 + (−1)k )k
2.3 2.6 π− k −1 2.9
k=1
2 + 3k
k
k=1
k k=1
ek

3. Define ak recursively by a1 = 1 and

( ( ))−1
k 1
ak = (−1) 1 + k sin ak−1 , k > 1.
k

∑
∞
Prove that ak converges absolutely.
k=1

√ ∑
∞
4. Suppose that ak ≥ 0 and ak → a as k → ∞. Prove that
k ak xk converges absolutely for
k=1
all |x| < 1
a
if a ̸= 0 and for all x ∈ R if a = 0.

5. For eachof the following, find all values of p ∈ R for which the given series converges abso-
lutely.

∑
∞
1 ∑
∞
kp ∑
∞
2kp k!
5.1 5.3 5.5
k=2
k lnp k k=1
pk k=1
kk
∑∞
1 ∑∞
1 ∑∞
√
5.2 5.4 √ 5.6 ( k 2p + 1 − k p )
k=2
lnp k k=2
k(k p − 1) k=1
206 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

6. Suppose that akj ≥ 0 for k, j ∈ N. Set

∑
∞
Ak = akj
j=1

∑
∞
for each k ∈ N, and suppose that Ak converges.
k=1
(∞ ) (∞ )
∑
∞ ∑ ∑
∞ ∑
6.1 Prove that akj ≤ akj
j=1 k=1 k=1 j=1
(∞ ) ( )
∑
∞ ∑ ∑
∞ ∑
∞
6.2 Show that akj = akj
j=1 k=1 k=1 j=1

∑
∞ ∑
∞
7. Suppose that ak converges absolutely. Prove that |ak |p converges for all p ≥ 1.
k=1 k=1

∑
∞ ∑
∞
8. Suppose that ak converges conditionally. Prove that k p ak diverges for all p ≥ 1.
k=1 k=1
) (
a2
9. Let an > 0 for n ∈ N. Set b1 = 0, b2 = ln , and
a1
( ) ( )
ak ak−1
bk = ln − ln , k = 3, 4, ...
ak−1 ak−2
an
9.1 Prove that r = lim if exists and is positive, then
n→∞ an−1

∑∞ ( ) ∑∞
1 k−1
lim ln(an ) = limn 1− bk = bk = ln r.
n→∞ n→∞
k=1
n k=1

an+1 1
9.2 Prove that if an ∈ R\{0} and → r as n → ∞, for some r > 0, then |an | n → r as
an
n → ∞.
8.4. ALTERNATING SERIES 207

8.4 Alternating series

Theorem 8.4.1 (Abel’s Formula) Let {ak }k∈N and {bk }k∈N be real sequences, and for each pair
of integers n ≥ m ≥ 1 set
∑
n
An,m := ak .
k=m

Then
∑
n ∑
n−1
ak bk = An,m bn − Ak,m (bk+1 − bk )
k=m k=m

for all integers n > m ≥ 1.

Proof. Since Ak,m − Ak−1,m = ak for k > m and Am,m = am , we obtain

∑
n ∑
n
ak bk = am bm + ak bk
k=m k=m+1
∑n
= am bm + (Ak,m − Ak−1,m )bk
k=m+1
∑n ∑
n
= am bm + Ak,m bk − Ak−1,m bk
k=m+1 k=m+1
∑
n ∑
n−1
= am bm + Ak,m bk − Ak,m bk
k=m+1 k=m
∑
n−1 ∑
n−1
= am bm + Ak,m bk + An,m bn − Ak,m bk − Am,m bm+1
k=m+1 k=m+1
∑
n−1
= Am,m bm + An,m bn − Am,m bm+1 − Ak,m (bk+1 − bk )
k=m+1
∑
n−1
= An,m bn − Am,m (bm+1 − bm ) − Ak,m (bk+1 − bk )
k=m+1
∑
n−1
= An,m bn − Ak,m (bk+1 − bk )
k=m

The proof is complete.

208 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

Theorem 8.4.2 (Dirichilet’s Test) Let {ak } and {bk } be sequences in R. If the sequence of
∑
n
partial sums sn = ak is bounded and bk ↓ 0 as k → ∞, then
k=1

∑
n
ak bk converges.
k=1

∑
n
Proof. Let sn = ak be bounded. Assume that bk is decreasing and converges to zero.
k=1
There is an M > 0 such that
∑
n
|sn | = ak ≤ M for all n ∈ N.
k=1

By the triangle inequality, for n > m > 1.

∑
n
|An,m | = ak = |sn − sm−1 | ≤ |sn | + |sm−1 | ≤ M + M = 2M.
k=m

Let ε > 0. Then there is an N ∈ N such that

ε
k≥N implies |bk | < .
2M
Since bk is decreasing and converges to zero, bk − bk+1 > 0 and bk > 0 for all k ∈ N.
By Abel’s Formula and telescoping, for n > m ≥ N , we obtain

∑
n ∑
n−1
ak bk = An,m bn − Ak,m (bk+1 − bk )
k=m k=m
∑
n−1
≤ |An,m ||bn | + Ak,m (bk+1 − bk )
k=m
∑
n−1
≤ 2M |bn | + |Ak,m ||bk+1 − bk |
k=m
∑
n−1
≤ 2M bn + 2M (bk − bk+1 )
k=m

= 2M bn + 2M (bm − bn )
ε
= 2M bm < 2M · = ε.
2M
∑
n
Thus, ak bk converges.
k=1
8.4. ALTERNATING SERIES 209

Corollary 8.4.3 (Alternating Series Test (AST)) If ak ↓ 0 as k → ∞, then

∑
∞
(−1)k ak converges.
k=1

∑
∞
Moreover, if ak converges, then
k=1

∑
∞
(−1)k ak converges conditionally.
k=1

∑
∞ ∑
∞
k
Proof. Since the partial sums of (−1) are bounded, (−1)k ak converges by Dirichilet’s Test.
k=1 k=1

∑
∞
(−1)k
Example 8.4.4 Prove that converges conditionally.
k=1
k

1 ∑ ∞
Solution. If ak = , we see that ak is decreasing and converges to 0. By AST, we have (−1)k ak
k k=1
∑∞ ∑∞
1
converges. It is clear that |(−1)k ak | = diverges by p-Series Test (p = 1).
k=1 k=1
k
∑∞
(−1)k
We conclude that converges conditionally.
k=1
k

∑
∞
(−1)k
Example 8.4.5 Prove that converges conditionally.
k=2
ln k

1
Solution. Let ak = . Since k + 1 > k > 0, ln(k + 1) > ln k. It implies that
ln k
1 1
< for all k > 1.
ln(k + 1) ln k
∑
∞
(−1)k
Then ak is decreasing and converges to 0. By AST, we obtain converges.
k=2
ln k
∑∞
1 ∑∞
(−1)k
By Example 8.2.11, . We conclude that converges conditionally.
k=2
ln k k=2
ln k
210 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

∑
∞
sin(kx)
Example 8.4.6 Prove that S(x) = converges for each x ∈ R.
k=1
k

Solution. Let x ∈ R. If x = 2ℓπ where ℓ ∈ Z, then

∑
∞
sin(2kℓπ)
= 0 < ∞.
k=1
k
{ }
1 1
For case x ̸= 2ℓπ for all ℓ ∈ Z. It’s easy to see that is decreasing and lim = 0.
k k→∞ k
Define
∑n
Sn = sin(kx)
k=1

Use trigonometry properties and teleascoping, we have

( ( x )) ( ( x )) ∑
n
2 sin Sn = 2 sin sin(kx)
2 2 k=1
∑
n (x)
= 2 sin(kx) sin
k=1
2
n [ (
∑ x) ( x )]
= cos kx − − cos kx +
k=1
2 2
∑n [ ( ) ( )]
1 1
= cos x k − − cos x k +
2 2
k=1
( ) ( )
1 1
= cos x − cos x n + .
2 2
(x)
Since sin ̸= 0 for all x ̸= 2ℓπ. We obatin
2
( ( x )) ( ) ( )
1 1
2 sin Sn = cos x − cos x n +
2 2 2
( ( x )) ( ) ( )
1 1
2 sin |Sn | = cos x + cos x n + +≤1+1=2
2 2 2
1 (x)
|Sn | ≤ ( ( x )) = csc .
sin 2 2
So, Sn is bounded for each x ̸= 2ℓπ. By Dirichilet’s Test, it implies that
∑
∞
sin(kx)
converges for all x ̸= 2ℓπ.
k=1
k

Therefore, we conclude that

∑
∞
sin(kx)
S(x) = converges for all x ∈ R.
k=1
k
8.4. ALTERNATING SERIES 211

Exercises 8.4

1. Prove that each of the following series converges.

∑
∞ (π ) ∑
∞
sin(kx)
1.1 (−1) k
− arctan k 1.5 , x ∈ R, p > 0
k=1
2 k=1
kp
∑∞
(−1)k k 2 ∑∞
(−1)k 2 · 4 · · · (2k)
1.2 1.6
k=1
2k k=1
k 2 1 · 3 · · · (2k − 1)
∑
∞
(−1)k ∑
∞
(−1)k
1.3 1.7
k=1
k! k=1
ln(ek + 1)
∑∞
(−1) k ∑∞
arctan k
1.4 , p>0 1.8
k=1
kp k=1
4k 3 − 1

2. For each of the following, find all values x ∈ R for which the given series converges.

∑
∞
xk ∑∞
(x + 2)k
2.1 2.4 √
k=1
k k=1
k k+1
∑
∞
x3k ∑
∞
2k (x + 1)k
2.2 2.5
k=1
2k k=1
k!
∑∞ ∞ (
∑ )k
(−1)k xk k(x + 3)
2.3 √ 2.6
k=1
k2 + 1 k=1
cos k

3. Using any test coveredin this chapter, find out which of the following series converge abso-
lutely, which converge conditionally, and which diverge.
√
∑
∞
(−1)k k 3 ∑∞
(−1)k k + 1
3.1 3.5 √
k=1
(k + 1)! k=1
kk k
∑∞
(−1)(−3) · · · (1 − 2k) ∑
∞
(−1)k sin k
3.2 3.6
k=1
1 · 4 · · · (3k − 2) k=1
k!
∑∞
(k + 1)k ∑∞
(−1)k
3.3 , p>e 3.7 √
k=1
pk k! k=1
k2 + 1
√
∑∞
(−1)k k ∑∞
(−1)k ln(k + 2)
3.4 3.8
k=1
k+1 k=1
k
212 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

∑
∞
4. ABEL’S TEST. Suppose that ak converges and bk ↓ b as k → ∞. Prove that
k=1

∑
∞
ak bk converges.
k=1

5. Use Dirichilet’s Test to prove that

∑
∞
cos(kx)
S(x) =
k=1
k2

converges for all x ∈ R.

∑
∞
6. Prove that ak cos(kx) converges for every x ∈ (0, 2π) and every ak ↓ 0.
k=1
What happens when x = 0 ?
∑
∞ ∑
∞
7. Suppose that ak converges. Prove that if bk ↑ ∞ and ak bk converges, then
k=1 k=1

∑
∞
bm ak → 0 as m → ∞.
k=m
8.4. ALTERNATING SERIES 213

Index

Abel’s formula, 205 Chord, 119

Absolute value, 9 Closed, 74

Additive identity, 1 Closure, 80

Additive inverse, 1 Closure properties, 1

Additive properties, 5, 48 Continuity of linear function, 106

Additive rule, 125 Continuous, 101, 106

Unifromly continuous, 115
Alternating series test (AST), 207
Unifrom continuity of linear function, 115
Associative properties, 1
Commutative properties, 1
Approach, 83
Comparison test, 191
Approximatation property infimum (API), 28 Comparison theorem, 51, 169
Approximatation property supremum (APS), 23 Comparison theorem for function, 90
Archimedean properties (AP), 24 Composition, 105

Base, 2 Converge, 36, 83

Bernoulli’s inequality, 133 Converge absolutely, 197

Converge conditionally, 197
Binomial formula, 19
Decreasing, 59, 140
Bolzano-Weierstrass theorem, 62
Stricly decreasing, 140
Bounded, 26, 32, 110
Density
Bounded above, 21, 32
of Irrationals, 27
Bounded below, 26, 32
of Rationals, 26
Bounded convergent theorem (BCT), 43
Derivative, 119
Cancellation law, 4 Differentiable, 119, 122
Cauchy, 65 Distributive properties, 1

Cauchy criterior, 196 Dirichlet function, 157

Cauchy’s theorem, 66 Dirichlet test, 206

Chain rule, 127 Diverges, 52, 179

Change of variable, 176 Divergent test, 181

214 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

Does not exist (DNE), 40 Limit comparison test, 191

Dyadic partition, 148 Limit supremum (limsup), 198

Exponent, 2 Linear property, 166

Extreme value theorem (EVT), 110 Lower bound, 26

Fundamental theorem of calculus, 172 Mathematical induction, 17

Gauss’formula, 18 Mean value theorem (MVT), 131

Geometric series, 182 Generalized mean value theorem, 134

Harmonic series, 180 Monotone, 59, 140

Stricly monotone, 140
Increasing, 59, 140
Stricly increasing, 140 Monotone convergent theorem (MCT), 60

Infinite series, 179 Multiplicative identity, 1

Infimum, 26 Multiplicative inverse, 1

Integers, 2 Multiplicative law, 9

Integral test, 188 Multiplicative property, 5, 49

Integration by part, 174 Natural numbers, 2

Interval, 11 Neighborhood, 72
Closed interval, 11
Nested, 61
Open interval, 11
Nested interval property, 61
Interior point, 77
Nonnegative, 5
Intermedaite value theorem (IVT), 112
Norm, 147
Inverse function, 33
One-to-one, 31
Inverse function theorem, 143
Onto, 31
Isolated point, 77
Open, 69
Irrationals, 2
p-series test, 190
L’Hôspital’s rule, 135

Least element, 17 Partial sum, 179

Limit, 83 Partition, 147

Limit of constant function, 85 Positive, 5

Limit of linear function, 86
Positive definite, 9
Limit point, 77
Product rule, 126
Left-hand limit, 93
Right-hand limit, 92 Quotient property, 50
8.4. ALTERNATING SERIES 215

Quotient rule, 126 Sequence, 35

Infinite sequence, 35
Rationals, 2
Subsequence, 41
Ratio test, 201
Sequential Charaterization of limit (SCL), 87
Reciprocal property, 49
Sign-preserving theorem, 111

Refinement, finer, 147 Squeeze theorem, 45

Riemann integrable, integrable, 155 Squeeze theorem for function, 89

Lower integral, 157
Symmetric law, 10
Upper integral, 157
Sum telescopes, 152
Riemann sum, 163 Telescopic series, 181
Lower Riemann sum, 149
Supremum, 21
Upper Riemann sum, 149
Tangent line, 119
Roll’s theorem, 130
Transtive property, 5
Root test, 200
Triangle inequality, 12
Scalar multiplicative property, 48 Apply triangle inequality, 13

Scalar multiplicative rule, 125 Trichotomy property, 5

Secant line, 119 Upper bound, 21

216 CHAPTER 8. INFINITE SERIES OF REAL NUMBERS

Reference

Gerald B. Folland. (1999). Real Analysis Modern Technique and Their Applications.
John Wiley & Sons, Inc., New York.

Halsey L. Royden and Prtrick M. Fitzpatrick. (2010). Real Analysis (Fourth Edition).
Pearson Education, Inc. New Jersey.

Pual Glendinning. (2012). Maths in minutes. Quercus Editions Ltd, London, England.

William R. Wade. (2004). An Introduction Analysis (Third Edition). Pearson

Education. Inc., New Jersey.
8.4. ALTERNATING SERIES 217

Vita

Assistant Professor Thanatyod Jampawai, Ph.D.

• B.Sc. (Mathematics, 2nd class honours), Chulalongkorn University, 2006

• M.Sc. (Mathematics), Chulalongkorn University, 2009

• Ph.D. (Mathematics), Chulalongkorn University, 2014

• Assistant Professor of Mathematics, Division of Mathematics, Faculty of Education,

Suan Sunandha Rajabhat University.

Contact me

Email: [email protected]

Website: www.eledu.ssru.ac.th/thanatyod_ja

Office: 1144