A.

M
.A

1
.E
l-S
ay
ed
 A.
M
.A

i
.E
l-S
ay
ed
 Advanced Calculus

Prof. Dr. Ahmed M. A. El-Sayed

Faculty of Science
Alexandria University

e.mail: [email protected]
ed
[email protected]
ay
http://ahmed-el-sayed.8m.com
l-S
.E

.
.A

.
http://math-frac.org
M
A.

∞
X sin nx
F (x) =
n=1
np

Z ∞ Z b
f (x) dx = lim f (x) dx
a b→∞ a

ii
Contents

1 Sequence of Real Numbers 1

1.1 Sequence of real numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Bounded sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.2 Convergent sequences . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.3 Divergent sequences . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Cauchy sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Monotonic sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Subsequences . . . . . . . .
1.5 Exercise I . . . . . . . . . .
. . .
. . .
.
.
.
.
.
.
.
.
ed.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
7
ay
l-S

2 Infinite Series of Real Numbers 9

.E

2.1 Series of real numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.2 Convergence and divergence . . . . . . . . . . . . . . . . . . . . . . . . . 9
.A

2.3 Series of nonnegative terms . . . . . . . . . . . . . . . . . . . . . . . . . . 11

M

2.4 Tests of convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

A.

2.5 Absolute convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

2.6 Alternating series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.7 Some general tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.8 Multiplication of series . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.9 Exercises II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3 Infinite products 23
3.1 Convergence and divergence . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 Absolute convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Exercises III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4 Sequences and Series of Functions 26

4.1 Sequences of functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.1.1 Bounded sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.1.2 Convergence and divergence . . . . . . . . . . . . . . . . . . . . . . 27
4.1.3 Test for uniform convergence . . . . . . . . . . . . . . . . . . . . . . 30
4.2 Series of function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.3 Convergence of the series of functions . . . . . . . . . . . . . . . . . . . . 31

iii
 4.3.1 Abel’s test and Dirichlet’s test . . . . . . . . . . . . . . . . . . . . 32
4.4 Uniform convergence and continuity . . . . . . . . . . . . . . . . . . . . . . 32
4.5 Uniform convergence and integrability . . . . . . . . . . . . . . . . . . . . 33
4.6 Uniform convergence and differentiation . . . . . . . . . . . . . . . . . . . 34
4.7 Exercises IV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

5 Power series 37
5.1 Radius of convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.2 Uniform convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.3 Multiplication of power series . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.4 Exercises V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

6 Improper integrals 41
6.1 First kind of improper integrals . . . . . . . . . . . . . . . . . . . . . . . . 41
6.2 Tests of convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.3 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.4 Absolute convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6.5 Second kind of improper integrals . . . . . . . . . . . . . . . . . . . . . . . 44
6.6 Third kind of improper integrals . . . . . ed. . . . . . . . . . . . . . . . . . . 46
ay
6.7 Uniform convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
l-S

6.8 Properties of uniform convergence . . . . . . . . . . . . . . . . . . . . . . 46

.E

6.9 Exercises VI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
.A

7 Introduction to
M

Fourier analysis 49
A.

7.1 Periodic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

7.2 Fourier series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
7.3 Half range series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
7.4 Parseval’s identity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
7.5 Bessel’s inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7.6 Differentiation and integration of Fourier series . . . . . . . . . . . . . . . . 54
7.7 Exercise VII . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

iv
Chapter 1

Sequence of Real Numbers

1.1 Sequence of real numbers

The sequence of real numbers is a set of real numbers in a definite order of arrangement
and formed according to a definite rule.
ed
Definition 1 The infinite sequence of real numbers is a function
ay
f :N →R
l-S

designated by f (n) or by {an } or

.E

{a1 , a2 , a3 , · · · , an , · · · }
.A
M

where N = {1, , 2 , 3, · · · } is the set of natural numbers and R is the set of all real
A.

numbers.
Thus a sequence is a set of numbers
a1 , a2 , a3 , · · · , an , · · ·
in a definite order of arrangement and formed according to a definite rule. Each number
in the sequence is called a term; an is called the nth. term.

Examples 1
(1) { an } = { − n } = { − 1, − 2, − 3, · · · }

(2) { an } = { 1/n } = { 1, 1/2, 1/3, · · · }

(3) { an } = { 1 + (−1)n /n }

(4){ an } = { 2−n }

(5) { an } = { (1 + 1/n)n }

1
1.1.1 Bounded sequences
A sequence {an } is said to be bounded above if there exists a constant number k1 such that

an ≤ k1 , ∀ n ∈ N

and bounded below if there exists k2 such that

an ≥ k2 , ∀n ∈ N.

The sequence {an } is bounded if it is bounded below and above, or if there exists a real
number k > 0 such that
| an | ≤ k, ∀n ∈ N.

Examples 2
(1) The sequence {an } = {(−1)n } is bounded, this is since an ≥ −1 and an ≤ 1
⇒ |an | ≤ 1.
ed
ay
(2) The sequence {an } = {1/n} is bounded, this is since an > 0 and an ≤ 1 ⇒ |an | ≤ 1.
l-S
.E

1.1.2 Convergent sequences

.A

Limit of a sequence
M

A number a is called the limit of the sequence { an } if,

A.

∀  > 0, there exists N () > 0 such that |an − a| <  ∀ n > N ().

In such case we write

lim an = a or an → a.
n→∞

And we say that the sequence { an } converges to the number a.

1.1.3 Divergent sequences

The sequence {an } is said to be diverge if it is not converge.

The sequence {an } is said to be divergent to +∞ if,

∀ M > 0, ∃ N (M ) such that an > M whenever n > N (M ).

In this case we write limn→∞ an = +∞.

If limn→∞ (−an ) = +∞, we write limn→∞ an = −∞ and we say that {an } diverges to −∞.
Of course, there are divergent sequences which do not diverge to +∞ or −∞.

2
Examples 3
(1) The sequence { 1/n} converges to zero. Since

|an − 0| = |1/n| = 1/n <  ⇒ n > 1/,

then,
∀  > 0, ∃ N () = [1/] such that |an − 0| < , ∀ n > N ().
It must be notice that if  = 0.01, then N () = 100, and if  = 0.001, then N () = 1000
and so.

(2) The sequence { (−1)n /n} converges to zero. Since

|an − 0| = |(−1)n /n| = 1/n <  ⇒ n > 1/,

then
∀ > 0, ∃ N () = [1/] such that |an − 0| <  ∀n > N ()

ed
(3) The sequence {n} diverges to ∞ and the sequence {(−1) n2 } diverges to −∞.
ay
l-S

Lemma 1.1 Given real numbers a and b such that a ≤ b +  for every  > 0.
.E

Then a ≤ b.
Proof. If b < a, then the inequality a ≤ b +  is violated for  = (a − b)/2 because
.A
M

a−b a+b a+a

b+=b+ = < = a.
A.

2 2 2

Theorem 1.1
(i) Every convergent sequence is bounded.
(ii) Every convergent sequence has a unique limit.
Proof. Let {an } be convergent to a, then

∀ > 0, ∃ N () such that |an − a| < , ∀n > N ().

This implies that

a −  < an < a + , ∀n > N ().
Putting g = min {a − , a1 , a2 , · · · , an } and G = max {a + , a1 , a2 , · · · , an },
then we have
g ≤ an ≤ G, ∀n
which proves that the sequence {an } is bounded.
Now let {an } converges to the two limits a, b, then

∀ > 0, ∃ N1 () such that |an − a| < /2, ∀n > N1 ().

3
 and ∀ > 0, ∃ N2 () such that |an − b| < /2, ∀n > N2 ().
So we have
|a − b| = |(an − b) − (an − a)| ≤ |an − a| + |an − b| <
< /2 + /2 = , ∀ n > max (N1 , N2 ) ⇒
∀  > 0, ∃ N () = max (N1 , N2 ) such that

|a − b| < , ∀ n > N ⇒ a = b.

And the limit is unique.

Remark
The two conditions (i) and (ii) of Theorem 1.1 are necessarily conditions. This means that
if the sequence {an } does not satisfy these two conditions, then it diverge.

Examples 4 ed
ay
l-S

(1) The sequence {an }, an = 1 ∀n converges to the value a = 1.

.E

(2) The sequence {an }, an = 1/n, n ∈ N, converges to the value 0.

.A
M

(3) The sequence {an }, an = 1 + (−1)n , n ∈ N, has the range { 0, 2 } and diverges.
A.

Squeeze Theorem. Suppose that the two sequences {an } and {bn } are conver-
gent to the limit L. If there is an integer no such that

an ≤ c n ≤ b n , ∀ n ≥ no
then the sequence {cn } converges to L, too.
Corollary. If, limn→∞ |an | = 0, then limn→∞ an = 0.
Proof. The results follows from the inequality

−| an | ≤ an ≤ | an |

Neighborhood
The neighborhood of the point a is the set

Nr (a) = { x : |x − a| < r }.

This shows that Nr (a) is the open interval (a − r, a + r) and N (a) is the open interval
(a − , a + )

4
Limit point of a sequence
A number b is said to be a limit point of the sequence {an } if every neighborhood of b,
Nr (b) contains an infinite numbers of member of the sequence {an },
i.e., if ∀ r > 0 we have an ∈ Nr (b) for infinite numbers of values of n ⇒ |an − b| < r for
infinite numbers of values of n.

Bolzano-Weierstrass theorem
Every bounded sequence has at least one limit point.

Theorem 1.2 The set of limit points of a bounded sequence has the greatest ( sup )
and the least ( inf ) members. The greatest is called upper limit and the least is called the
lower limit and denoted by

lim¯ an = lim sup an and limn→∞ an = lim inf an =

n→∞ n→∞ n→∞

Theorem 1.3 A bounded sequence of real numbers converges if and only if

lim sup an = lim inf an = a ⇒ an → a

ed
ay
1.2 Cauchy sequence
l-S
.E

A sequence {an } is called a Cauchy sequence if

.A

∀  > 0, ∃ N () > 0, such that |am − an | <  ∀ m > n > N ().
M
A.

Theorem 1.4 A necessary and sufficient condition for the convergence of a sequence {an }
is that the sequence is a Cauchy sequence.
Proof. (1) To prove the necessity, let an → a. Then

∀ > 0, ∃ N () such that |an − a| < /2, ∀n > N ().

and ∀ > 0, ∃ N () such that |am − a| < /2, ∀m > N ().
⇒ ∀ > 0, ∃ N () such that |am − an | ≤ |am − a| + |an − a| <
/2 + /2 =  ∀m > n > N () ⇒
every convergent sequence is a Cauchy sequence.
(2) Let {an } be a Cauchy sequence, then

∀  > 0, ∃ N () > 0, such that |an+1 − an | <  ∀ n > N () ⇒

an −  < an+1 < an + , n > N ().

Let
g = min{a1 , a2 , · · · , an+1 , an − }

5
and
G = max{a1 , a2 , · · · , an+1 , an + },
then
g ≤ an+1 ≤ G
i.e., the sequence is bounded, then from Bolzano-Weierstrass theorem the sequence has at
least one limit point.
Let p1 , p2 be two limit points of the sequence, then

∃  > 0, such that |an − p1 | < /3, and |am − p2 | < /3 for infinitly many n, m.

Now
|p1 − p2 | = |(an − p1 ) − (am − p2 ) − (am − an )| ≤
≤ |an − p1 | + |am − p2 | + |am − an | < 
for infinitely many m, n ⇒ p1 = p2 and the Cauchy sequence of real numbers is convergent.

Theorem 1.5 If an → a and bn → b, then

(1) an ± bn → a ± b ed
ay
(2) an .bn → a.b
l-S

(3) if b 6= 0, then an /bn → a/b.

Theorem 1.6 If the sequence {an } is bounded and bn → 0, then the sequence {an .bn }
.E

converges to 0.
.A
M

1.3 Monotonic sequences

A.

The sequence {an } is said to be monotonic increasing ( or nondecreasing) if an+1 > an (or
an+1 ≥ an ) ∀n ∈ N
and is said to be monotonic decreasing (or non increasing) if an+1 < an (or an+1 ≤ an )
∀n ∈ N .

Theorem 1.7 Every bounded monotonic sequence converges.

Corollary The monotonic sequence converges if and only if it is bounded.

1.4 Subsequences
Let {an } be a given sequence. For every increasing sequence {n1 , n2 , n3 , · · · nk , · · · } of
natural numbers the sequence {ank } is called a subsequence of {an }
For example {a1 , a4 , a7 , a10 , · · · } is a subsequence of {an } but {a10 , a4 , a7 , a1 , a13 , · · · }
is not.
In other words the linear order of a subsequence of {an } should be the same as that of the

6
original sequence.
As a consequence of Bolzano-Weierstrass theorem the student can prove the following the-
orem.

Theorem 1.8 Every bounded sequence contains a convergent subsequence.

Theorem 1.9 A sequence {an } converges to a if and only if every subsequence of {an }
converges to a.
Proof. Let an → a, then

∀  > 0 ∃ N () > 0 such that n > N () ⇒ |an − a| < .

If {ank } is a subsequence of {an }, then nk ≥ n which would implies that

∀  > 0 ∃ N () > 0 such that nk > N () ⇒ |ank − a| <  ⇒ ank → a.

i.e., every subsequence of {an } converges to a.

Now let every subsequence of {an } converges to a. Since {an } is a subsequence of it

self, then {an } converges to a.
ed
ay
l-S

Equivalent sequences
.E

The sequences {an } and {bn } are said to be equivalent if limn→∞ (an − bn ) exists and is
.A

zero. It is not necessary here that the two sequences are convergent.
M

Theorem 1.10 If the two sequences {an } and {bn } are equivalent, then either both
A.

of them converge to the same limit or neither of them converge.

Proof. If neither of the two sequences converge, then the theorem is satisfied. If however
one of them converges an → a say, then

bn = (bn − an ) + an → 0 + a ⇒ bn → a.

Examples 5
3 3 +n+1
(1) The two sequences { nn+1
2 } and {
n
n2
} are equivalent but neither of them converge.

n
(2) The two sequences { n+1 } and { n+1
n
} are equivalent and both of them has the limit
1.

1.5 Exercise I
(1) Prove Theorems 1.5, 1.6, 1.7 and 1.8.

7
(2) Discuss the boundedness, monotonicity and convergence of the following sequence.
(i) {(−1)n } (ii) {1 + (−1)n } (iii) {(−1)n+1 /n}

(3) Discuss the existence of limit points and the convergence of the following sequences.
2
(i) {1/n2 } (ii) {(−1)n n} (iii) {1 + (−1)n /n} (iv) { n2n+1 }

(4) If {an } and {bn } are two Cauchy sequences. Prove that the sequences
an
{an ± bn }, {an .bn }, { }, bn 6= 0
bn
are Cauchy sequences.

(5) If an → a and an ≥ 0 ∀n, prove that a ≥ 0

(6) If an → a, bn → b and an ≤ bn , prove that a ≤ b.

(7) If an → a, bn → b and an ≤ cn ≤ bn , prove a ≤ c ≤ b

ed
ay
(8) Prove that
l-S

1 1 1 1
bn = + + + ··· + → 0
.E

(n + 1) 2 (n + 2) 2 (n + 3) 2 (2n)2
.A

(9) Discuss for convergence the following sequences

M
A.

−5 n 7 tan−1 ln n
(1) {6( ) } (2) {8 − ( )n } (3) {tan−1 n} (4) { } (5) {(−1)n }
6 8 n n
n2 4n2 + 1 cos n en
(6) { } (7) { 2 } (8) { } (9) { 4 } (10) {e−n ln n}
ln (n + 1) 2n − 1 n n
4n3 + 5n + 1
(11) {(1 + 1/n)n } (12) {(−1)n n3 n−3 } (13) {2−n sin n} (14) { }
2n3 − n2 + 5
n2 n2
(15) { − (16) {n sin1/n} (17) {cos nπ} (18) {4 + sin nπ/2}
2n − 1 2n + 1
√ √ √
(19) {n1/n } (20) { n + 1 − n} (21) { n2 + n − n2 }.

8
Chapter 2

Infinite Series of Real Numbers

2.1 Series of real numbers

Formally an infinite series of real or complex numbers an is written as
∞
X
ed
an = a1 + a2 + a3 + · · · + an + · · · (2.1)
ay
n=1
l-S

Consider the sequence { Sn } defined by

.E

n
.A

X
Sn = a1 + a2 + · · · + an = ak .
M

k=1
A.

P
The sequence { Sn } is called the sequence of partial sums of the series an and the
number Sn is called the nth. partial sum of the series (1). The number an is called the nth.
term of the series (1).

2.2 Convergence and divergence

The convergence of the series (1) depends on the convergence of the sequence of partial
sums { Sn }.
If the sequence of partialPsums {Sn } converges to s ( limn→∞ Sn = s), then the series (1)
converges and we write an = s.
If the sequence { Sn } does not converge, then the series (1) is said to be diverge.
If the sequence { Sn } is bounded and the series (1) is divergent, then the series (1) is said
to be finitely oscillating.

Example 2.1
1
P
Prove that the series an , a n = n(n+1)
is convergent and find its sum.

9
 1 1
Proof. Since an = n
− n+1
, then
n
X 1 1 1 1 1 1 1 1
Sn = − = (1 − ) + ( − ) + · · · ( − )=1− → 1
k=1
k k + 1 2 2 3 n n + 1 n + 1

1
P P
i.e., Sn converges to 1, then the series converges and an = n(n+1)
= lim Sn = 1.

Theorem 2.1 (Cauchy criteria) The series (1) converges if and only if
m
X
given  > 0, ∃ N () > 0, such that m > n > N () ⇒ |Sm − Sn | ≤ | ak | < .
k=n+1

Proof. The series (1) converges if and only if the sequence { Sn } converges. By the
Cauchy theorem of convergence of the sequence, the sequence { Sn } converges if and only
if ∀  > 0, ∃ N () such that

|Sm − Sn | = |an+1 + an+2 + · · · + am | < , for m > n > N ().

Theorem 2.2 If the series (1) converges, then limn→

ed an = 0.
ay
∞
Proof. Taking m = n + 1 in theorem 2.1 we get
l-S
.E

∀  > 0, ∃ N () > 0, such that | an+1 | < , ∀ n > N () ⇒ an → 0.

.A
M

Remark The condition of Theorem 2.2 is necessary but not sufficient.

A.

Example 2.2 P 1
Show that the series n
does not converge
Proof. If the series converges, then
1 1
∀  > 0, say  = , ∃ N such that |S2n − Sn | < ∀ 2n > N.
4 4
But
1 1 1 1 1 1
|S2n − Sn | = | + + ··· + | = + + ··· + >
n+1 n+2 2n n+1 n+2 2n
1 1 1 1
+ + ··· + = >
2n 2n 2n 2
which prove that the series diverges. ( Note that an → 0).

Theorem 2.3 If
∞
X ∞
X ∞
X
an = s, then an = s + ao and c an = c s
n=1 n=0 n=1

10
where c is any number independent of n. Also if
∞
X ∞
X ∞
X
an diverges, then c an and an are divergent.
n=1 n=1 n=0

P P P
Theorem 2.4 If an = s1 and bn = s2 , then (an ± bn ) = s1 ± s2
Proof. The result follows from the identity
n
X n
X n
X
(an ± bn ) = an ± bn by making n → ∞.
k=1 k=1 k=1

The same proof shows that

(1) If any two of the three series are convergent, then the third is also convergent.

(2) If one of the series is divergent and the other convergent, then the third is neces-
sary divergent.

ed
(3) If two of the series are divergent, then no conclusion can be drawn about the behavior
of the third.
ay
l-S
.E

2.3 Series of nonnegative terms

.A

In this section we shall suppose that an is real and nonnegative for all n ∈ N .
M

This implies that

A.

Sn+1 − Sn = an+1 ≥ 0
and so the sequence { Sn } is monotonic increasing.
P
Theorem 2.5 A series an , an ≥ 0, is convergent if and only if the sequence of
partial sums { Sn } is bounded.
proof. an ≥ 0 ⇒ Sn+1 − Sn = an+1 ≥ 0 ⇒ the sequence P { Sn } is monotonic
increasing and hence converges ( and consequently the series an converges ) if and only
if it is bounded.

Example 2.3P 1
The series n2
is convergent.
proof. Since an = n12 is positive and { Sn } is bounded,
n n n n
X 1 X 1 X 1 X 1 1 1
Sn = 2
=1+ 2
<1+ =1+ − < 1 + (1 − ) < 2,
k=1
k k=2
k k=2
k(k − 1) k=2
k−1 k n

then the series converges.

Theorem 2.6 (Geometrical series)

11
 P n
Let r ≥ 0. The series r is convergent if r < 1 and divergent if r ≥ 1.
Proof. Let r < 1. Then
1 − rn 1
Sn = 1 + r + r2 + · · · + rn = < ⇒ { Sn } is bounded
1−r 1−r
and the series is convergent.
If r ≥ 1, then Sn = 1 + r + r2 + · · · + rn ≥ 1 + 1 + 1 + · · · + 1 ≥ n → ∞ as n → ∞.

2.4 Tests of convergence

Theorem
P 2.7 (Comparison test) P
Let cn be convergent series of positive terms and let dn be divergent one.
X
(1) If 0 ≤ an ≤ cn , ∀ n ≥ 1, then an is convergent
X
(2) If an ≥ dn , ∀ n ≥ 1, then the series an is divergent
Proof. (1) SinceP the series
P
ed
cn is convergent series of positive terms, then P
the sequence
ay
n n
of partial sums { Pck } is bounded and hence (an ≤ cn ) the sequence {
k=1 k=1 ak } is
l-S

bounded ⇒ the series an is convergent.

.E

Pn Pn P
(2) Since Sn = k=1 ak ≥ k=1 dn → ∞ as n → ∞ the series an diverges.
.A
M

Corollary
A.

X
(1) If ∃ No and k > 0 such that 0 ≤ an ≤ kcn , ∀ n ≥ No , then an is convergent
X
(2) If ∃ N1 and k > 0 such that an ≥ kdn , ∀ n ≥ N1 , then an is divergent.

Examples 2.4
Discuss the convergence of the series
1 1 1 1 1 1 1 1
(1) 1 + 2
+ 2 + 2 + · · · + n + · · · and (2) 1 + √ + √ + √ + · · · + √ + · · · .
2 3 4 n 2 3 4 n
Solution (1) Since nn ≥ 2n for n ≥ 2, then 0 ≤ an = n1n ≤ (1/2)n but (1/2)n is
P
convergent (why?) and by the comparison test the series (1) is convergent.
√ √ P
(2) Since n ≤ n for n ≥ 1, then 1/ n ≥ 1/n. But the series 1/n diverges, then

12
by the comparison test the series (2) is also diverges.
p p
P
Generally p ≤ 1 ⇒ n ≤ n ⇒ 1/n ≥ 1/n. But the series 1/n diverges, then the
1/np , p ≤ 1 diverges.
P
series

Theorem 2.8 If
an
an > 0, bn > 0 and → c 6= 0 as n → ∞,
bn
then the two series
∞
X X
an and bn are both convergent or both divergent.
n=1

Example P 2.5
The series sin n1 is divergent. This is since

an sin n1
= 1 → 1 as n → ∞.
bn n

Theorem 2.9 ( Ratio test I) ed

ay
P
Let an be a series of positive terms.
l-S

an+1 X
(1) If ≤ α < 1, ∀ n, α independent of n, then an converges.
.E

an
.A

an+1 X
(2) If > 1, ∀ n, then the series an diverges.
M

an
A.

proof. (1)
X
an ≤ an−1 α ≤ an−2 α2 ≤ · · · ≤ ao αn , α < 1 ⇒ αn converges
P
and by the comparison test the series an converges.
(2) X
an+1 > an ⇒ a1 < a2 < · · · < an < · · · ⇒ an 6→ 0 ⇒ an dinerges

Theorem 2.10 (Ratio test II )

X an+1
Let an be a series of positive terms such that lim = α, then
n→∞ an
P
(1) the series an converges if α < 1
P
(2) the series an diverges if α > 1

13
(3) the test fails when α = 1.
Proof. (1)
an+1 an+1
→ α ⇒ ∀  > 0, ∃ N () > 0 such that α −  < < α + , n > N ().
an an
Choose  > 0Psuch that α +  < 1, then an+1
an
< α +  < 1. Hence by the ratio test I we
deduce that an converges.
an+1
(2) Choose  > 0 whateverPsmall such that α −  > 1, then an
> 1. Hence by the
ratio test I we deduce that an diverges.

(3) When α = 1 we find that an+1

P1 P 1
an
→ 1. For the two series n
and n2
we have
an+1
an
→ 1 but the first diverges and the second converges. Then the test fails when α = 1.

Examples 2.6
Discuss the convergence of the series

1+
1
+
1
+ ··· +
1 ed
+ · · · and
2 22 23
+ + + ··· +
2n
+ ···
ay
1.2 1.2.3 1.2.3 · · · .n 1 2 3 n
l-S

Solution :
1 1 an+1 n! 1
.E

(1) an = and an+1 = ⇒ = = → 0 < 1, as n → ∞.

n! (n + 1)! an (n + 1)! n+1
.A

Hence the series converges.

M
A.

2n 2n+1 an+1 2n
(2) an = and an+1 = ⇒ = → 2 as n → ∞
n n+1 an n+1
and the series diverges.
Root test X √
Let an , an > 0 be such that lim n an = α, then
(1) The series converges if α < 1

(2) The series diverges if α > 1

(3) If α = 1 the test fails.

Proof.
√ √
n
an → α ⇒ ∀  > 0, ∃ N () > 0, such that α −  < n an < α + .
(1) Choose  > 0, whatever small such that (α + ) < 1 ⇒ an < (α + )n and by the
comparison test the series converges.

(2) Choose  > 0, whatever small such that (α − ) > 1 ⇒ an > (α − )n and by
the comparison test the series divergent.

14
 Exercises 2.1
Determine the whether following series converges or diverges.
X √n X √n 5
n n + 2)
X
(1) (2) (3) (99)
n2 + 1 3n + 4 n2 102n
X 32n 1 1 1 X 1 3/2
(4) n
n−1
(5) + + + · · · (6) (1 + √ )−n
5 2 4 8 n
X n+1 2 X 1 X 2
(7) ( )n (8) n
.(9)
n n n + en
3

X n+1 X 2 n X n!
(10) (11) n! ( ) (12)
n3 + 1 n nn
X nn X 10 + 2n X (n!)2
(13) (14) (15)
10n+1 n (2n)2
X (2n)! X n X 1
n
(16) n
(17) ( ) (18) √3
2 2n + 1 n ln n
ed
ay
l-S

1 1.4 1.4.7 1.4.7 · · · (3n − 2)

(19) + + + ··· +
.E

2 2.4 2.4.6 2.4.6 · · · (2n)

1.3.5 · · · (2n − 1)
.A

1.3 1.3.5
(20) 1 + + + ··· + .
2! 3! n!
M
A.

15
Cauchy’s integral test
Let anP = f (n), n ≥ 1 is nonnegative monotonic decreasing
R n integrable function. Then the
series an and the integral ( the sequence) In = 1
f (x) dx converges or diverges
together.
Proof. Since an is monotonic decreasing, then n < x < n + 1 ⇒ an > f (x) > an+1 and
Z n+1 Z n+1 Z n+1 Z n+1
an dx > f (x) dx > an+1 dx ⇒ an > f (x) dx > an+1 .
n n n n

hence
Z 2 Z 3 Z n
a1 > f (x) dx > a2 , a2 > f (x) dx > a3 , ··· , an−1 > f (x) dx > an .
1 2 n−1

Adding we get
n
X Z n
Sn − an > In > Sn − a1 , where Sn = ak and In = f (x) dx.
k=1 1

Let the integral In be convergent, i.e., limn→∞ In is finite ⇒ the sequence {In } is convergent
which implies that it is bounded, then ed
Pthe sequence {Sn } is bounded. Since it is monotonic,
ay
then it is convergent
P and the series an converges.
l-S

Now if the series an is convergent, then the sequence {Sn } is bounded and consequently
R n+1
R n sequence {In } is bounded. Since the sequence
the {In } is monotonic ( 1 f (x) dx >
.E

Rn
f (x) dx), then it is convergent and limn→∞ 1 f (x) dx is finite.
.A

1
M

Example 2.7
A.

P 1
Consider the series np
, p > 0. Let f (x) = x1p , then this function satisfies the assumptions
of the integral test ( Why ? ). Now
Z n Z n
dx 1 1
In = f (x) dx = p
= (x1−p )n1 = (n1−p − 1), p 6= 1.
1 1 x 1−p 1−p
Now p < 1 ⇒ 1 − p > 0 and In → ∞ as n → ∞ which implies that the series is divergent.
−1
p > 1 ⇒ 1 − p < 0 and In → 1−p as n → ∞ which implies that the series is convergent.
R n dx
For p = 1 we find that 1 x = ln n → ∞ as n → ∞ and the series diverges.

Exercise 2.2
P∞ 1
Test for convergence the series n=2 n(logn)P .

2.5 Absolute convergence

P P
The series an is said to be converges absolutely if the series |an | converges.

16
Theorem 2.11 Every absolutely convergent series is convergent.
Proof. The result follows by the comparison test and the relation an ≤ |an |.
P
Conditional convergence The series an is said to be convergent conditionally if
it is convergent but not absolutely convergent.

Example P 2.8
sin nx
The series n2
converges absolutely. this is since

sin nx 1 1 X 1
|an | = | | ≤ | = and the series converges.
n2 n2 n2 n2

Exercise 2.3
Discuss the absolute convergence
X xn ∞ ∞
(−1)n X sin nx X (1 + cos2 nx)
(1) (2) (3) (−1)n (4) (−1)n
n! n3 np np
n=1
ed n=1
ay
2.6 Alternating series
l-S
.E

A series whose terms are alternatively positive and negative is called an alternating series.
.A

Leibnitz test
M

If the alternating series (−1)n an , an > 0 is such that

P
A.

(1) an+1 ≤ an , ∀ n,

(2) lim an = 0, as n → ∞.
Then the series converges.
Proof The sum of the series to 2M terms is

S2M = (a1 − a2 ) + (a3 − a4 ) + · · · + (a2M −1 − a2M )

and satisfies

S2M = a1 − (a2 − a3 ) − (a4 − a5 ) − · · · − (a2M −2 − a2m−1 ) − a2M .

Since an ≥ an+1 , then

S2M ≥ 0, S2 ≤ S4 ≤ S6 · · · ≤ S2M ≤ a1 .

Therefore {s2M } is bounded and monotonic increasing sequence and thus has a limit S.
Since
S2M +1 = S2M + a2M +1 , lim S2M = S and lim a2M +1 = 0
M →∞ n→∞

17
it follows that
lim S2M +1 = S + 0 = S.
M →∞

Thus the partial sums of the series approach the limit S and the series converges.

Example 2.9 P (−1)n

Discuss for convergence the series np
, p > 0.
p
Here an = 1/n , p > 0 and the series is alternating. Since
(1) (n + 1)p ≥ np , then an+1 ≤ an

(2) n ≥ 1 and p > 0 ⇒ an → 0 as n → ∞.P

1/np converges when p > 1 and
P
Then the series is convergent. Now since |an | =
diverges when p ≤ 1, then the series is conditionally convergent when p ≤ 1 and the series
converges absolutely for p > 1.

2.7 Some general tests

1. Raabe’s test ed
ay
P
Let an be a series of positive terms. If
l-S

an
lim n ( − 1) = α,
.E

n→∞ an+1
.A

then the series (i) converges when α > 1, (ii) diverges when α < 1 and (iii) the test fails
M

when α = 1.
A.

Example 2.10
Test for convergence the series

α 1 + α (1 + α)(2 + α) (1 + α)(2 + α)(3 + α)

+ + + + ···
β 1 + β (1 + β)(2 + β) (1 + β)(2 + β)(3 + β)

Solution Here
(1 + α)(2 + α) · · · (n − 1 + α) an+1
an = , n = 2, 3, · · · and lim =1 ⇒
(1 + β)(2 + β) · · · (n − 1 + β) an

the ratio test fails. But

an n+β
lim n ( − 1) = lim n ( − 1) = β − α.
n→∞ an+1 n→∞ n+α
Then by Raabe’s test the series converges if β − α > 1, diverges if β − α < 1.
1+α
Now when β − α = 1, then β = 1 + α ⇒ an = n+α . Since α < n α, then n + α <
n+α 1+α
P
n(1 + α) ⇒ 1+α < n ⇒ n+α > 1/n. So by the comparison with 1/n the series

18
diverges when β − α = 1.

2. Logarithmic
P test
Let an be a series of positive terms such that
an
lim (n log − 1) = α,
n→∞ an+1
then the series converges if α > 1 and diverges if α < 1.

Example 2.11
Test for convergence
x 22 x2 33 x3 n n xn
1+ + + + ··· + + ··· .
1! 2! 3! n!
Ignoring the first term we can write
nn xn an+1 n+1 n
an = ⇒ lim = lim ( ) x = x e,
n! n→∞ an n→∞ n
ed
ay
then by the ratio test the series converges if x e < 1 and converges if x e > 1. Now for
x e = 1, x = 1/e ⇒
l-S
.E

an n n an n n+1
=( ) e ⇒ n log = n [1 + n log ] = n [1 − log ]=
an+1 n+1 an+1 n+1 n
.A
M

n
= n [1−n log(1+1/n)] = n [1−n(1/n−1/2n2 +1/3n3 −· · · )] ⇒ lim (n log −1) = −1/2 < 1
A.

n→∞ n+1
and the series diverges.

3. Dirichlet’s test and Abel’s test

Theorem 2.12 If {an } and {bn } are two sequences of real numbers, define An = a1 + a2 +
a3 + · · · . Then we have the identity
n
X n
X
ak bk = An bn+1 − Ak (bk+1 − bk ).
k=1 k=1
P∞ P∞
Therefore k=1 a k b k converges if both the series k=1 Ak (bk+1 − bk ) and the sequence
{ An bn+1 } converges.
Proof. Writing Ao = 0, we have
n
X n
X n
X n
X
ak bk = (Ak − Ak−1 ) bk = Ak bk − Ak bk+1 + An bn+1 .
k=1 k=1 k=1 k=1

The second assertion follows at once from this identity.

19
 P
Theorem 2.13 (Dirichlet’s test) Let an be a series of complex numbers whose
partial sums form a boundedPsequence. Let { bn } be a decreasing sequence which con-
verges to 0. Then the series an bn converges

Proof. Let An = a1 + a2 + · · · + an and assume that |An | ≤ M for all n. then

limn→∞ An bn+1 = 0. P P
Therefore, to stablish convergence of an bn we need only show that An (bn − bn+1 ) is
convergent.
Since bn is decreasing and An is bounded, we have
|Ak (bk+1 − bk )| ≤ M (bk − bk+1 ).
P
But the series (bP k − bk+1 ) is convergent (Why ?). Hence the comparison test implies
the convergence of An (bn+1 − bn ).
P P
Theorem 2.14 (Abel’s test) The series an bn converges if an converges and if
{ bn } is a monotonic convergent sequence.

ed
P
Proof. Convergence of an and {bn } establishes the existence of the limit limn→∞ An bn+1 ,
ay
where An = a1 + a2 + · · · + an .
l-S

Also, an is a bounded sequence. The remainder of the proof is similar to that of Theorem
2.13.
.E
.A

Examples 2.12
M
A.

P (n3 +1)1/3 −n
(1) For the series log n
we have an = (n3 + 1)1/3 − n, bn = 1/log n. Since the series
P
an converges (Why ?) and P the sequence {bn } is monotonic decreasing (Why ?), then
from Abel’s test the series an bn converges.

(2) For the series

1 1 1 (−1)n 1
1− 2
+ 2
− 2
+ · · · , we have an = 2
and bn = .
3.2 5.3 7.4 n 2n − 1
P
Since the series converges and the sequence {bn } is monotonic decreasing, then from
an P
Abel’s test the series an bn converges.

2.8 Multiplication of series

X X n
X X
Given an and bn , we put cn = ak bn−k , n = 1, 2, · · · and call cn
k=0

20
the Cauchy product of the two given series.

Example 2.13
The series ∞
X (−1)n 1 1 1
√ = 1 − √ + √ − √ + ···
n=0
n+1 2 3 4
is converges ( Why ?). We inform the product of this series with it self and obtain
∞ n
X 1 1 1 1 1 n
X 1
cn = 1 − ( √ + √ ) + ( √ + √ √ + √ ) + · · · + (−1) p +· · ·
n=0
2 2 3 2 2 3 k=0
(n − k + 1)(k + 1)

which diverges ( Why ?). This proves that the product of two convergent series may
actually divergent.

Theorem 2.15 Suppose

∞
X ∞
X
(i)
ed
an converges absolutly (ii) an = A,
ay
n=0 n=0
l-S

∞
X n
X
(iii) bn = B, (iv) cn = ak bn−k , n = 1, 2, · · ·
.E

n=0 k=1
.A

P P
Then the series cn converges with sum cn = AB.
M
A.

2.9 Exercises II
(I) Determine whether the series is absolutely converges, conditionally converges and di-
verges.
X 1 X (−1)n X (−1)n+1 X (−1)n
(1) (−1)n √ (2) (3) (4)
2n + 1 n2/3 ln (n + 1) n3
X (−1)n+1 n X (−1)n n X (−1)n ln n X (−1)n
(5) (6) (7) (8)
n2 + 4 ln n n np
X 5(−1)n X X (−10)n X n!
(9) (10) (−1)n e−n (11) (12) .
n3 + 1 n! (−5)n
(II) Test for convergence the following series
∞
1.2 3.4 5.6 X 1
(1) 2 2 + 2 2 + 2 2 + · · · (2)
3 .4 5 .6 7 .8 n=2
n log n (log log n)P

21
 ∞
X 1 1 1 1 1
(3) (4) √ + √ + √ + · · · + √ + ···
n=1
n log (1 + 1/n) 10 20 30 10n
1 1 1 1 1 1 1
(5) √ +√ + ··· + √ + · · · (6) + + + ··· + + ···
3
7 3
8 3
n+6 2 5 10 1 + n2
∞ X rn
X 1 X
3 1/3
X
−(1+1/n)
(7) , p > 0, (8) [(n +1) −n] (9) n (10) , r>0
n=2
(log n)p nn
1 2 1.2 2 1.2.3 2 1.2.3.4 2 X rn
(11) ( ) + ( ) + ( ) +( ) + · · · (12) , r > 0.
3 3.5 3.5.7 3.5.7.9 n!
(III) Test the convergence and divergence for the different values of x.

x x2 x3 x2 x3 x4 x x2 x3
(1) + + + · · · (2) √ + √ + √ + · · · (3) + + + ···
1 2 3 2 1 3 2 4 3 1 2! 3!
X 3.6.9. · · · 3n X 2.4.6. · · · 2(n + 1)
(4) xn , x > 0. (5) xn−1 , x > 0.
7.10.13. · · · (3n + 4) 3.5.7 · · · (2n + 3)
P
(IV) Given that the series an converges absolutely. Show that each of the following series
also converges absolutely ed
ay
l-S

X X an X a2n
(i) a2n (ii) (iii) .
.E

1 + an 1 + a2n
.A

an , an > 0 is convergent. Prove that the series (an an+1 )1/2

P P
(V) Given that the series
M

is convergent.
A.

P P
(VI) Given that the series an is divergent. Prove
P that n an is also divergent.
(VII) Assuming the convergence of the series an show that the following series are
convergent
X an X an X n+1 X 1
(1) (2) (3) ( ) an (4) (1 + )n an .
n log n n n

22
Chapter 3

Infinite products

Definition 3.1 Given a sequence {an }, let

p1 = a1 , p2 = a1 .a2 · · · , pn = a1 .a2 . · · · an = Πnk=1 ak . (3.1)
The symbol Π∞ n=1 an = a1 .a2 . · · · .an , · · · is called infinite product, pn is called the nth.
partial product and an the nth. factor of the product.
ed
ay
l-S
.E

3.1 Convergence and divergence

.A

(1) If an 6= 0, ∀n we say that the product pn converges if there exists p 6= 0 such that
M

pn → p.
A.

In this case p is called the value of the product and we write p = Π∞

n=1 an .
If pn → 0 we say that the product diverges to zero.

(2) If there exists an N such that n > N implies an 6= 0, we say that Π∞

n=1 an converges
∞
provided that Πn=N +1 an converges.

(3) The product diverges if it does not converge or if infinitely many factors an are zero.

Remarks
(1) The value of p can be zero if and only if a finite number of factors are zero.

(2) The convergence of the product is not affected by inserting or removing a finite number
of factors, zero or not.

Examples
(1) The product Π∞
n=1 (1 + 1/n) diverges.
n+1 2 3 4 n n+1
an = , pn = . . · · · . = n + 1 → ∞ as n → ∞.
n 1 2 3 n−1 n

23
(2) The product Π∞
n=2 (1 − 1/n) diverges.

n−1 1 2 3 n−1 1
an = , pn = . . · · · = → 0 as n → ∞.
n 2 3 4 n n
Theorem 3.1 (Cauchy) The infinite product Πan converges if and only if

∀  > 0, ∃ N such that n > N ⇒ |an+1 .an+2 . · · · .an+k − 1| < , k = 1, 2, 3 · · ·

Theorem 3.2 Let an > 0. Then the product Π(an + 1) converges if and only if the series
P
an converges

Proof. For x > 0, we have

x2 x3 xn
1+x≤1+x+ + + ··· + + · · · = ex .
2! 3! n!
Consider the two sequences

Sn = a1 + a2 + · · · + an and pn = (1 + a1 )(1 + a2 ) · · · (1 + an ).
ed
ay
These two sequences are increasing sequences. Also
l-S

and pn < eSn .

.E

pn > Sn
.A

Now (1)
M

pn > Sn ⇒, if pn → p, (pn > 0 ⇒ p 6= 0)

A.

then pn is bounded and so P Sn is bounded. Now Sn is bounded and increasing, then Sn

converges and so the series an converges.

(2) pn < eSn ⇒ if Sn converges, then it is bounded which implies that pn is bounded.
Now pn is bounded and increasing, then it converges to p 6= 0 say and the product converges.

3.2 Absolute convergence

The product Π(1+an ) is said to be converge absolutely if the product Π(1+|an |) converges.

Theorem 3.3 Every absolutely convergent product Π(1 + an ) is convergent.

3.3 Exercises III

(1) Prove theorem 3.3

24
 xn
(2) Discuss for the values of x and α the convergence of the product Π(1 + nα
), α > 0.

(3) Discuss for the values of x and α the absolute convergence of the product
n xn
Π(1 + (−1)
nα
), α > 0.

(4) If each sum Sn = nk=1 ak of the series

P P P
an is not zero and an → S 6= 0, show that
∞ an
the infinite product a1 Πn=2 (1 + Sn−1 ) converges to the value S.

(5) Discuss the convergence of the infinite products

3
2 1 ∞ n −1
(i) Π∞
n=2 (1 − ) (ii) Π∞
n=2 (1 − 2) (iii) Πn=2 3 .
n(n − 1) n n +1

ed
ay
l-S
.E
.A
M
A.

25
Chapter 4

Sequences and Series of Functions

4.1 Sequences of functions

Let A and B be two subsets of real (complex) numbers and fn : A → B, ∀ n = 1, 2, · · ·
a function with domain A and co-domain B, then the sequence { fn (x) } is a sequence of
functions.
ed
ay
l-S

Examples 4.1
.E

(1) Let fn : R → R, fn (x) = nx , n ∈ N and x ∈ R, then the sequence { x

} is a sequence
.A

n
of function.
M
A.

(2) Let A = [0, 1] and fn : A → R, fn = xn , n ∈ N and x ∈ A, then the sequence

{ xn } is a sequence of functions on [0, 1].

4.1.1 Bounded sequences

1- The sequence {fn (x)}, n = 1, 2, · · · is said to be bounded if there exists a function
f (x) such that
| fn (x) | ≤ |f (x)|, ∀x ∈ A
2- The sequence {fn (x)}, x ∈ A is said to be bounded on A if there exists the sequence
{an } such that
| fn (x) | ≤ |an |, ∀n = 1, 2, · · ·
3-The sequence {fn (x)} is said to be uniformly bounded on A if there exists a constant
M > 0 such that
|fn (x)| ≤ M ∀x ∈ A and all n.
The number M is called a uniform bound of {fn (x)}.

26
4.1.2 Convergence and divergence
(1)- Convergence at a point A sequence of functions { fn (x) } with domain A is said
to be convergent at xo ∈ A if the sequence of real numbers { fn (xo ) } is convergent.

(2)- Point wise convergence A sequence of functions { fn (x) } with domain A is

said to be converges pointwise on A if it converge at every point in A. i.e., We say that
the sequence of functions { fn (x) }, n = 1, 2, 3, · · · converges pointwise on the set A to the
function f (x) if

∀  > 0, ∃ N = N (, x) such that n > N () ⇒ |fn (x) − f (x)| < , ∀ x ∈ A

and then we write fn (x) → f (x) point wise on A.

(3)- Boundedly convergence A sequence of functions {fn (x)} is said to be buondedly

convergence on [a, b] if
i- The sequence converges pointwise on [a, b]
ii- The sequence is uniformly bounded.
ed
(4)- Uniform convergence We say that the sequence of functions { fn (x) }, n =
ay
1, 2, 3, · · · converges uniformly on the set A to the function f (x) if
l-S
.E

∀  > 0, ∃ N = N ()(depends only on ) such that n > N () ⇒ |fn (x)−f (x)| < , ∀ x ∈ A
.A

and then we write fn (x) → f (x) uniformly on A.

M
A.

Examples 4.2
x x 1
(1) Let { fn (x) } = { }, x ∈ R, then lim = x lim = 0, x ∈ R
n n→∞ n n→∞ n
n
(2) Let { fn (x) } = { x }, then for x ∈ [0, 1],
we find 
0 0 ≤x<1
fn (x) → f (x) = =
1 x = 1
n
(3) Let fn (x) = , x ∈ R, then fn (x) → 0, ∀ x ∈ R.
x2 + n2
1 + nx
(4)Letfn (x) : R → R be defined by fn (x) = , then fn (x) → x, ∀ x ∈ R.
n

Our chief interest in this chapter is to determine whether the important properties ( such
as continuity, differentiability and integrability are preserved under the limit operation.

To say that the function f (x) is continuous at the point x = c we mean that limx→c f (x) =

27
f (c). Hence to ask whether the limit function of the sequence of functions is continuous, is
the same as to ask whether

lim lim fn (x) = lim lim fn (x).

x→c n→∞ n→∞ x→c

We shall show by means of several examples that the limit processes cannot in general be
interchanged without affecting the results.

Examples 4.3
m
(1) Let Sm,n = , m, n = 1, 2, 3, · · · , then lim Sm,n = 1, ∀ fixed n ⇒ lim lim Sm,n = 1.
m+n m→∞ n→∞ m→∞

On the other hand

lim Sm,n = 0, ∀ fixed m ⇒ lim lim Sm,n = 0 ⇒

n→∞ m→∞ n→∞

lim lim Sm,n 6= lim lim Sm,n

m→∞ n→∞ n→∞ m→∞
n ed
(2) The sequence of functions { x }, x ∈ [0, 1], is a sequence of continuous functions, but
ay
the limit function f (x) where
l-S
.E


0 0 ≤x<1
fn (x) → f (x) = =
1 x = 1
.A
M

is discontinuous at x = 1.
A.

(3) Consider the sequence

sin nx 1
fn (x)} = { √ }, x ∈ R. Then lim fn (x) = lim √ sin nx = 0 = f (x), x ∈ R and so f 0 (x) = 0.
n n
√
Now fn0 (x) = n cos nx which does not converge

d d
⇒ lim fn (x) 6= lim fn (x)
dx n→∞ n→∞ dx

(4) Let fn (x) = n2 x(1 − x2 )n , x ∈ [0, 1]. Then we have

Z 1
n2
lim fn (x) = 0, x ∈ [0, 1] and fn (x) dx = → ∞, as n → ∞
n→∞ 0 2n + 2
Z 1 Z 1
⇒ lim fn (x) dx 6= lim fn (x) dx.
n→∞ 0 0 n→∞

28
 Examples 4.4
x
(1) Consider again the sequence { fn (x) } = { n
}. This sequence is convergent pointwise
on R but not uniformly, this is since

x |x| |x|
| fn (x) − 0 | = | | = < ⇒ n> = N (, x).
n n n
(2) Consider the sequence { nx } for x ∈ [0, 1]. Since

|x| 1 1
|fn (x) − 0| = < ⇒ n > = N ()
n n 
1
Hence ∀ , ∃ N () = [ ] such that n > N () ⇒ |fn (x) − 0| < , ∀ ∈ [0, 1].

sin nx
(3) For the sequence { √n } we have

sin nx 1 1
| √ − 0| ≤ √ <  ⇒ n > N (), N () = [ 2 ]
n n 
ed
ay
i.e., the sequence converges uniformly on R.
l-S

Theorem 3.1 (Cauchy criterion for uniform convergence)

.E

The sequence of functions {fn (x)}, x ∈ A, converges uniformly on A if and only if

.A
M

∀  > 0, ∃ N () such that |fm (x) − fn (x)| ≤ , whenever m > n > N (), x ∈ A.
A.

Examples 4.5
(1) Prove that the sequence {x(1 + 1/n)}, x ∈ R is uniformly convergent on any bounded
interval.
Proof. Let x ∈ [a, b] ⊂ R ⇒ |x| < |b|. Now since

1 1 n+m 2|b|
|fm (x) − fn (x)| = |x|| − | < |b| < <
m n mn n
2|b| 2|b|
⇒ n> > N () = [ ]
n n
i.e., ∀  > 0, ∃ N () such that |fm (x) − fn (x)| < , m > n > N () and x ∈ [a, b]
which proves that the sequence converges uniformly on [a, b].

x
(2) By the same way we can prove that the sequence { nx+1 }, x ∈ R converges uniformly.

29
4.1.3 Test for uniform convergence
Theorem 4.2 Suppose that lim fn (x) = f (x) pointwise on A, put Mn = sup|fn (x) − f (x)|.
Then
fn (x) → f (x) uniformly on A if and only if Mn → 0 as n → ∞.
Proof. Let fn (x) → f (x) uniformly, then ∀  > 0, ∃ N () such that n > N () ⇒
|fn (x) − f (x)| < , x ∈ A ⇒
sup |fn (x) − f (x)| < , n > N () i.e., n > N () ⇒ Mn < 
x∈A

which proves that Mn → 0, as n → ∞.

Conversely, suppose that Mn → 0, as n → ∞
i.e., ∀  > 0, ∃ N () > 0 such that n > N () ⇒ sup |fn (x) − f (x)| < .
x∈A

But x ∈ A and n > N () ⇒ |fn (x) − f (x)| < sup |fn (x) − f (x)| <  ⇒
x∈A

fn (x) → f (x) uniformly on A.

ed
ay
Examples 4.6
l-S
.E

(1) Consider again the sequence { nx }, x ∈ [0, 1]. fn (x) → 0 pointwise on [0, 1]. Now
sup |fn (x) − 0| = 1/n → 0 as n → ∞, then fn (x) → 0 uniformly on [0, 1].
.A
M

(2) Consider the sequence {xn }, x ∈ [0, 1]. Since

A.


0 0 ≤x<1
fn (x) → f (x) =
1 x = 1,
then Mn = sup |fn (x) − f (x)| = 1 and does not converge to 0 and the sequence converges
pointwise only.

n
(3) Consider the sequence { x2 +n2 }, x ∈ R. Then we find that

fn (x) → 0 forevery x ∈ R and Mn = sup |fn (x) − 0| = 1/n → 0 as n → ∞,

then fn (x) → 0 uniformly on R.

4.2 Series of function

Pn
Let SP
n (x) = k=1 fn (x) be a sequence ( of partial sums) of functions defined on A ⊂ R,
then fn (x) is a series of functions.

30
4.3 Convergence of the series of functions
P
The series fn (x) converges ( at a point on its domain, or pointwise, or uniformly ) if and
only if the sequence {Sn (x)} converges.
If the sequence {Sn (x)} converges to the function f (x) for every x ∈ A we write
∞
X
f (x) = fn (x) on A.
n=0

Theorem 4.3 (CauchyPcriterion)

The series of functions fn (x) converges uniformly on A if and only if for every  > 0
there exists N () > 0 such that
m
X
m > n > N () ⇒ |Sm − Sn | = | fk (x)| < , x ∈ A.
k=n+1

Theorem 4.4 (Weierstrass M-test)

Let {Mn } be a sequence of P
nonnegative real numbers such that |fn (x)| < M
Pn , n = 1, 2, · · ·
and x ∈ A. P ed
Then the series fn (x) converges uniformly on A if the series Mn converges.
ay
Proof. If Mn converges, then ( Cauchy criterion ) ∀  > 0, there exists N () such that
l-S
.E

m
X m
X m
X
m > n > N () and x ∈ A ⇒ | fk (x)| < fk (x) < Mk <  ⇒
.A

k=n+1 k=n+1 k=n+1

M

fn (x) → f (x) as n → ∞ and x ∈ A.

A.

Examples 4.7

1
P
The series x2 +n2
, x ∈ R is uniformly converges on R. This is since

1 1 X X 1
|fn (x)| = | | < = Mn and Mn = converges
x2 + n 2 n2 n2
P sin nx
(2) The series x2 +n2
, x ∈ R is uniformly converges on R. This is since

sin nx 1 X X 1
|fn (x)| = | | < = Mn and Mn = converges
x2 + n 2 n2 n2
Dirichlet’s Theorem Suppose that
(1) the sequence {an } is monotonic decreasing sequence of positive terms having limit zero,

(2) there exists a constant P such that for x ∈ [a, b]

|u1 (x) + u2 (x) + · · · un (x)| < P for all n > N.

31
Then the series ∞
X
a1 u1 (x) + a2 u2 (x) + · · · = an un (x)
n=1

is uniformly convergent in [a, b].

Exercises 4.8

(I) Test for uniform convergence

X 1 X 1 X (−1)n
(1) , x ∈ R (2) , p > 0, and x ∈ R (3) , p > 0, and x ∈ R.
x2 + n 3 x2 + np x2 + n p

4.3.1 Abel’s test and Dirichlet’s test

Theorem P 4.5 (Abel)
The series fP
n (x)gn (x) converges uniformly on [a, b] if
(1) The series fn (x) is uniformly converges on [a, b]

(2) The sequence {gn (x)} is uniformly bounded

ed
ay
(3) The sequence {gn (x)} is monotonic for every fixed x ∈ [a, b].
l-S
.E

Theorem P 4.6 (Dirichlet)

.A

The series fn (x)gn (x) converges

Pn uniformly on [a, b] if
(1) The sequence An (x) = k=1 fk (x) is uniformly bounded on [a, b]
M
A.

(2) The sequence {gn (x)} converges uniformly to 0

(3) The sequence {gn (x)} is monotonic decreasing for every fixed x ∈ [a, b].

4.4 Uniform convergence and continuity

Theorem 4.7 Assume that fn (x) → f (x) uniformly on A. If each term fn (x) is continuous
at a point c ∈ A, then the limit function f (x) is also continuous at the point c.
Proof. The uniform convergence of the sequence {fn (x)} to f (x) implies that

∀  > 0, ∃ N () such that n > N () and x ∈ A ⇒ |fn (x) − f (x)| < . (4.1)
3
Since fn (x), n = 1, 2, · is continuous at the point c ∈ A, then

∀  > 0, ∃ δ() such that |fn (x) − fn (c)| < , whenever |x − c| < δ(). (4.2)
3

32
Finally the uniform convergence of the sequence {fn (x)} implies the convergence of the
sequence {fn (c)} (Why ?), the

∀  > 0, ∃ N () such that n > N () ⇒ |fn (c) − f (c)| < . (4.3)
3
Adding (1) - (3) we get

|f (x) − f (c)| = |f (x) − fn (x) + fn (x) − fn (c) + fn (c) − f (c)| ≤

|f (x) − fn (x)| + |fn (x) − fn (c)| + |fn (c) − f (c)| < , whenever |x − c| < δ
⇒ f (x) is continuous at the point c ∈ A.

4.5 Uniform convergence and integrability

Riemann definition of integrals
Let P be a partition of the interval I = [a, b] and f (x) be defined on I. The real number
n
R(P, f ) =
X
ed
f (tk )(xk − xk−1 ), tk ∈ [xk , xk−1 ]
ay
k=1
l-S

is called the Riemann sum of the function f corresponding to the partition P = (xo , x1 , · · · , xn ).
.E

The function f (x) is called Riemann integrable on I if limn→∞ R(P, f ) exists as max (xk −
.A

xk−1 ) → 0, 1 ≤ k ≤ n and then

M

Z b
A.

f (x) dx = lim R(P, f )

a n→∞

Theorem 4.8 Let {fn (x)} be a sequence of integrable functions. If the sequence {fn (x)}
converges uniformly to the function f (x), then the function f (x) is integrable on [a, b] and
Z b Z b
lim fn (x) dx = lim fn (x) dx
n→∞ a a n→∞

Example 4.10
∞ Z π ∞
X sin nx X 1
let f (x) = 3
. Prove that f (x)dx = 2 .
n=1
n 0 n=1
(2n − 1)4

Solution Since we have | sinn3nx | ≤ n13 . Then by the Weierstrass M test the series is
uniformly convergent for all values of x, in particular 0 ≤ x ≤ π, and can be integrated
term by term. Thus
Z π Z π X ∞ ∞ Z π
sin nx X sin nx
f (x)dx = 3
dx = 3
dx
0 0 n=1
n n=1 0 n

33
 ∞ ∞
X 1 − cos nπ 1 1 X 1
= 4
= 2( 4 + 4 + · · · ) = 2
n=1
n 1 3 n=1
(2n − 1)4
Theorem 4.9 ( Arzela ) Assume that the sequence {fn (x)} is boundedly convergent on
[a, b] and suppose that each term of the sequence {fn (x)} is (Riemann) integrable on [a, b].
If the limit function f (x) = lim fn (x) is integrable on [a, b], then
Z b Z b Z b
lim fn (x) dx = lim fn (x) dx = f (x) dx.
n→∞ a a n→∞ a

Example 4.11
Verify Arzela’s theorem for the sequence {fn (x)} = {xn }, x ∈ [0, 1].

4.6 Uniform convergence and differentiation

Theorem 4.10 Assume that each term of the sequence {fn (x)} is a real-valued function
having a finite derivative at each point on an interval (a, b). Assume that for at least one
ed
point xo ∈ (a, b) the sequence {fn (xo )} converges. Assume further that there exists a func-
ay
tion g such that fn0 (x) → g(x) uniformly on (a, b). Then
l-S

(i) There exists a function f such that fn (x) → f (x) uniformly on (a, b)
.E

(ii) For each x ∈ (a, b) the derivative f 0 (x) and equals g(x).
.A
M

Theorem 3.11 Assume that fn is real valued differentiable functions on (a, b).
A.

Assume
P that P 0
fn (xo ) converges and fn (x) = P
g(x) uniformly on (a, b), then
(1) there exists a function f such that fn (x) = f (x) uniformly on (a, b).
0
(2) f (x) exists on (a, b) and

d X X d
fn (x) = f 0 (x) = fn (x).
dx dx

34
4.7 Exercises IV
P
(1) Prove that if the series fn (x) converges uniformly to f (x) and each fn (x) is continuous
at c ∈ A, then f (x) is continuous at c ∈ A and
∞
X ∞
X ∞
X
lim fn (x) = limx→c fn (x) = fn (c)
x→c
n=1 n=1 n=1
P
(2) Let fn (x) = 1, 2, · · · be integrable on [a, b]. If fn (x) converges uniformly to f (x),
then f (x) is integrable and
∞
Z bX ∞ Z
X b Z b
f (x) dx = fn (x) dx = f (x) dx.
a n=1 n=1 a a

P
(3) Discuss the differentiability of the series fn (x) and show when
∞ ∞
d X X d
fn (x) = fn (x).
dx n=1 dx
n=1
ed
ay
(4) Assume that fn (x) → f (x) and gn (x) → g(x) uniformly on A. Prove that fn (x) ±
l-S

gn (x) → f (x) ± g(x) uniformly on A and fn (x).gn (x) → f (x), g(x) uniformly on A if each
.E

of fn (x) and gn (x) are bounded on A

.A

1 x
(5) Let fn (x) = nx+1 and gn (x) = nx+1 , x ∈ (0, 1). Prove that fn (x) → 0 pointwise
M

but not uniformly and gn (x) → 0 uniformly on (0, 1)

A.

(6) Prove that if fn (x) = 11 + n2 x2 , x ∈ [0, 1], then {fn (x)} converges pointwise but
not uniformly on [0, 1].
P P
(7) Discuss the uniform convergence of an .sin nx and an .cos nx on R

x
P
(8) Discuss the uniform convergence of (1+nx2 )nα
for the values of x and α.

(9) Let fn (x) = xn . Prove that the sequence {fn (x)} converges pointwise but not uni-
formly on [0, 1]. Let g(x) be a continuous function on [0, 1] with g(1) = 0, prove that the
sequence {xn g(x)} converges uniformly on [0, 1].

x
(10) Let fn (x) = 1+nx 2 , x ∈ R. Find the limit function f (x) of the sequence {fn (x)}
and the limit function g(x) of the sequence {fn0 (x)}.
a - Prove that fn0 (x) exists for every x but f 0 (0) 6= g(0). For what values of x f 0 (x) = g(x).
b - In what subinterval of R does fn (x) → f (x) and fn0 (x) → g(x) uniformly.

(11) Let {fn (x)} be a sequence of real function such that fn (x) → f (x) uniformly on

35
[a, b]. Prove that
Z x Z x
φn (x) = fn (t) dt → φ(x) = f (t) dt uniformly on [a, b]
a a
P 1
(12) For what values of x the series 1+n2 x
converges uniformly and absolutely. Is the
limit f (x) of the series continuous in the case of convergence.
2
(−1)n x n+n
P
(13) Prove that the series 2 converges uniformly in every bounded interval
but does not converge absolutely for any values of x.

ed
ay
l-S
.E
.A
M
A.

36
Chapter 5

Power series

A power series is a series of functions of the form

∞
X
2 3 n
ao + a1 x + a2 x + a3 x + · · · + an x + · · · = an x n
n=1

ed
where ao , a1 , a2 , · · · are constants called coefficients of the power series.
ay
l-S

Examples 5.1
.E

xn is a power series with an = 1.

P
(1) The series
.A
M

P xn
(2) The series n!
is a power series with an = 1/n!
A.

By the way the convergence and divergence of the power series is the same as the series of
functions. But here we have some special results.

5.1 Radius of convergence

an x n =
P P
Consider the power series un , to determine the convergence of this series,
apply the ratio test, then

un+1 an+1 |x| 1 an+1

lim | | = lim | | |x| = , where = lim | |,
n→∞ un n→∞ an R R n→∞ an

then the series converges if |x|

R
< 1 ⇒ |x| < R and diverges if |x| > R.
This R is called the radius of convergence and the interval (−R, R) is called interval of
convergence.

Examples 5.2

37
(1) The series 1 + x + x2 + · · · + xn + · · · =
P n
x has the radius of convergence R = 1 and
the series converges in (−1, 1). Now x = 1 ⇒ an = 1, x = −1 ⇒ an = (−1)n ⇒ an
does not converge to zero and the series diverges. So the series converges when |x| < 1 and
diverges when |x| ≥ 1.

(2) Determine the interval of convergence of the series

∞
(2x) (2x)2 (2x)3 X (2x)n
− + − ··· = (−1)n .
1 2 3 n=1
n
n
The series has the radius of convergence R1 = lim | (−1) n
n+1 (−1)n
| = 1. Then the series con-
verges when |2x| < 1 and diverges when |2x| > 1. Now 2x = 1 ⇒ x = 1/2, the series will
be 1−1/2+1/3−1/4+· · · and the series converges (Why ?). also 2x = −1 ⇒ x = −1/2,
the series will be 1 + 1/2 + 1/3 + 1/4 + · · · and the series diverges (Why ?). Then the
series converges in (−1/2, 1/2].
P xn
(3) Determine the interval of convergence of the series .
ed n!
ay
1 1 n! 1
= lim | | = lim =0 ⇒ R=∞ ⇒
l-S

R n→∞ (n + 1)! 1 n→∞ n + 1

.E

the series converges in (−∞, ∞)

.A
M

5.2 Uniform convergence

A.

an xn converges in (−R, R), then it is uniformly con-

P
Theorem 5.1 If the power series
verges in [−r, r], 0 < r < R.

|an |rn . Since

P P
Proof. Let un (x) =
|an+1 | rn+1 an+1 r
n
= | |r = < 1.
|an | r an R
an xn is uniformly convergent in [−r, r]
P
Then by the Weierstrass -M test the power series

Example 5.3
Prove that
x3 x5 x7 π 1 1 1
tan−1 x = x − + − + · · · and = 1 − + − + ···
3 5 7 4 3 5 7
Proof.
∞
X
Since (1 + u2 )−1 = (−1)n u2n , |u| < 1. Then for x < 1 we have
0

38
 Z x Z x ∞
X
−1 2 −1
tan x = (1 + u ) du = (−1)n u2n du ⇒
0 0 0

x3 x5 x7
tan−1 x = x −
+ − + ···
3 5 7
and from the convergence of the series at x = 1 we get the result.

5.3 Multiplication of power series

an xn with radius P
P
Theorem 5.2 P Given the two power series f (x) = of convergence
n
R and g(x) = bn x with radius of convergence r, then f (x)g(x) = cn x n , x ∈
(−R, R) ∩ (−r, r).

5.4 Exercises V
an xn . Let 0 < r < R. Prove
ed
P
(1) Let R be the radius ofPconvergence of the power series
that the function f (x) = an xn is continuous in [−r, r]
ay
l-S

an xn . Let 0 < r < R.

P
(2) Let R be the radius of convergence of the power series
.E

n
P
Prove that the function f (x) = an x is integrable on [−r, r] and
.A

x x ∞ ∞ Z x ∞
an x n
Z Z
M

X X X
n n
f (t) dt = an t dt = an t dt = , ∀x<R
A.

0 0 n=0 n=0 0 n=0

n + 1

and the last series has the radius of convergence R.

an xn . Let 0 < r < R.

P
(3) Let R be the radius of convergence of the power series
Prove that the function f (x) = an xn is differentiable on [−r, r] and
P

∞ ∞ ∞
d d X n
X d n
X
f (x) = an x = an x = n an xn−1 , ∀ x < R
dx dx n=0 n=0
dx n=0

and the last series has the radius of convergence R. Also prove that an = f n (0)/n!, n =
1, 2, 3, · · · .

an xn and bn xn converge on some interval

P P
(4) Prove that if the two power series
(−R, R), R > 0 to the same function f (x), then an = bn , ∀n.

(5) Prove theorem 5.2

39
(6) Find the interval of convergence of the following power series
X (x − 1)n X (x − 3)2n+1 X (x − 2)2n X (x − 5)n
(i) (ii) (−1)n+1 (iii) (−1)n+2 (iv)
n! (2n + 1)! (2n)! n2

(7) Evaluate the integral

1 2
1 − e−x
Z
dx
0 x2
to 3 decimal place accuracy.

(8) Prove that

x2 x3 x4 1 1 1
ln(1 + x) = x − + − + ··· and ln 2 = 1 − + − + ··· .
2 3 4 2 3 4

ed
ay
l-S
.E
.A
M
A.

40
Chapter 6

Improper integrals

In the Riemann-Stieltjes definition of integration over the interval [a, b] both the interval
and the function (integrand) are bounded. We now consider the cases when either the
interval or the function is unbounded.

6.1 First kind of improper integrals ed

ay
l-S

It is an integral of the form

.E

Z ∞ Z b
(1) f (x) dx or (2) f (x) dx
.A

a −∞
M

where f (x) is bounded and integrable on any bounded interval [a, b].
A.

Definition 6.1 The improper integral (1) is said to be convergent if

Z ∞ Z b
f (x) dx = lim f (x) dx exists
a b→∞ a

and the improper integral (2) is said to be convergent if

Z b Z b
f (x) dx = lim f (x) dx exists
−∞ a→−∞ a

If the limits does not exist, then the integrals are said to be divergent.
In like mannar, we define
Z ∞ Z a Z ∞
f (x) dx = f (x) dx + f (x) dx
−∞ −∞ a

where a is a real number, and call the integral convergent or divergent according as the
integrals on the right converge or not.

41
Examples 6.1
Discuss the convergence and divergence of the integrals
Z ∞ Z ∞ Z ∞
dx −α x dx
(1) p
, a > 0, (2) e dx, a > 0 and (3) x −x
a x a −∞ e + e

Solution (1) For p = 1 we have

Z ∞ Z b
dx dx
= lim = lim (ln b − ln a) = ln ∞ − ln a = ∞
a x b→∞ a x b→∞

Since for p 6= 1, we have

b
x1−p x=b b1−p a1−p
Z
dx
= |x=a = −
a xp 1−p 1−p 1−p
and, then
∞
b1−p a1−p a1−p
Z
dx
p>1 ⇒
a xp
= lim (
b→∞ 1 − p
− ed
1−p
) = −
1−p
ay
and
l-S

Z b
dx
p<1 ⇒ → ∞.
.E

a xp
.A

Then the first integral converges if p > 1 and diverges for p ≤ 1.

M
A.

(2) Let α 6= 0, then

b
e−α x b
Z
1
e−α x
dx = − |a = (e−aα a − e−bα ) ⇒
a α α
b Z b
e−aα
Z
−α
lim e x dx = if α > 0 and lim e−α x dx = ∞ if α < 0.
b→∞ a a b→∞ a
R∞
Also at α = 0 we have a dx = ∞.
Then the integral (2) converges if α > 0 and diverges if α ≤ 0.
Z ∞ Z ∞
x dx du
(3) Letting u = e we find that x −x
= 2
−∞ e + e 0 1 + u2
Z b
du
= lim = 2 lim tan−1 b = 2 tan−1 ∞ = π
b→∞ 0 1 + u2 b→∞

and the integral converges.

42
6.2 Tests of convergence
Theorem 6.1 (Comparison
R∞ test) Let R0 ≤ f (x) ≤ g(x), x ≥ a, then
∞
(1) The integral a f (x) dx converges if a g(x) dx converges.
R∞ R∞
(2) The integral a
g(x) dx diverges if a
f (x) dx diverges.

Exercises 6.1
Discuss the convergence of the integrals
Z ∞ Z b Z ∞
dx
(1) 2
(2) cos x dx (3) e−α x cos β x dx, α > 0
0 1 + x −∞ 0
Z ∞ Z ∞ Z ∞
dx dx cos2 x
(4) √ dx (5) dx (6) dx
1 x3 + x 0 ex + 1 2 x2
Z ∞ √
Z ∞
ln x
Z ∞
√
dx
√
ed x
ay
(7) dx (8) dx (9) dx.
3 x 2 x− x3
1 1+x
l-S
.E

Theorem 6.2 (Ratio test) Let 0 ≤ f (x), g(x), x ≥ a, then

.A

Z ∞ Z ∞
f (x)
(1) If lim = k, k 6= 0 or ∞, then f (x) dx and g(x) dx
M

x→∞ g(x) a a
A.

converges or diverges together.

R∞ R∞
(2) If k = 0 and a
g(x) dx converges, then a
f (x) dx converges.
R∞ R∞
(3) If k = ∞ and a
g(x) dx diverges, then a
f (x) dx diverges.

6.3 Examples
(1) Test for convergence the two integrals
Z ∞ Z ∞
dx 1 + x
(1) 3
√ , (2) √ dx.
1 x + x 2 1 + x4
From the first integral, for x ≥ 1 we have
Z ∞
1 1 1
f (x) = 3 √ ≤ 3 . But dx
x + x x 1 x3

43
converges, ( Why?), then the first integral converges.
For the second integral, let
1 + x 1 f (x)
f (x) = √ and g(x) = , then lim = 0 = k.
1 + x4 x x→∞ g(x)
R∞ R∞
But 2 g(x) dx diverges ( Why?), then by the ratio test the integral 2 f (x) dx
diverges.

Exercise 6.2
Discuss the convergence or divergence of the integrals
Z ∞ Z ∞ ∞
x2
Z
x+1 2
(1) √ dx (2) e−x dx (3) dx.
2 1 + x4 1 1 3x4 + 5

6.4 Absolute convergence

R∞
RThe
∞
improper integral a f (x) dx is said to be absolutely convergent if the integral
ed
|f (x)| dx is convergent.
ay
a
If the integral converges but not absolutely converge, then it is said to be conditionally
l-S

convergent.
.E

Theorem 6.3 Every absolutely convergent integral is convergent.

.A

R∞
Theorem R ∞6.4 Let f (x) ≥ 0 and |g(x)| ≤ M. If f (x) dx is convergent, then the
M

a
integral a f (x) g(x) dx is convergent.
A.

Example 6.2R
∞ R∞ R∞
The integral 0 cos x
x2 +1
dx converges absolutely, this is since 0 | cos x
x2 +1
| dx ≤ 0 1
x2 +1
dx
which is convergent.

6.5 Second kind of improper integrals

Rb
It is an integral of the form a f (x) dx, where a and b are finite but f (x) is not bounded for
some points in [a, b], hence these points are called singular points and we have the following
three cases
(1) If x = a is the only singular point, then
Z b Z b
f (x) dx = lim+ f (x) dx.
a →0 a+

(2) If x = b is the only singular point, then

Z b Z b−
f (x) dx = lim+ f (x) dx.
a →0 a

44
(3) If c ∈ (a, b) is the singular point, then
Z b Z c Z b
f (x) dx = f (x) dx + f (x) dx =
a a c
Z c−1 Z b
= lim f (x) dx + lim+ f (x) dx.
1 →0+ a 2 →0 c+2

(4) Cauchy principal value

It may be happen that the limits does not exists when 1 and 2 → 0 in dependently in
(3). In such case it is possible to choose 1 = 2 = , if
Z b Z c− Z b
f (x) dx = lim+ ( f (x) dx + f (x) dx
a →0 a c+

exists, we call it the Cauchy principal value.

Examples 6.3
Discuss the convergence of the integrals
Z b
dx
Z b
dx ed
ay
(1) p
dx (2) p
dx.
a (x − a) a (b − x)
l-S

Solution Since
.E

Z b Z b
dx dx
dx = lim+
.A

p p
a (x − a) → 0 a+ (x − a)
M

1 1 1 1
= lim+ ( − p−1 ) = (b − a)1−p , p < 1,
A.

→ 0 1 − p (b − a) p−1 () 1−p

then the integral converges if p < 1 and diverges if p ≥ 1, since for P = 1, we have
Z b
dx
= lim [ln (b − a) − ln ] → − ∞.
a x−a → 0

The same result can be proved for the second integral.

Rb
Theorem 6.5 Let f (x) be unbounded only at the lower limit of the integral a
f (x) dx
and limx→ a+ (x − a)p f (x) = A. Then
Rb
(1) a f (x) dx converges if p < 1 and A is finite
Rb
(2) a f (x) dx diverges if p ≥ 1 and A 6= 0 (A may be infinite).

Theorem 6.6 Let f (x) be unbounded only at the upper limit x = b and
limx→ b− (b − x)p f (x) = A. Then
Rb
(1) a f (x) dx converges if p < 1 and A is finite
Rb
(2) a
f (x) dx diverges if p ≥ 1 and A 6= 0 (A may be infinite).

45
Exercise 6.4
Discuss the convergence and divergence of the integrals
Z 2 Z 4 Z 1
dx ln x dx cos x dx
(1) √ , (2) 2
, (3) √ .
1
3
x −1 3 (x − 3) 0 x

6.6 Third kind of improper integrals

It can be expressed in terms of the first and second kinds. For example
Z ∞ Z 1 Z ∞
n−1 −x n−1 −x
x e dx = x e dx + xn−1 e−x dx
0 0 1

which is convergent (Why ?).

6.7 Uniform convergence

The integral φ(α) =
R∞ ed
f (x, α) dx is said to be uniformly convergent in [α1 , α2 ] if
ay
a
l-S

Z ∞ Z n
∀  > 0, ∃ N () only, such that n > N () ⇒ | f (x, α) dx| = |φ(α)− f (x, α) dx| < 
.E

n a
.A

∀ α ∈ [α1 , α2 ].
M
A.

Weierstrass M-test
If we can find a function M (x) ≥ 0 such that
Z ∞
(i) | f (x, α) | ≤ M (x), α ∈ [α1 , α2 ], x > a (ii) M (x) dx converges,
a
R∞
then the integral a
f (x, α) dx is uniformly and absolutely convergent in [α1 , α2 ]

Exercise 5.5
Discuss the uniform convergent of the integrals
Z ∞ Z ∞ ∞
sin2 α x
Z
cos α x cos α x
(1) dx (2) dx (3) dx
0 x2 + 1 1 xp 0 1 + xp

6.8 Properties of uniform convergence

R∞
1- If f (x, α) is continuous for x ≥ a and α ∈ [α1 , α2 ] and if the integral a
f (x, α) dx is
uniformly convergent for α ∈ [α1 , α2 ], then

46
 R∞
(1) φ(α) = a
f (x, α) dx is continuous in [α1 , α2 ]
Z ∞ Z ∞
(2) lim f (x, α) dx = lim f (x, α) dx, α ∈ [α1 , α2 ]
α→αo a a α→αo
Z α2 Z α2 Z ∞ Z ∞ Z α2
(3) φ(α) d α = f (x, α) dx d α = f (x, α) d α dx, α ∈ [α1 , α2 ]
α1 α1 a a α1

(4) If f (x, α) is continuous and has continuous derivative with respect to α for x ≥ a, α ∈
R ∞ (x,α)
[α1 , α2 ] and if a ∂f∂α dx converges uniformly in [α1 , α2 ], then
Z ∞
dφ(α) ∂f (x, α)
= dx, a does not depend on α
dα a ∂α
Remark
∂f (x,α)
If f (x, α) and ∂α
are continuous in both x and α and
Z g2 (α)
φ(α, x) = f (x, α) dx,
g1 (α)

then ed
ay
l-S

Z g2 (α)
∂φ(α, x) ∂f (x, α) dg2 (α) dg1 (α)
= dx + f (g2 (α), α) − f (g1 (α), α) .
∂α ∂α dα dα
.E

g1 (α)
.A

Example 6.4 Prove that

M

Z ∞
dx π
= α−3/2
A.

2
(α + x )2 4
0
R∞
Proof. Consider the integral 0 x2dx +α
. Since
Z ∞ Z ∞
1 1 dx π dx
≤ , and = ⇒ converges ∀ α ≥ 1.
x2 + α x2 + 1 0 x2 + 1 2 0 x2 + α
Z ∞
dx 1 x π
Also 2
= lim √ tan−1 √ |b0 = √ .
0 x +α b→∞ α α 2 α
Now
∞
−1
Z
∂ 1 1 1 ∂
2
= 2 , and ≤ 2 ⇒ f (x, α) dx
∂α x + α (x + α)2 2
(x + α)2 (x + α) 0 ∂α
converges uniformly for all α ≥ 1, and then
Z ∞ Z ∞
d ∂ dx dx d π π
φ(α) = 2
= − 2 2
= − √ = − α−3/2 .
dα 0 ∂ α x +α 0 (x + α) dα 2 α 4

47
6.9 Exercises VI
Discuss the convergence and the uniform convergence of the integrals
Z ∞ Z ∞ Z ∞ Z ∞
1 cos x 1 cos α x
(1) dx (2) dx (3) dx (4) dx
0 1 + xp 0 1 + xp 0 α + xp 0 α + xp
Z ∞ −α x Z ∞ −α x Z ∞
e e 1
(5) p
dx (6) p
dx (7) dx.
0 1+x 0 α+x 0 xp

ed
ay
l-S
.E
.A
M
A.

48
Chapter 7

Introduction to
Fourier analysis

7.1 Periodic functions

ed
A function f (x) is said to have a period T or to be periodic with period T if for all
ay
x, f (x + T ) = f (x), where T is a positive constant.
l-S

Examples 7.1
.E

(1) Since sin (x + 2nπ) = sin x, n = 1, 2, 2, · · · . Then the function sin x is periodic
.A

with period 2π
M
A.

(2) Also the function cos x is periodic with period 2π, this is since cos (x+2nπ) = cos x, n =
1, 2, 3, · · · .

(3) The function tan x is periodic with period π, this is since tan (x+nπ) = tan x, n =
1, 2, 3, · · · .

(4) Any constant function is periodic.

Exercise 7.1
I- Graph each of the following functions (1) f (x) with period 10, where

3, 0 < x < 5
f (x) =
− 3, − 5 < x < 0.

(2) f (x) with period 2π, where


sin x, 0 ≤ x ≤ π
f (x) =
0, π < x < 2π.

49
(3) f (x) with period 6, where

 0, 0 ≤ x < 2
f (x) = 1, 2 ≤ x < 4
0 4 ≤ x < 6.


II- Prove the following relations

Z L Z L 
mπ x nπ x mπ x nπ x 0, m 6= n
cos cos dx = sos sos dx =
−L L L −L L L L, m = n.
Z L
mπ x nπ x
cos sos dx = 0, for any values of m, n.
−L L L
Z L Z L
mπ x mπ x
cos = sos dx = 0, m = 1, 2, 3, · · ·
−L L 0 L

III- Consider the series

∞
A +
X
(an cos
nπx ed
+ bn sin
nπx
). (7.1)
ay
n=1
L L
l-S

If the series (7.1) converges uniformly to the function f (x) in the interval (−L, L), then by
.E

using the properties of the sin and cos functions ( in exercise II) to prove prove that
.A

Z L Z L
1 nπx 1 nπx ao
M

an = f (x) cos dx, bn = f (x) sin dx, and A = . (7.2)

A.

L −L L L −L L 2

From this fact we can study the Fourier series corresponding to the function f (x) defined
on an interval (−L, L) or (c, 2L) where c and L are any real numbers.

7.2 Fourier series

Let f (x) be defined in the interval (L, −L) and outside of this interval by f (x + 2L) = f (x)
i.e. assume that f (x) has the period 2L.
The Fourier series ( Fourier expansion ) corresponding to f (x) is given by the series (7.1)
where the Fourier coefficients an and bn are given by (7.2).
If f (x) has the period 2L , the coefficients an and bn can be determined equivalently from
Z c+2L Z c+2L
1 nπx 1 nπx
an = f (x) cos dx, bn = f (x) sin dx.
L c L L c L
Theorem (Dirichlet) Suppose that
(1) f (x) is defined and single valued except possibly at a finite number of points in (−L, L).

50
(2) f (x) is periodic outside (−L, L) with period 2L.

(3) f (x) and f 0 (x) are sectionally continuous in (−L, L),

then the series (7.1) with coefficients (7.2) converges to f (x) at each continuous point
x ∈ (−L, L) and to 1/2(f (x + 0) + f (x − 0)) for other points of (−L, L)

7.3 Half range series

If f (x) is even function on (−L, L), then
Z L ∞
2 nπx X nπx
an = 0, and bn = f (x) sin dx and f (x) has sin series bn sin .
L 0 L n=1
L

If f (x) is odd function on (−L, L), then

Z L ∞
2 nπx X nπx
bn = 0, and an = f (x) cos dx and f (x) has cos series an cos .
L 0 L n=1
L

Examples 7.2
ed
ay
l-S

(1) Find the Fourier series corresponding to the function

.E


0 −5 ≤x<0
.A

f (x) = =
3 0≤x≤5
M
A.

with period 5.
Solution The period is 2L = 10 ⇒ L = 5. Choose the interval c to c + 2L as −5 to 5 ,
so that c = −5. Then
Z c+2L
1 0
Z Z 5
1 nπx nπx nπx
an = f (x)cos dx = ( 0 cos dx + 3 cos dx = 0, if n 6= 0
L c L 5 −5 L 0 L
3(1−cosnπ)
and if n = 0, then ao = 3. Also we can get bn = nπ
. The corresponding Fourier
series is
∞
3 X 3(1 − cosnπ) nπx 3 6 nπ 1 3nπ
+ sin = + (sin + sin + . . . ).
2 n=1 nπ 5 2 π 5 3 5

From the convergence of the series to f (x) we have

 3

 2
, x = −5
 0, −5<x<0


3
f (x) = = 2
, x = 0
3, 0<x<5



 3
, x = 5

2

51
with period 10. Then the series converges to the function f (x) for −5 < x < 5.

(2) Expand f (x) = x2 , 0 < x < 2π in a Fourier series with period (i) 2π and (ii)
not specified period.
Solution The period 2L = 2π ⇒ L = π. Choose c = 0, we can get an = 4/π 2 , n 6=
0, ao = 8π 2 /3 and bn = −4π/n. Then
∞
2 4π 2 X 4 4π
f (x) = x = + ( 2 cos nπ − sin nπ), 0 < x < 2π
3 n=1
n n

and the series converges to 2π 2 at x = 0 and x = 2π. From this result, at x = 0 we can
deduce that ∞ ∞
4π 2 X 4 2
X 1 π2
+ = 2π ⇒ = .
3 n=1
n2 n=1
n2 6
If the period is not specified, the Fourier series can not be determined uniquely in general.

Exercises 7.2
ed
ay
(1) Expand f (x) = x, 0 < x < 2 with period 2.
l-S
.E

(2) Expand f (x) = x, 0 < x < 2 in a half range (i) sine series (ii) cosine series.
.A

(3) Expand f (x) = sin x, 0 < x < π in a Fourier cosine series.

M
A.

(4) Expand f (x) = cos x, 0 < x < π in a Fourier sine series.

7.4 Parseval’s identity

Assuming that the Fourier series (7.1) corresponding to the function f (x) converges uni-
formly to f (x) in (−L, L), then
∞
ao X nπx nπx
f (x) = + (an cos + bn sin ).
2 n=1
L L

Multiplying by f (x) and integrating we get Parseval’s identity

L ∞
ao 2
Z
1 2
X
[f (x)] dx = = (a2n + b2n )
L −L 2 n=1

52
7.5 Bessel’s inequality
Let an and bn be the Fourier coefficients corresponding to the function f (x). If f (x) is
sectionally continuous on (−L, L) , then
∞
a2o X 2 1 L
Z
2
+ (an + bn ) ≤ {f (x)}2 dx (7.3)
2 n=1
L −L
which is the Bessel’s inequality.
Proof. Let
M
ao X nπx nπx
SM (x) = + (an cos + bn sin ), M = 1, 2, · · · (7.4)
2 n=1
L L
This sequence is the sequence of partial sums of the Fourier series corresponding to f (x).
Now Z L
2
(f (x) − SM (x)) ≥ 0 ⇒ (f (x) − SM (x))2 dx ≥ 0 ⇒
−L
Z L Z L Z L
2
2 f (x) SM (x) dx − {SM (x)} dx ≤ {f (x)}2 dx. (7.5)
−L −L ed −L
ay
Multiplying both sides of (7.4) by 2f (x) and integrating from −L to L, we get
l-S

Z L M
a2 X 2
.E

2 f (x) SM (x) dx = 2L ( o + (an + b2n )), M = 1, 2, · · · (7.6)

−L 2
.A

n=1

Squaring (7.4) and integrating from −l to L, we obtain

M
A.

Z L M
2 a2o X 2
{SM (x)} dx = L ( + (an + b2n )), M = 1, 2, · · · (7.7)
−L 2 n=1

Substituting of (7.6) and (7.7) into (7.5) and dividing by L, we get

M
a2o X 2 1 L
Z
2
+ (an + bn ) ≤ {f (x)}2 dx, M = 1, 2, · · · (7.8)
2 n=1
L −L

Now the sequence in the left hand side of (7.8) is bounded and monotonic (Why ?), then
it converges. Taking the limit as n → ∞ of (7.8) we get Bessel’s inequality.

Riemann’s theorem
If f (x) is sectionally continuous, then
Z π Z π
lim f (x) sin nx dx = lim f (x) cos nx dx = 0
n→∞ −π n→∞ −π

Proof. From Bessel’s inequality we have an → 0 and bn → 0 and from the definition of
an and bn with L = π we obtain the results.

53
7.6 Differentiation and integration of Fourier series
Differentiation and integration of Fourier series can be justified by using the theorems of
differentiation and integration of the series of functions which hold for series in general.
The following theorem for integration is especially useful.

Theorem The Fourier series corresponding to the function f (x) may be integrated term
R x term from a to x where a, x ∈ (−L, L) and the resulting series converges uniformly to
by
a
f (x) dx provided that the function f (x) is sectionally continuous in (−L, L).

7.7 Exercise VII

(1) Find the Fourier series corresponding to the function f (x), 0 < x < 2 in a half range
sine series.
(2) By integrating the resulting series, in (1), term by term find the Fourier series for the
function f (x) = x2 , 0 < x < 2 and then use the result to evaluate the value of the series
∞
ed
ay
X (−1)n
.
l-S

n=1
n2
.E

(3) Prove Parseval’s identity.

.A
M
A.

54