CHAPTER 2 : SEQUENCES OF REAL NUMBERS

S. ARICHE

UNIVERSITY SAAD DAHLAB OF BLIDA

Contents

1. Bounded sequences 1
2. The Convergent Sequences 3
3. Limit Theorems 5
4. Monotonous Sequences 6
5. Particular Sequence 7
6. Extracted Sequences (Subsequences) 8
7. Adjacent Sequences 9
8. The Cauchy Sequences 10
9. Recurring Sequences 11
9.1. The case where f is increasing 11
9.2. The case where f is decreasing 12

1. Bounded sequences

Definition 1.1. A sequence is a real-valued function f whose domain is the set of positive integers N.
The numbers f (1), f (2), · · · are called the terms of the sequence.
Let U : N → R be a sequence. Then the values of U are U (1), U (2), U (3), · · · , U (n), · · · . It is customary
to write Un instead of U (n) in this case.

Specifying a sequence: There are several ways to specify a sequence:

(1) By giving the function. For example:
1
(a) Un = n. This is the sequence
1 1 1 1
{1, , , , · · · , , · · · }.
2 3 4 n
n−1
(b) Un = n . This is the sequence {0, 12 , 23 , 34 , · · · , n−1
n , · · · }.
(c) Un = (−1)n n2 . This is the sequence {−1, 4, −9, 16, · · · , (−1)n n2 , · · · }.
(2) By giving the first few terms to establish a pattern, leaving it to you to find the function. This
is risky (it might not be easy to recognize the pattern and/or you can be misled).
(a) {Un } = {0, 1, 0, 1, 0, 1, · · · }. The pattern here is obvious; can you devise the function? It is
(
1 − (−1)n 0, n odd
Un = where Un =
2 1, n even.
n2 +1
(b) {Un } = {2, 25 , 10 17 26
3 , 4 , 5 , · · · }, so Un = n .
1
2 S. ARICHE

(c) {Un } = {2, 4, 8, 16, 32, · · · , }. What is U6 ? What is the function? While you might say 64
π
and Un = 2n , the function I have in mind gives U6 = 6:
 
n π 64
Un = 2 + (n − 1)(n − 2)(n − 3)(n − 4)(n − 5) − .
720 120
(3) By a (
recursion formula. For example:
U1 = 1,
(a) 1
Un+1 = n+1 Un
 1 1 1
The
( first five terms are 1, 2 , 6 , 24 , · · · .
U1 = 1,
(b)
Un+1 = 12 (Un + 1).
The first 5 terms are {1, 1, 1, 1, 1, , · · · }. (Un )n is a constant sequence.

Definition 1.2. Let (Un )n∈N be a sequence of real numbers.

(1) If a real number M satisfies
Un ≤ M for all n ∈ N,
then M is called an upper bound of (Un )n and the sequence (Un )n is said to be bounded above.
(2) If a real number m satisfies
m ≤ Un for all n ∈ N,
then m is called an lower bound of (Un )n and the sequence (Un )n is said to be bounded below.
(3) The sequence (Un )n is said to be bounded if it is bounded above and bounded below.

Proposition 1.3. Let (Un )n∈N be a sequence of real numbers. The sequence (Un )n is bounded if and
only if
∃k > 0, |Un | ≤ k, ∀n ∈ N.

Definition 1.4. A sequence (Un )n∈N of real numbers is said to be :

• increasing if Un ≤ Un+1 for all n ∈ N,
• strictly increasing if Un < Un+1 for all n ∈ N,
• decreasing if Un ≥ Un+1 for all n ∈ N,
• strictly decreasing if Un > Un+1 for all n ∈ N,
A sequence with either of these four properties is said to be monotonic.
1
Example 1.5. 1) Un = n+1 , n ∈ N. We have
−1
Un+1 − Un = < 0, then (Un )n is decreasing.
(n + 1)(n + 2)
2) Un = 2n , n ∈ N.

Un+1 − Un = 2n > 0, ∀n ∈ N, then (Un )n is increasing.

3) Un = (−1)n , n ∈ N: (Un )n is not monotonic.

Definition 1.6. A sequence (Un )n is constant if it is increasing and decreasing in the same time.
√
Example 1.7. Un = 2, n ∈ N.

Definition 1.8. A sequence (Un )n is said to be stationary if

∃n0 ∈ N, ∀n ≥ n0 : Un+1 = Un .
 CHAPTER 2 : SEQUENCES OF REAL NUMBERS 3

(
√1 , n ≤ 3,
Example 1.9. Un = √n+1
2, n > 3.

Remark 1.10. If the geeneral term of a sequence (Un )n is positif, then

Un+1
• (Un )n is increasing if Un ≥ 1, for all n ∈ N,
Un+1
• decreasing if Un ≤ 1, for all n ∈ N.

Example 1.11. Let Un = 2n > 0, ∀n ∈ N.

Un+1
We have Un = 2 ≥ 1, then (Un )n is increasing.

2. The Convergent Sequences

Definition 2.1. A sequence (Un )n of real numbers converges to the real number l if for each  > 0 there
exists a number N () such that

∀n ∈ N, n ≥ N () ⇒ |Un − l| < .

If (Un )n converges to l we will write lim Un = l. The number l is called the limit of the sequence (Un )n .
n→∞
A sequence that does not converge to a real number is said to diverge.
n2 −n−1
Example 2.2. Prove that lim 2 = 12 .
n→∞ 2n −1
Solution. Let  > 0 be given. We need to solve
n2 − n − 1 1
− < .
2n2 − 1 2
First simplify:
n2 − n − 1 1 2(n2 − n − 1) − (2n2 − 1) −2n − 1 2n + 1
− = = = 2 .
2n2 − 1 2 2(2n2 − 1) 2(2n2 − 1) 4n − 2
Therefore,
2n + 1 2n + n 3n 3
≤ 2 = 2 = ,
4n2 − 2 4n − 2n2 2n 2n
and we have
n2 − n − 1 1 3
− ≤ , valid for n ∈ N.
2n2 − 1 2 2n
3 3
Solving for n is now easy: 2n < . The solution is n > 2 .
3
Thus, with N () = E( 2 ) + 1 , the implication
n2 − n − 1 1
n ≥ N () ⇒ − < .
2n2 − 1 2
Theorem 2.3. If (Un )n converge to l1 and (Un )n converge to l2 , then l1 = l2 . That is, the limit of a
convergent sequence is unique.

Proof : Let suppose the opposite, i.e. the sequence (Un )n converge to two different limits l1 and l2 , then

∀ > 0, ∃n1 ∈ N, ∀n ∈ N, n ≥ n1 : |Un − l1 | < ,

and
∀ > 0, ∃n2 ∈ N, ∀n ∈ N, n ≥ n2 : |Un − l2 | < .
|l1 −l2 |
Let  = 2 > 0 and let n0 = max(n1 , n2 ) sush that n ≥ n0 , then
|l1 − l2 | |l1 − l2 |
|Un − l1 | < and |Un − l2 | < .
2 2
4 S. ARICHE

Therefore,
|l1 − l2 | |l1 − l2 |
|l1 − l2 | = |l1 − Un + Un − l2 | ≤ |l1 − Un | + |Un − l2 | = |Un − l1 | + |Un − l2 | < + = |l1 − l2 |,
2 2
a contradiction. Therefore, l1 = l2 .

Remark 2.4. The sequence Un is divergente if:

• lim Un = ∞,
n→∞
• (Un )n has two different limits.

Example 2.5. (1) Un = −n, n ∈ N. We have lim Un = −∞ so (Un )n is divergent.

n→∞
n
(2) Vn = (−1) , n ∈ N. We have

V2n → 1 and V2n+1 → −1,

thus (Vn )n is divergent.

Theorem 2.6. If (Un )n converges, then (Un )n is bounded.

Proof : Suppose that (Un )n converges to l. Then for all  > 0, there exists a positive integer n0 such
that
∀n ≥ n0 : |Un − l| < .

which is equivalent of
∀n ≥ n0 : Un ∈]l − , l + [.

Therefor, it follows that l −  < Un < l +  and then (Un )n is bounded for all n ≥ n0 .
On the other hand, (Un )n is bounded for all n = 0, · · · , n0 − 1. Thus if we set

M = max{U0 , U1 , U2 , · · · , Un0 −1 , l + } and m = min{U0 , U1 , U2 , · · · , Un0 −1 , l − },

we deduce that
m ≤ Un ≤ M for all n ∈ N.

Remark 2.7. The inverse of the theorem is false:

(Un )n is bounded ; (Un )n is convergent.

Example 2.8. Un = (−1)n , (Un )n is bounded but is not convergent.

Theorem 2.9. Arithmetic operations on convergent sequences

Let (Un )n , (Vn )n , (Wn )n , be given sequences. Let K and L be real numbers. Assume that

lim Un = K and lim Vn = L,

n→∞ n→∞

Then the following statements hold:

(A) If Wn = Un + Vn , n ∈ N, then lim Wn = K + L.
n→∞
(B) If Wn = Un Vn , n ∈ N, then lim Wn = KL.
n→∞
Un K
(C) If L 6= 0 and Wn = Vn , n ∈ N, then lim Wn = L.
n→∞
 CHAPTER 2 : SEQUENCES OF REAL NUMBERS 5

3. Limit Theorems

Proposition 3.1. Let (Un )n and (Vn )n be given sequences such that

lim Un = 0 and the sequence (Vn )n is bounded,

n→∞

then lim Un Vn = 0.
n→∞

Proof : We have lim Un = 0 then

n→∞

∀ > 0, ∃n0 ∈ N, ∀n ≥ n0 : |Un | < ,

and as (Vn )n is bounded, we deduce that

∃k > 0; |vn | ≤ k, ∀n ∈ N.

Soit  > 0, ∃n0 ∈ N, ∀n ≥ n0 : |Un Vn | = |Un ||Vn | < k, and therefore

∃n1 = n0 , ∀n ≥ n1 , |Un VN | < 1 = k,

we conclude: lim Un Vn = 0.
n→∞

n sin n
Example 3.2. 1) Un = n2 +1 , n ∈ N. We have
n sin n n
= 2 sin n,
n2 + 1 n +1
Let
n
Wn = and Vn = sin n.
n2 + 1
We have (Vn )n is bounded and (Wn )n converges to 0, then lim Un = 0.
n→∞
2) Un = (−1)n e−n , n ∈ N. Let Wn = e−n and Vn = (−1)n . We have (Vn )n is bounded and (Wn )n
converges to 0, then lim Un = 0.
n→∞

Theorem 3.3. Let (Un )n and (Vn )n be given sequences. Let K and L be real numbers. Assume that
(1) lim Un = K.
n→∞
(2) lim Vn = L.
n→∞
(3) There exists a natural number n0 such that

Un ≤ Vn , for all n ≥ n0 .

Then k ≤ L.

Corollary 3.4. Let (Un )n be given sequence that converges such that lim Un = K. Then
n→∞
(1) If Un ≥ 0 then K ≥ 0.
(2) If Un ≤ 0 then K ≤ 0.

Theorem 3.5. Let (Un )n , (Vn )n and (Wn )n be given sequences. Assume the following
(1) The sequence (Un )n converges to the limit L.
(2) The sequence (Vn )n converges to the limit L.
(3) There exists a natural number n0 such that

Un ≤ Wn ≤ Vn for all n > n0 .

Then the sequence (Wn )n converges to the limit L.

6 S. ARICHE

cos n
Example 3.6. 1) Un = n2 +1 . We have
−1 cos n 1
−1 ≤ cos n ≤ 1 ⇒ ≤ 2 ≤ 2 ,
n2 +1 n +1 n +1
and then lim ncos n
2 +1 = 0.
n→∞
n
√ 1
P
2) Un = n2 +k
. For all n such that 1 ≤ k ≤ n, we have
k=1
p p p
n2 + 1 ≤ n2 + k ≤ n2 + n,

and then
1 1 1
√ ≤√ ≤√ ,
n2 +n n2 +k n2 +1
Summing up we obtain
n n n
X 1 X 1 X 1
√ ≤ √ ≤ √ ,
k=1
n2 +n k=1
n2 +k k=1
n2 +1
which is equivalent to
n n
√ ≤ Un ≤ √ ,
n2+n 2
n +1
and then lim Un = 1.
n→∞

Proposition 3.7. Suppose that (Un )n and (Vn )n are sequences such that Un ≤ Vn for all n ≥ 1.
(1) If Un → +∞, then Vn → +∞.
(2) If Vn → −∞, then Un → −∞.

4. Monotonous Sequences

The following theorem gives powerful tools for establishing convergence of a sequence.

Theorem 4.1. A seaquence of real numbers Un )n , that is increasing and bounded above, is convergente,
and we have
U0 = inf{Un , n ∈ N},
lim Un = sup{Un , n ∈ N}.
n→∞

Proof : Since the sequence (Un )n is bounded above, then it admits a supremum noted l. Using the
caraterisation of the supremum, we obtain

∀ > 0, ∃n0 ∈ N, l −  < Un0 ≤ l.

On the other hand, the sequence is increasing and bounded above by l, then

∀n ≥ n0 , Un0 ≤ Un ≤ l,

we conclude that
∀ > 0, ∃n0 ∈ N, ∀n ≥ n0 : l −  < Un ≤ l + ,
which implies that (Un )n converges to l.

Theorem 4.2. A seaquence of real numbers Un )n , that is decreasing and bounded below, is convergente,
and we have
U0 = sup{Un , n ∈ N},
lim Un = inf{Un , n ∈ N}.
n→∞
 CHAPTER 2 : SEQUENCES OF REAL NUMBERS 7

Remark 4.3. If the sequence (Un )n is increasing non bounded above, or decreasing non bounded below,
then (Un )n is divergent. In fact, suppose that (Un )n is increasing non bounded above, then

∀A ∈ R, ∃n0 ∈ N, Un0 > A,

since the sequence is increasing, we obtain ∀n ≥ n0 : A < Un0 < Un which means that (Un )n is divergent.
1
Example 4.4. Let the sequence Un = 1 + n2 .

(1) Prove that (Un )n is monotone.

(2) Deduce that (Un )n is convergente.
(3) Let A = {Un , n ∈ N}. Calculate sup(A), inf(A), min(A) and max(A) if they exist.

Solution: 1) Let prove the monotonus.

−2n − 1
Un+1 − Un = ≤ 0, ∀n ∈ N∗ ,
n2 (n + 1)2
then (Un )n is decreasing and it is monotone.
2) Let deduce that (Un )n is convergent. We have for all n ∈ N∗ , n ≥ 1 then n2 ≥ 1 and 0 ≤ 1
n2 ≤ 1, we
deduce that
1
(4.1) 1≤1+ ≤ 2.
n2
(Un )n is decreasing and is bounded below by 1, which implies that it is convergent, and it converges
exactly to the infimum of the set {Un , n ∈ N}.
3) The set A is bounded due to (4.1), so it admits a supremum and a infimum. Since the sequence (Un )n
is decreasing bounded below, applying theorem 4.2, we deduce that it converges to the infimum inf(A)
and we have

inf(A) = inf{Un , n ∈ N} = lim Un = 1 and sup(A) = sup{Un , n ∈ N} = U1 = 2.

n→∞

Moreover 1 ∈
/ A, then min(A) does not exist, and U1 = 2 ∈ A then max(A) = 2.

Remark 4.5. Let A = {Un , n ∈ N}.

(1) If (Un )n is an increasing sequence which is bounded above, then max(A) exists if and only if
(Un )n is stationary (i.e. ∃n0 ∈ N, ∀n ≥ n0 : Un+1 = Un )
(2) If (Un )n is an decreasing sequence which is bounded below, then min(A) exists if and only if (Un )n
is stationary

5. Particular Sequence

Let (an )n be a given sequence. Then the expression

a1 + a2 + a3 + · · · + aN + · · ·

is called an infinite series. We often abbreviate it by writing

∞
X
a1 + a2 + a3 + · · · + aN + · · · = an .
n=1

For each natural number n we calculate the (finite) sum of the first n terms of the series

Sn = a1 + a2 + a3 + · · · + an .
8 S. ARICHE

∞
P
We call Sn a partial sum of the infinite series an . (Notice that (Sn )n is a new sequence.)
n=1
If the sequence (Sn )n converges and if
lim Sn = S,
n→∞
∞
P
then the infinite series is called convergent and we write an = S. The S is called the sum of the serie.
n=1
n
Let take for example the sequence Un = a , a ∈ R which is a geometric sequence. We know that:
• If a = 1, Un = 1 : constant sequence then convergent.
• If a = −1, Un = (−1)n : divergent sequence.
• If |a| < 1: (Un )n is convergent and

lim Un = lim an = 0.
n→∞ n→∞

• If |a| > 1: (Un )n is divergent and lim Un = ∞.

n→∞

Example 5.1. lim ( 31 )n = 0, lim (− 14 )n = 0, lim 2n = +∞.

n→∞ n→∞ n→∞

Let define now the sequence (Sn )n by

n
X
Sn = ak = 1 + a + a2 + · · · + an .
k=0

(1) If a = 1, Sn = n + 1 : then lim Sn = +∞.

n→∞
(2) If a 6= 1 then
1 − an+1
Sn = a0 ,
1−a
• If |a| < 1: (Un )n is convergent and
1
lim Sn = ∈ R.
n→∞ 1−a
• If |a| > 1: lim Sn = ∞.
n→∞
1−(−1)n+1
• If a = −1, Un = 2 : has not a limit.

Example 5.2. Calculate

∞
1 1 1 1 X 1
+ + + ··· + n + ··· = .
2 4 8 2 n=1
2n
1 n
1 1 1−( 2 )
We have a = 2 ∈] − 1, 1[ then Sn = 2 1− 21 = 1 − ( 12 )n , then (Sn )n converges and

S = lim Sn = 1.
n→∞

6. Extracted Sequences (Subsequences)

Definition 6.1. Given a sequence (Un )n . Let ϕ : N → N a function strictly increased, then the sequence
(Vn )n defined by Vn = Uϕn , ∀n ∈ N, is said to be an extracted sequence or a subsequence.

Example 6.2. 1) Let the sequence Un = (−1)n . We have U2n = 1 and U2n+1 = −1 are two sub sequence
of (Un )n .
√
2
2) Let vn = cos(n π4 ), n ∈ N. We have V8n = cos(2nπ) = 1 and U8n+1 = cos(2nπ + π4 ) = 2 are two
subsequence of (Vn )n .

Theorem 6.3. If (Un )n converges to l, then every subsequence (Uϕn )n of (Un )n also converges to l.
 CHAPTER 2 : SEQUENCES OF REAL NUMBERS 9

Corollary 6.4. If (Un )n has a subsequence (Vn )n that converges to l1 and a subsequence (Wn )n that
converges to l2 with l1 6= l2 , then (Un )n does not converge.

Example 6.5. 1) Let Un = cos(n π4 ), n ∈ N. We have

√
2
U8n → 1 and U8n+1 →
2
then (Un )n diverges.
n
2) Let the sequence Un = n(−1) . We have U2n = 2n and lim U2n = +∞. Since the subsequence (U2n )n
n→∞
of (Un )n diverges then (Un )n is also divergent.

7. Adjacent Sequences

Definition 7.1. Two real sequences (Un )n and (Vn )n are called adjacent if (Un )n is increasing, (Vn )n is
decreasing and lim (Vn − Un ) = 0.
n→∞
n
1 1
P
Example 7.2. Let Un = k! , V n = Un + n!n .
k=0
• Monotonic of (Un )n :
n+1 n
X 1 X 1 1
Un+1 − Un = − = > 0,
k! k! (n + 1)!
k=0 k=0
then (Un )n is increasing.
• Monotonic of (Vn )n :
1 1 −1
Vn+1 − Vn = Un+1 + − Un − = < 0,
(n + 1)!(n + 1) n!n (n + 1)!(n + 1)n
then (Vn )n is decreasing.
1
• lim (Vn − Un ) = lim = 0.
n→∞ n→∞ n!n
All the conditions are satisfied, we deduce that (Un )n and (Vn )n are adjacent.

Proposition 7.3. Let (Un )n and (Vn )n be two adjacent sequences such that (Un )n is increasing, (Vn )n
is decreasing then Vn − Un ≥ 0, ∀n ∈ N.

Proposition 7.4. If (Un )n and (Vn )n are two adjacent sequences, then they are convergent and have the
same limit.

Proof : Let (Un )n and (Vn )n be two adjacent sequences, then (Un )n is increasing, (Vn )n is decreasing
and lim (Vn − Un ) = 0. On the other hand, Proposition 7.4 implies that
n→∞

U0 ≤ U1 ≤ · · · ≤ Un−1 ≤ Un ≤ Vn ≤ Vn−1 ≤ · · · ≤ V0 ,

which means that

∀n ∈ N : Un ≤ V0 and Vn ≥ U0 .
In summary, we conclude that, (Un )n is increasing bounded above by V0 then converges, and (Vn )n is
decreasing bounded below by U0 then converges also.
In the other hand, we can write Vn = (Vn − Un ) + Un and therfore

lim Vn = lim (Vn − Un ) + Un = lim Un .

n→∞ n→∞ n→∞

Proposition 7.5. If (U2n )n and (U2n+1 )n are two subsequences of (Un )n that have the same limit, then
(Un )n is convergent and converges also to the same limit.
10 S. ARICHE

8. The Cauchy Sequences

Definition 8.1. A sequence (Un )n is a Cauchy sequence if

∀ > 0, ∃n0 ∈ N, ∀p, q ∈ N : p > q ≥ n0 ⇒ |Up − Uq | < .

Definition 8.2. A sequence (Un )n is a Cauchy sequence if

∀ > 0, ∃n0 ∈ N, ∀n ≥ n0 , ∀p > 1 : |Un+p − Un | < .

In other terms,

lim |Un+p − Un | = 0, ∀p > 1.

n→∞

n
sin k
P
Example 8.3. Prove that the sequence defined by Un = 2k
, is a Cauchy sequence.
k=1
We have
n+p n n+p
X sin k X sin k X sin k
|Un+p − Un | = | k
− k
| = | |
2 2 2k
k=1 k=1 k=n+1
n+p n+p
X sin k X 1
≤ | |≤ .
2k 2k
k=n+1 k=n+1

On the other hand,

n+p
X 1 1 1 1 1
= ( + 2 + ··· + p)
2k n
2 2 2 2
k=n+1

1 1 1 − ( 21 )p 1 1
= ( ; ) = n (1 − p ).
2n 2 1 − 21 2 2

Then,
1 1
|Un+p − Un | ≤ (1 − p ).
2n 2
We conclude finally that lim |Un+p − Un | = 0, ∀p > 1.
n→∞

Theorem 8.4. Every convergent sequence is a Cauchy sequence.

Proof : Suppose that Un converges to l. Let  > 0. There exists a positive integer n0 such that
|Un − l| < /2 for alln > n0 . Let p > q > n0 , then

|Up − Uq | = |Up − l + l − Uq | ≤ |Up − l| + |Uq − l| < /2 + /2.

Then (Un )n is a Cauchy sequence.

Theorem 8.5. Every Cauchy sequence is bounded.

Theorem 8.6. A sequence (Un )n is convergent if and only if it is a Cauchy sequence.

Remark 8.7. (Un )n is not a Cauchy sequence ⇒ (Un )n is divergent.

 CHAPTER 2 : SEQUENCES OF REAL NUMBERS 11

9. Recurring Sequences

Definition 9.1. A recurring sequence is an equation that recursively defines a sequence where the next
term is a function f of the previous terms . It gives you a connection between two consecutive terms.
These two terms are usually Un+1 and Un . However they could be given as Un and Un−1 . The starting
value U1 , would have to be provided. Note that the starting value can also be U0 , and we write
(
U0 be given,
Un+1 = f (Un ), ∀n ≥ 1.
This sequence is well defined if and only if ∀n ∈ N, Un ∈ D(f ).

Example 9.2. (
U0 = 1
(Un )n : √
Un+1 = 1 + Un , ∀n ≥ 1.
√
Here f : x 7→ f (x) = 1 + x and D(f ) = [−1, +∞[.

Definition 9.3. Let f : A → A be an application. If a sequence (Un )n defined by the recurrence formula
Un+1 = f (Un ), converges to a point l ∈ R and if f is continous at this point l, then l verifies

l = f (l) if l is inreasing,

and we say that l is a fixed point.

Remark 9.4.
1) In all this section, we suppose that f is continous.
2) Usually, to study the nature of a recurring sequence we study the monotony of the function f .

9.1. The case where f is increasing.

Proposition 9.5. Let f : A → A be an increasing application and (Un )n be a sequence defined by the
recurrence formula (
U0 ∈ Abe given,
Un+1 = f (Un ), ∀n ≥ 1.
Then, we have Un ∈ A, ∀n ∈ N and (Un )n is monotonous. More precisely:
(1) If U0 ≥ U1 , then (Un )n is decreasing.
(2) If U0 ≤ U1 , then (Un )n is increasing.
(
U0 = 0
Example 9.6. Let Un :
Un+1 = 7U n +4
3Un +3 , ∀n ≥ 1.
(1) Prove that 0 ≤ Un ≤ 2, ∀n ∈ N.
(2) Prove that (Un )n is monotonous.
(3) Let A = {Un , n ∈ N}. Determine inf(A), sup(A).

Solution: 1) By induction, we will prove that 0 ≤ Un ≤ 2, ∀n ∈ N.

For n = 0, P (0) is true. Suppose that 0 ≤ Un ≤ 2, ∀n ≥ 1 and prove that 0 ≤ Un+1 ≤ 2, ∀n ≥ 1.
7Un +4
We have Un+1 = 3U n +3
≥ 0, ∀n ≥ 1 and
7Un + 4 Un − 2
Un+1 − 2 = −2= ≤ 0.
3Un + 3 3Un + 3
Therfore 0 ≤ Un ≤ 2, ∀n ∈ N.
12 S. ARICHE

7x+4
2) We have Un+1 = f (Un ), then f : x 7→ 3x+3 , and D(f ) = R − {−1}.
9
f 0 (x) = > 0, then f is increasing.
(3x + 3)2
Since U0 = 0 < U1 = 34 , we obtain that (Un )n is increasing.
3) From (Q1), (Un )n is bounded then A is bounded also and we deduce that there exists inf(A) and
sup(A).
Since (Un )n is bounded above and increasing, then it is convergent and converges to a point l that verifies
7l + 4
l = f (l) =
3l + 3
and then l = 2 (the second solution of the equation is refused l = − 32 ). We concluse

sup(A) = lim Un = l = 2.
n→∞

Moreover U0 = 0 ∈ A, then inf(A) = 0.

9.2. The case where f is decreasing. In this case, (Un )n is not monotonous, and we have to study
the two subsequece (U2n )n and (U2n+1 )n .

Proposition 9.7. If f is decreasing, then (U2n )n and (U2n+1 )n are monotonous with opposite monotony,
i.e.
• If U0 < U2 then (U2n )n is increasing and (U2n+1 )n is decreaing.
• If U0 > U2 then (U2n )n is decreasing and (U2n+1 )n is increaing.

Proposition 9.8. A sequence (Un )n converges if and only if (U2n )n and (U2n+1 )n are adjacents.