0% found this document useful (0 votes)
65 views8 pages

2021-2022 Final Exam

The document is the instructions for a final exam consisting of 2 exercises. Exercise 1 has 4 questions regarding Taylor expansions and derivatives. Exercise 2 has 4 questions proving properties of the function g(x) = tan(x)/x, including finding its Taylor expansion, limit at 0, and determining if it is differentiable at 0 after extending it by continuity.

Uploaded by

Pavle Pakalović
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
65 views8 pages

2021-2022 Final Exam

The document is the instructions for a final exam consisting of 2 exercises. Exercise 1 has 4 questions regarding Taylor expansions and derivatives. Exercise 2 has 4 questions proving properties of the function g(x) = tan(x)/x, including finding its Taylor expansion, limit at 0, and determining if it is differentiable at 0 after extending it by continuity.

Uploaded by

Pavle Pakalović
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Analysis MAA102 - Bachelor

Final Exam.
Tuesday, January 25, 2022
Duration: 2 hours.

Foreword: The use of electronic devices (cell phones, computers, . . . ) is forbidden.


The lecture notes, documents which have been distributed and personal notes are not
allowed.
It is strongly advised to read carefully the statements of all the exercises. The writing
should be concise and clear, special attention should be drawn on the precise statement
of the results of the course which are used.
The different exercises are independent and, if you are not able to answer one question,
you can nevertheless answer the next questions of an exercise using the results of the
previous questions.

Final word: Exercise 5 (on page 4) is a ”Bonus exercise” and should only be considered
if you have finished all other exercises.

It is not necessary to do all the exercises to get a very good grade.

1
Exercise 1. (Special exercise for BX24). We define the function
2024
!
X (−1)n+1
h(x) := exp xn .
n
n=1

1.1 - For all n ⩾ 1, compute the n-th derivative of x 7→ ln(1 + x).

We set f (x) := ln(1 + x). We claim that


1
f (n) (x) = (−1)n+1 (n − 1)! .
(1 + x)n

Indeed, the formula holds for n = 1 since the derivative of f is given by


1 1
f ′ (x) = = (−1)2 0! .
x+1 (1 + x)1

Now, assuming that the formula is true for n, we get


−n
f (n+1) (x) = (−1)n+1 (n − 1)!
(1 + x)n+1
1
= (−1)n+2 n! ,
(1 + x)n+1

which proves that the formula holds for n+1. The principle of mathematical induction
implies that the formula holds for all n ⩾ 1.

1.2 - Prove that there exists a function x 7→ η(x) which tends to 0 as x tends to 0, such
that
2024
X (−1)n+1
ln(1 + x) = xn + x2024 η(x),
n
n=1

for all x > −1.

The function x 7→ ln(1 + x) is C ∞ and hence it admits a Taylor expansion at any


order at 0. Therefore, there exists a function x 7→ η(x) which tends to 0 as x tends
to 0, such that
2024
X f (n) (0) n
ln(1 + x) =: f (x) = f (0) + x + x2024 η(x)
n=1
n!
2024
X (−1)n+1 n
= x + x2024 η(x),
n=1
n

and this completes the proof.

1.3 - Find the Taylor expansion of the function h at order 2024, at x = 0.

2
Thanks to the previous question, we can write
2024
X (−1)n+1 n
x = ln(1 + x) − x2024 η(x).
n=1
n

Hence we get

h(x) = exp ln(1 + x) − x2024 η(x) = (1 + x) exp exp −x2024 η(x) .


 

Now, we use the fact that the exponential function admits a Taylor expansion at order
1 at 0. So we can write
exp z = 1 + z + z ψ(z),
where the function ψ tends to 0 as z tends to 0. Using this, we conclude that
h(x) = (1 + x) 1 − x2024 η(x) − x2024 η(x) ψ −x2024 η(x)


= 1 + x + x2024 γ(x),

where the function γ tends to 0 at 0.

Exercise 2. For all x ∈ − π2 , π2 − {0}, we define




Å ã1/x2
tan x
g(x) := .
x
2.1 - Prove that the function g is well defined, continuous and differentiable on (− π2 , π2 )−
{0}.

Observe that the function g can be written as


Å Å ãã
1 tan x
g(x) = exp ln .
x2 x

Now, the function x 7→ tan x π π



x is well defined, continuous and differentiable on − 2 , 2 −
{0} as the quotient of continuous and differentiable functions. Moreover, this func-
tan x

tion is positive. Therefore, the function x →
7 ln x is well defined, continuous
and differentiable on − π2 , π2 − {0} as the composition of continuous and differen-

tiable functions. It then follows that g is well defined, continuous and differentiable
on − π2 , π2 − {0} as the composition of continuous and differentiable functions.


2.2 - Prove that the Taylor expansion of the function x 7→ tan x at order 5 at x = 0 is
given by
x3 2x5
tan(x) = x + + + x5 ψ(x),
3 15
where x 7→ ψ(x) tends to 0, as x tends to 0.

3
We have
x3 x5
sin x = x − + + x5 η(x),
6 120
and
x2 x4
cos x = 1 − + + x4 θ(x),
2 24
where both x 7→ η(x) and x 7→ θ(x) tend to 0, as x tends to 0. Using the expansion
1
= 1 + z + z 2 + z 2 ξ(z),
1+z
where z 7→ ξ(z) tends to 0, as X tends to 0. Using this, we conclude with little work
that
x2 2x4
tan(x) = 1 + + + x4 ψ(x),
3 15
where x 7→ ψ(x) tends to 0, as x tends to 0.

2.3 - Prove that the function g has a limit as x tends to 0 and compute this limit.

Using the previous question together with the second order Taylor expansion of y 7→
ln(1 + y) at y = 0,
y2
ln(1 + y) = y − + y 2 ζ(y),
2
where y 7→ ζ(y) tends to 0, as y tends to 0, we get

x2 2x4 x4
Å ã
tan x
ln = + − + x4 ψ(x)
x 3 15 18
x2 7x4
= + + x4 ψ(x).
3 90
where x 7→ ψ(x) tends to 0, as x tends to 0. This implies that

ln tan x

x 1
lim 2
= ,
x→0,x̸=0 x 3
and, using the continuity of x 7→ exp x, we conclude that

lim g(x) = exp(1/3) = 3 e.
x→0,x̸=0

2.4 - We extend by continuity the function g at x = 0 i.e. we agree that, by definition

g(0) := lim g(x).


x→0, x̸=0

Is the function g (extended by continuity) differentiable at 0? If this is the case, what is


the value of the derivative of g at 0?

4
We have seen in the previous question that

x2 7x4
Å ã
tan x
ln = + + x4 ψ(x),
x 3 90
therefore,
tan x
Ç å
ln x
g(x) = exp
x2
1 7x2
Å ã
= exp + + x2 ψ(x)
3 90
Å 2 ã
7x
= e1/3 exp + x2 ψ(x) ,
90
where x 7→ ψ(x) tends to 0, as x tends to 0. Using the Taylor expansion of y 7→ exp y
at order 1 at y = 0, we conclude that

7x2
Å ã
g(x) = e1/3 1 + + x2 ζ(x) ,
90
where x 7→ ζ(x) tends to 0, as x tends to 0. In particular

g(x) − g(0)
Å ã
1/3 7x
=e + x ζ(x) ,
x−0 90
tends to 0 as x tends to 0. So, g is differentiable at 0 and g ′ (0) = 0.

Exercise 3. Let f : R → R be function which is continuous on R and differentiable on


R − {0}.

3.1 - In this question, we assume that

lim f ′ (x),
x→0, x̸=0

exists. Prove that f is differentiable at x = 0. Hint: Use the Mean Value Theorem
between 0 and x.

Assume that x > 0. The Mean Value Theorem implies that there exists cx ∈ (0, x),
such that
f (x) − f (0) = f ′ (cx ) (x − 0).
Since f ′ has a limit at 0, we conclude that

lim f ′ (cx ) = ℓ
x→0,x̸=0

and this implies that f is differentiable at 0.

5
3.2 - In this question, we assume that f is differentiable at x = 0. Is it true that

lim f ′ (x),
x→0, x̸=0

exists? Hint: Find a counterexample.

The function Å ã
1
f (x) = x2 sin
x
is differentiable on R but
Å ã Å ã
′ 1 1
f (x) = 2 x sin − cos ,
x x2

does not have a limit at 0.

Exercise 4. Given α > 0, we define the sequence (un )∞


n=1 by u1 > 0 and

n
un+1 := un ,
n+α
for all n ⩾ 1.

4.1 - Prove that the sequence (un )∞


n=1 converges.

The sequence is positive and decreasing. Therefore, it converges.

4.2 - We define
vn := ln (nα un )
Compute wn := vn+1 − vn in terms of n.

We have
wn := vn+1 − vn
Å ã Å ã
n+1 un+1
= α ln + ln
n un
Å ã
1  α
= α ln 1 + − ln 1 +
n n

4.3 - Prove that the sequence (n2 wn )∞


n=1 converges and compute its limit.

6
Using a Taylor expansion of x 7→ ln(1 + x) at order 2 at 0, we get

lim n2 wn − α(1 − α) = 0.

n→+∞

α (1−α)
Therefore, the sequence (n2 wn )∞
n=1 converges to 2 .

4.4 - Prove that the sequence (nα un )∞


n=1 converges.


X
Thanks to the previous question, we know that the infinite series wn converges.
n=1
Hence the sequence (vn )∞
n=1 converges and, since the exponential function is continu-
ous, we conclude that the sequence (nα un )∞
n=1 converges to some positive real number.


X
4.5 - Prove that the infinite series un converges if α > 1 and diverges when α ∈ (0, 1].
n=1


X 1
Since the infinite series α
converges when α > 1 and diverges when α < 1, we
n=1
n
can use the previous question together with the comparison test to conclude that the

X
infinite series un converges if α > 1 and diverges when α ∈ (0, 1].
n=1


X
4.6 - Prove that the infinite series (−1)n un converges for all α > 0.
n=1

The sequence (un )∞n=1 is decreasing and tends to 0. A result from the lecture implies

X
that the alternative series (−1)n un converges.
n=1

4.7 - Assume that α > 1 is fixed and that (an )∞n=1 is a real valued sequence such that
an > 0 for all n ⩾ 1 and
an+1 α
⩽1− ,
an n

X
for all n ⩾ 2. What can you say about the infinite series an ?
n=1

We choose γ ∈ (1, α) and compute


Å ã
1  α
ln ((n + 1)γ an+1 ) − ln (nγ an ) ⩽ γ ln 1 + + ln 1 − ⩽ 0,
n n
for n large enough, say n ⩾ N . This implies that the sequence (nγ an )∞
n=N is decreas-
ing and hence Å ãγ
N
an ⩽ aN ,
n

7

X
and the comparison test implies that the infinite series an converges.
n=1

4.8 - Assume that α ∈ (0, 1) is fixed and that (an )∞


n=1 is a real valued sequence such that
an > 0 for all n ⩾ 1 and
an+1 α
⩾1− .
an n

X
What can you say about the infinite series an ?
n=1

We choose γ ∈ (α, 1) and compute


Å ã
γ γ 1  α
ln ((n + 1) an+1 ) − ln (n an ) ⩾ γ ln 1 + + ln 1 − ⩽ 0,
n n

for n large enough, say n ⩾ N . This implies that the sequence (nγ an )∞
n=N is increas-
ing and hence Å ãγ
N
an ⩾ aN ,
n

X
and the comparison test implies that the infinite series an diverges.
n=1

Exercise 5. Bonus exercise. To be considered only if you have finished all


other exercises. Let f : R → R is a C n+1 function where n ∈ N. We assume that

f (1) = f (0) = f ′ (0) = . . . = f (n) (0) = 0.

Prove that there exists x ∈ (0, 1] such that f (n+1) (x) = 0.

We apply Taylor-Lagrange Theorem between 0 and 1. This leads to


n
X f (k) (0) f (n+1) (c)
f (1) = f (0) + +
k! (n + 1)!
k=1

where c ∈ (0, 1). Under the assumptions, this reduces to

f (n+1) (c)
0= ,
(n + 1)!

and hence f (n+1) (c) = 0. This completes the proof.

You might also like