MA101-CALCULUS

Lecture 3
SEQUENCES OF REAL NUMBERS

G Sankara Raju Kosuru

Department of Mathematics
IIT Ropar
[email protected]

August 2, 2024
Recall
Sequence of real numbers
A sequence in R is a function from N to R. Denoted by hxn i or {xn }∞
n−1 or
simply {xn } or (xn ).
Convergent sequence
A sequence hxn i is said to be Convergent to ` for every ε > 0 ∃N(= N() ∈
N such that
|xn − `| < ε ∀n > N.
Equivalently, xn ∈ (` − , ` + ) for all n ≥ N.
lim xn = ` or hxn i 7→ ` or xn → ` or x is the limit of hxn i.
n→∞

Properties
1 Every convergent sequence is bounded
2 If a sequence is convergent, then the limit is unique
3 Algebra of limits
4 Sandwich (or squeeze) theorem
G. S. Raju (IIT Ropar) SEQUENCES 2 / 14
Examples
1. h sinn n i converges 0.
Note that −1 ≤ sin n ≤ 1.
Then it follows that − n1 ≤ sinn n ≤ n1 ∀n ∈ N.
∴ −1 1 sin n
n → 0 & n → 0 by Sandwich/Squeeze theorem, we get n → 0.
2. Let x ∈ R with |x| < 1. Define xn = x n for n ∈ N. Discuss the
convergence of the sequence hxn i.
1
If x = 0, then clearly xn −→ 0. Suppose x 6= 0. Then |x| > 1 and
1
hence we can write = 1 + k with k > 0
|x|
 
1 n n
|x|n = (1 + k) = 1 + nk + k 2 + . . . + k n > nk ∀n ∈ N.
2
(using Binomial theorem)
1
∴ 0 < |xn | = |x|n < → 0 as n → ∞.
nk

1
Thus, 0 < |x n | = |x|n < nk . By Sandwich theorem, we have x n → 0.
G. S. Raju (IIT Ropar) SEQUENCES 3 / 14
Theorem (Monotone converges theorem)
A bounded increasing sequence converges. (In fact it converges to its
lub/suprimum)

Proof: Let hxn i be an increasing and bounded sequence. Let

L = sup{xn : n ∈ N}. We claim that xn → L. Suppose the error  > 0 be
given. by definition of the suprimum, there exists N such that

L −  < xN ≤ L. (1)

As {xn } is increasing, we have xn ≤ xn+1 . Thus we get

L −  < xn < L +  for all n ≥ N.

Hence lim xn = L.
n→∞
G. S. Raju (IIT Ropar) SEQUENCES 4 / 14
Theorem
A bounded deceasing sequence converges. (In fact it converges to its
glb/infimum)

Theorem
A monotone sequence is convergence if and only if it is bounded.
Example 1
1 1 1
Consider xn = 1 + 1! + 2! + · · · + n! . Clearly hxn i is an increasing sequence.
 
1 1 1 1 1
Also xn ≤ 1 + + + 2 + · · · + n−1 = 1 + 2 1 − n < 3.
1 2 2 2 2

∴ By Monotone convergence theorem, hxn i is Convergent.

n
X 1
This limit is known as the Euler constant e. Thus e = lim .
n→∞ k!
k=0
G. S. Raju (IIT Ropar) SEQUENCES 5 / 14
Cauchy sequence
Motivation
Let us recall the notation of convergent sequences. A sequence {xn } is said
to converges to x, if for given error  > 0, there exists a stage N = N()
such that
|xn − x| < , for n ≥ N. (2)

Therefore for m, n ≥ N,

|xn − xm | = |xn − x + x − xm |
≤ |xn − x| + |x − xm |
< +
G. S. Raju (IIT Ropar) SEQUENCES 6 / 14
Cauchy sequence
Definition
A sequence {xn } is said to be Cauchy if for given error  > 0, there exists
a stage N = N() such that |xn − xm | < , for n, m ≥ N.

To verify a sequence to be Cauchy, it is not mandatory to have the limit.

Theorem
A sequence is Cauchy if and only if it is convergent (and hence bounded).

Let {xn } be a sequence. If {xn } is not Cauchy, then it is not a convergent

sequence.
G. S. Raju (IIT Ropar) SEQUENCES 7 / 14
Example
n
1 1 1 X 1
Let xn = + + ··· = for n ∈ N. Then is the sequence hxn i
1 2 n k
k=1
convergent.
Slon: Chooose the error  = 12 . Then for any N ∈ N, we have
2N N
X 1 X1
|x2N − xN | = −
k k
k=1 k=1
2N
X 1 1 1 1
= = + + ··· +
k N +1 N +2 2N
k=N1
1 1 1 N 1
> + + ··· + = = = .
2N 2N 2N 2N 2
Thus the sequence is not Cauchy, hence not convergent.
* n +
X1
The harmonics sequence is unbounded.
k
k=1
G. S. Raju (IIT Ropar) SEQUENCES 8 / 14
Unboundedness of harmonic sequence
n
1 1 1 X 1
Let xn = + + ··· = for n ∈ N. Then the sequence hxn i
1 2 n k
k=1
unbounded.
Slon: For any n ∈ R.
n
2
X 1 1 1 1
|x2n | = = + + ··· + n
k 1 2 2
k=1
     
1 1 1 1 1 1 1 1
= + + + + + ··· 3 + ··· + + ··· + n
1 2 3 4 22 + 1 2 2n−1 + 1 2
     
1 1 1 1 1 1 1 1
> + + + 2 + + ··· 3 + ··· + + ··· + n
1 2 22 2 3 2 23 2
   2  n−1 
1 2 2 2
= 1+ + + + ··· +
2 23 23 2n
     
1 1 1 1 n
= 1+ + + + ··· + =1+
2 2 2 2 2

As n is arbitrary, hxn i is unbounded

G. S. Raju (IIT Ropar) SEQUENCES 9 / 14
Sub-sequences
Motivation
Consider the sequence h1, 1, 21 , 2, 13 , 3, 14 , 5, · · · i. That is x2n−1 = n for all
1
n ∈ N and x2n = for all n ∈ N.
n
1 Is it a bounded sequence?

2 Any terms of the sequence are interesting?

3 What is the patron of the terms?

Definition
Let {xn } be a sequence. Let A ⊆ N be an infinte set with unceasing order
(i.e. A = {nk : k ∈ N : nk < nk+1 for all k ∈ N}). The part {xnk } of the
sequence is known as a subsequence of {xn }.

It is well understood in this case that k ≤ nk .

In other words, a subsequence is formed by considering any infinite subcol-
lection of the terms of a sequence without changing the order.
G. S. Raju (IIT Ropar) SEQUENCES 10 / 14
Examples
1
1
n is a subsequence of h1, −1, 12 , −2, 31 , −3, 41 , −5, · · · i
2 hni is a subsequence of h1, 1, 21 , 2, 31 , 3, 41 , 5, · · · i
3 {x2n } is a subsequence of {xn } and is known as even sequence.
4 {x2n+1 } is a subsequence of {xn } and is known as odd sequence.

Theorem
1 Every subsequence of a bounded sequences is bounded.
2 Every subsequence of a convergent sequence is convergent. Further
the limit is same.
3 Every subsequence of a monotone sequence is monotone.

If two subsequences of a given sequence converge to distinct limits, then

the sequence diverges.

Theorem
Let hxn i be a Cauchy sequence. Suppose it has a convergence subsequence
hxnk i that converges to `. Then lim xn = `.
n→∞
G. S. Raju (IIT Ropar) SEQUENCES 11 / 14
Recursive sequences
Motivation
Let us recall the Fibonacci sequence {1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, · · · }.

Definition
A sequence is said to be a recurrence sequence, the terms of the sequence
are generated by solving a recurrence equation. That is the nth term of the
sequence is depend upon previous term(s) of the sequence for n ≥ 2.

G. S. Raju (IIT Ropar) SEQUENCES 12 / 14

Examples
 
1 2
1 x1 = 2 and xn+1 = xn + for all n ≥ 1. Then {xn } is a
2 xn √
recursive sequence. Does it converge?. lim xn = 2.
n→∞
Slon: Suppose lim xn = `(> 0). Hence lim xn = `.
  n→∞  n+1→∞
 
1 2 1 2 1 2
xn + → `+ Hence ` = `+ . That is
2 xn 2 `√ 2 `
2`2 − ` − 2 = 0. Thus ` = 2.
yn + yn+1
2 Let a, b ∈ R. Define y1 = a, y2 = b and yn+2 = and
0 √ 2
yn+2 = yn · yn+1 . Then {yn } and {yn0 } are recursive sequences.
√ √
3 Let z1 = 2 and zn+1 = 2 + xn . Then {zn } is a recursive sequence.
Application : Square root of a positive number
√
Let a > 0. Let us construct a sequence hxn i that
 converges
 to a.
1 a
Suppose x1 = a. For any n ∈ N set xn+1 = xn + .
2 xn √
Show that it is increasing and bounded and converges to a
G. S. Raju (IIT Ropar) SEQUENCES 13 / 14
Divergence sequences
Motivation
h(−1)n i. hni. h−ni

Definition
A sequence hxn i is said to be divergent (non-convergent sequence) if is it
is a not a convergence sequence. That is there exists 0 > 0 such that
{n : xn ∈
/ (a − 0 , a + 0 )} is an infinite set for any a ∈ R.
G. S. Raju (IIT Ropar) SEQUENCES 14 / 14