Problems and Solutions

in
Real and Complex Analysis,
Integration,
Functional Equations
and
Inequalities

by
Willi-Hans Steeb
International School for Scientific Computing
at
University of Johannesburg, South Africa
Preface

The purpose of this book is to supply a collection of problems in analysis.

Please submit your solution to one of th email addresses below.

e-mail addresses of the author:

[email protected]
[email protected]

Home page of the author:

http://issc.uj.ac.za
Prescribed books for problems.

1) Problems and Solutions in Theoretical and Mathematical Physics, Third

Edition, Volume I: Introductory Level

by

Willi-Hans Steeb
World Scientific Publishing, Singapore 2009
ISBN- 13 978 981 4282 14 7
http://www.worldscibooks.com/physics/7416.html

2) Problems and Solutions in Theoretical and Mathematical Physics, Third

Edition, Volume II: Advanced Level

by

Willi-Hans Steeb
World Scientific Publishing, Singapore 2009
ISBN-13 978 981 4282 16 1
http://www.worldscibooks.com/physics/7416.html

updated: January 16, 2018

v
vi
Contents

Preface v

Notation x

1 Sums and Products 1

1.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 12

2 Maps 26
2.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 35

3 Functions 41
3.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 60

4 Polynomial 73
4.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 73
4.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 86

5 Equations 91
5.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 91
5.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 98

6 Normed Spaces 102

6.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 103
6.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 106

7 Complex Numbers and Complex Functions 107

7.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 107
7.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 113

vii
8 Integration 118
8.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 118
8.2 Programming Problems . . . . . . . . . . . . . . . . . . . . 144
8.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 145

9 Functional Equations 152

9.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 152
9.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 156

10 Inequalities 157
10.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 157
10.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 169

Bibliography 172

Index 177

viii
x

Notation
:= is defined as
∈ belongs to (a set)
∈
/ does not belong to (a set)
∩ intersection of sets
∪ union of sets
∅ empty set
N set of natural numbers
Z set of integers
Q set of rational numbers
R set of real numbers
R+ set of nonnegative real numbers
C set of complex numbers
Rn n-dimensional Euclidean space
space of column vectors with n real components
Cn n-dimensional complex linear space
space of column vectors with n complex components
H Hilbert
√ space
i −1
<z real part of the complex number z
=z imaginary part of the complex number z
|z| modulus of complex number z
|x + iy| = (x2 + y 2 )1/2 , x, y ∈ R
T ⊂S subset T of set S
S∩T the intersection of the sets S and T
S∪T the union of the sets S and T
f (S) image of set S under mapping f
f ◦g composition of two mappings (f ◦ g)(x) = f (g(x))
v column vector in Cn
vT transpose of v (row vector)
0 zero (column) vector
k.k norm
x · y ≡ x∗ y scalar product (inner product) in Cn
x×y vector product in R3
A, B, C m × n matrices
det(A) determinant of a square matrix A
tr(A) trace of a square matrix A
rank(A) rank of matrix A
AT transpose of matrix A
 xi

A conjugate of matrix A
A∗ conjugate transpose of matrix A
A† conjugate transpose of matrix A
(notation used in physics)
A−1 inverse of square matrix A (if it exists)
In n × n unit matrix
I unit operator
0n n × n zero matrix
AB matrix product of m × n matrix A
and n × p matrix B
A•B Hadamard product (entry-wise product)
of m × n matrices A and B
[A, B] := AB − BA commutator for square matrices A and B
[A, B]+ := AB + BA anticommutator for square matrices A and B
A⊗B Kronecker product of matrices A and B
A⊕B Direct sum of matrices A and B
δjk Kronecker delta with δjk = 1 for j = k
and δjk = 0 for j 6= k
λ eigenvalue
 real parameter
t time variable
Ĥ Hamilton operator
Chapter 1

Sums and Products

1.1 Solved Problems

Problem 1. The harmonic series can be approximated by
n
X 1 1
≈ 0.5772 + ln(n) + .
j=1
j 2n

Calculate the left and rigt-hand side for n = 1 and n = 10.

Problem 2. The Bernoulli numbers B0 , B1 , B2 , . . . are defined by the

power series expansion
∞
x X Bj B1 B2 2
x
= xj ≡ B0 + x+ x + ···.
e − 1 j=0 j! 1! 2!

One finds B0 = 1, B1 = −1/2, B2 = 1/6, B3 = 0, B4 = −1/30. One has

Bj = 0 if j ≥ 3 and j odd. The Bernoulli numbers are utilized in the Euler
summation formula
n Z n n
X 1 X B2k (2k−1)
f (j) = f (t)dt+ (f (n)+f (1))+ (f (n)−f (2k−1) (1))+Rm (n)
j=1 1 2 (2k)!
k=1

where Z m
4
|Rm (n)| ≤ |f (2m) (t)|dt.
(2π)2m 1

1
2 Problems and Solutions

Calculate the sum

n
X
j2
j=1

using f (t) = t2 .

Problem 3. Let z ∈ C and |z| < 1. Calculate the sum

∞  
X 2s + n − 1
(1 − |z|2 )2s n |z|2n .
n=0
n

Problem 4. Let x ∈ R and r ∈ N with r ≥ 1. Find the sum

r−1
X
exp(−2πikx/r).
k=0

Problem 5. Let a1 , a2 , . . . , an ∈ [0, 1]. Show that there exists a real num-
ber x in the unit interval such that the average of the (unsigned) distances
from x to the aj ’s is 1/2.

Problem 6. Let f : Rm → R be a differentiable function. Show that

there exist m differentiable functions g1 , g2 , . . . , gm defined on Rm with
the properties
m
X
f (x) = f (0) + xj gj (x)
j=1

and
∂f
gj (0) = (0)
∂xj

where x = (x1 , x2 , . . . , xm ).

Problem 7. Let n be a positive integer and f (j) = j(j − 1)(j −P 2) with

n
j = 1, 2, . . . , n. Let aj := f (j + 1) − f (j). By calculating the sum j=1 aj
show that
n
X n(n + 1)(2n + 1)
j2 = .
j=1
6
 Sums and Products 3

Problem 8. Show that

 2
N N N  
X 1  −
X 1 X 1 1 1
 ≡2 − .
j=1
u − vj j=1
(u − vj )2 i<j
vi − vj u − v i u − vj
j=2
i=1

This identity plays a role in the Bethe ansatz.

Problem 9. Show that the series

∞
X sin(3j θ)
f (θ) =
j=0
2j

is convergent. Is the series df /dθ convergent?

Problem 10. Find the radius of convergence of the power series

∞  
X j+k j
z , k > 0.
j=0
j

Problem 11. Find the radius of convergence of the power series

∞
X j! mj
j
z , m = 1, 2, . . . .
j=1
j

Problem 12. Let (s0 , s1 , ..., sn−1 )T ∈ Rn , where n = 2k . This vector in

Rn can be associated with a piecewise constant function f defined on [0, 1)
k
2X −1
f (x) = sj Θ[j2−k ,(j+1)2−k ) (x)
j=0

where Θ[j2−k ,(j+1)2−k ) (x) is the step function

1 x ∈ [j2−k , (j + 1)2−k )

Θ[j2−k ,(j+1)2−k ) (x) :=
0x ∈/ [j2−k , (j + 1)2−k )

with the support [j2−k , (j + 1)2−k ). Let xj+1 = 4xj (1 − xj ) with j =

0, 1, 2, . . . and x0 = 1/3. Then
1 8 32 6272
x0 = , x1 = , x2 = , x3 = .
3 9 81 6561
4 Problems and Solutions

Find f (x) for this data set and then calculate

Z 1
f (x)dx.
0

Problem 13. Let a1 , a2 , . . . , an be a finite sequence of numbers. Its

Cesáro sum is defined as
s1 + s2 + · · · + sn
n
where
sk = a1 + a2 + · · · + ak
for each k, 1 ≤ k ≤ n. Suppose that the Cesáro sum of the 99-term sequence
a1 , a2 , . . . , a99 is 100. Find the Cesáro sum of the 100-term sequence 1,
a1 , a2 , . . . , a99 .

Problem 14. Each x ∈ [0, 1] can be written as

∞
X j
x=
j=1
2j

with j = 0 or j = 1. Define the function f : [0, 1] → [0, 1) as

∞
X 2j
f (x) = .
j=1
3j

The function f is known as Cantor function. Let x = 1/8. Find f (x).

Problem 15. Let s ≥ 2. Simplify the sum

∞ X
∞
X 1
.
j1 =1 j2 =1
(j 1 + j2 )s

Problem 16. Consider

∞
1+x X
2
= cj x j .
1−x−x j=0

Find cj .
 Sums and Products 5

Problem 17. The Cantor series approximation is defined as follows.

For arbitrary chosen integers n1 , n2 , . . . (equal or larger than 2), we can
approximate any real number r0 as follows

xj = integer part(rj ), j = 0, 1, 2, . . .
rj+1 = (rj − xj )nj

and
N
X xj
r0 ≈ x0 + .
j=1
n1 n2 · · · nj

The approximation error is

1
EN = .
n1 n2 · · · nN
Apply this approximation to r0 = 2/3 and the golden mean number with
nj = 2 for all j and N = 4.

Problem 18. Calculate the sum

√ √ √ √
q q q
S = 2 + 2 − 2 2 − 2 − 2 − 2.

Problem 19. Let n1 , n2 , n3 ∈ Z. Calculate

1 − eiπ(n1 +n2 +n3 ) .

√ √ √
Problem 20. Show that 7 + 2 10 has a square root the form x + y.

Problem 21. Calculate

∞
X j3 + 1
xj .
j=0
j!
3
Hint. Write j + 1 in the form

a + bj + cj(j − 1) + dj(j − 1)(j − 2).

Problem 22. Let n ∈ N. Find the sum

(n−1)/2
X
k2
k=−(n−1)/2
6 Problems and Solutions

where k runs in steps of 1.

Problem 23. Assume that the series

1, 14, 51, 124, 245, 426, . . .

is of the form
ak 3 + bk 2 + ck + d.
Find the integer coefficents a, b, c, d.

Problem 24. Find the sum

2 2 2 2
Sn = + + + ··· + .
1·2·3 2·3·4 3·4·5 n · (n + 1) · (n + 2)

Calculate Sn for n → ∞. Hint. Find a, b, c for

2 a b c
= + + .
n · (n + 1) · (n + 2) n n+1 n+2

Problem 25. Let a > 0. Show that

∞ ∞
X 1 1 X 1
= + 2z 2 − n2 a2
.
n=−∞
z − na z n=1
z

Problem 26. Let x 6= y, x 6= z, y 6= z. Find the sum

1 1 1
+ + .
(x − y)(x − z) (y − x)(y − z) (z − x)(z − y)

Problem 27. Given the expansion

x2 x3 x4
ln(1 + x) = x − + − + ···, −1 < x ≤ 1.
2 3 4
Calculate ln(11).

Problem 28. Let n ∈ N0 and p, a > 0. Show that

∞
X (k + n)! 1
exp(−2πkp/a) ≡ .
k!n! (1 − exp(−2πp/a))n+1
k=0
 Sums and Products 7

Hint: Start with

∞
X (k + n)!
f (n) := exp(−2πkp/a)
k!n!
k=0

and find f (n + 1).

Problem 29. Let (s1 , s2 ) ∈ Z2 . Let P (φ1 , φ2 ) be a probability density

and let θ1 , θ2 ∈ R. We can express the characteristic double sequence as
Z θ1 +2π Z θ2 +2π
χ(s1 , s2 ) = exp(i(s1 φ1 + s2 φ2 ))P (φ1 , φ2 )dφ1 dφ2 .
θ1 θ2

Find P (φ1 , φ2 ) as a function of χ(s1 , s2 ). Note that χ(0, 0) = 1.

Problem 30. Show that

π
= 4 arctan(1/5) − arctan(1/239).
4
Hint. Let θ = arctan(1/5). Thus tan(θ) = 1/5. Applying the double angle
formula for the tangent we have tan(2θ) = 5/12 and tan(4θ) = 120/119.
Finally apply that tan(π/4) = 1.

Problem 31. Let n ∈ N0 . Consider an infinite number of time variable

t = (t1 , t2 , t3 , . . .). Consider the sum
X xk0 tk11 tk22 · · ·
pn (x + t1 , t2 , t3 , . . .) = .
k0 +k1 +2k2 +3k3 +···=n
k0 !k1 !k2 ! · · ·
k0 ,k1 ,k2 ,k3 ,...≥0

Find p0 , p1 , p2 , p3 .

Problem 32. Let n be an integer with n ≥ 1. Simplify the series

1 1 1 1
S= + + + ··· + .
1·2 2·3 3·4 n(n + 1)

Problem 33. Let N be a positive integer and a, b be positive integers or

positive half-integers. Find
1
f (N, a, b) = (1 + (−1)2a + (−1)2b + (−1)2a+2b+N ).
2

Problem 34. Let −1 < x < +1 and n ≥ 2.

8 Problems and Solutions

(i) Show that

n−1
X nxn x(1 − xn )
Sn (x) = jxj = + .
j=0
x−1 (1 − x)2

(ii) Show that

x
lim Sn (x) = .
n→∞ (1 − x)2
(iii) Let x = 1/2. Show that limn→∞ Sn (1/2) = 2.

Problem 35. Let x, y, z ∈ R and x 6= y, x 6= z, y 6= z. Show that

x y z
+ + = 0.
(z − x)(x − y) (x − y)(y − z) (y − z)(z − x)

Problem 36. Let

x 6= 0, x 6= ±2π, x 6= ±4π, . . .

and n ∈ N. Show that

cos(x/2) − cos((n + 1/2)x)
sin(x) + sin(2x) + · · · + sin(nx) = .
2 sin(x/2)

For the values x = 0, x = ±2π, x = ±4π etc the sum is given by 0.

Problem 37. Let n = 0, 1, 2, . . .. The function

1 sin2 n+1



2
θ
Kn (θ) :=
n + 1 sin2 12 θ

is called the Fejér kernel.

(i) Show that
j=+n
X  
|j|
Kn (θ) := 1− eijθ .
j=−n
n+1

(ii) Show that Kn (θ) ≥ 0.

(iii) Show that for any continuous 2π periodic function f on R one has
Z π
1
Kn ? f (θ) := Kn (θ − α)f (α)dα
2π −π
j=+n
X   Z π 
|j| 1
= 1− e−ijα f (α)dα eijθ .
j=−n
n+1 2π −π
 Sums and Products 9

(iv) Show that Kn ? f (θ) → f (θ) uniformely in θ as n → ∞.

Problem 38. Let λ > 0.

(i) Calculate
∞
X je−λ λj
S1 (λ) = .
j=0
j!

(ii) Calculate
∞
X j(j − 1)e−λ λj
S2 (λ) = .
j=0
j!

Problem 39. (i) Let α > 0. Show that

∞
X (−1)k cos(kx) π cosh(αx) 1
= · − 2
k 2 + α2 2α2 sinh(αx) 2α
k=1

where −π ≤ x ≤ π.
(ii) Let α > 0. Show that
∞
X (−1)k sin(kx) π sinh(αx)
=− ·
k 2 + α2 2α2 sinh(απ)
k=1

for −π < x < π.

Problem 40. Let a > 0 and b > 0. Show that the continued fraction
1 1 1 1
a+ ···
b+ a+ b+ a+
can be written as r
a a2 a
+ + .
2 4 b

Problem 41. Let M, N ≥ 1. Find the sum

N
M X
X (m + n)!
.
m=1 n=1
(2m + n)!(m + 2n)!

Problem 42. Study the sum

∞
X `
x`+m+n .
(`m + `n + mn + 1)2
`,m,n=1
10 Problems and Solutions

Problem 43. Show that

∞
X 1
2 + 3n + 1 = 4π.
n=−∞
n 4 8

Problem 44. Let n ≥ 1. Find the sum

n
X 3
.
j=1
j(j + 3)

What happens for n → ∞?

Problem 45. Let F1 = 1, F2 = 2, Fk+1 = Fk−1 + Fk be the Fibonacci

numbers. Calculate
∞
X log2 (Fk+1 )
.
2k+1
k=1

Problem 46. Let Z be the set of integers. Simplify the sum

X sin(π(x − k)) sin(π(y − k))
.
π(x − k)π(y − k)
k∈Z

Problem 47. Let n ≥ 1. One has

n  
X n k
(1 + x)n = x . (1)
k
k=0

Show that
         
n n n n n
1· +2· +3· + · · · + (n − 1) · +n· = n2n−1 . (2)
1 2 3 n−1 n

Problem 48. Let n > 2 and A, h > 0. Find Vn given by

n−1
Ah X 2
Vn = j .
n3 j=1

Then find limn→∞ Vn . This concern the volume of a pyramide, h is the

height and A is the area of the base.
 Sums and Products 11

Problem 49. Let n ≥ 1 and c1 , . . . , cn be constants. Find the minima

of the function f : R → R
n
X
f (x) = (x − cj )2 .
j=1

Problem 50. Let x ∈ R. The sequence of functions { fk (x) } is defined

by f1 (x) = cos(x/2) and for k > 1 by

fk (x) = fk−1 (x) cos(x/2k ).

Thus
fk (x) = cos(x/2) cos(x/22 ) · · · cos(x/2k ).
Obviously, we have fk (0) = 1 for every k. Calculate limk→∞ fk (x) as a
function of x for x 6= 0.

Programming Problems

Problem 1. Let m, n ≥ 2. Calculate the finite sum

m X
X n
Sn,m = |j − k|.
j=1 k=1

Give a C++ implementation.

Problem 2. Let n ∈ N. Calculate the finite sum

n
X n
X
Sn = |j + k|.
j=−n k=−n

Problem 3. Let m ≥ 1 and n ≥ 1. Find the sum

min(m,n)
X
(n + m − 2k + 1).
k=0
12 Problems and Solutions

1.2 Supplementary Problems

Problem 1. Show that

1 + 2 tanh2 (λ) + 3 tanh4 (λ) + · · · + (n + 1) tanh2n (λ) + · · · ≡ cosh4 (λ).

Note that tanh(0) = 0 and cosh(0) = 1.

Problem 2. Show that

∞
X 1
= 2 ln(2) − 1.
j=2
j(2j − 1)

Problem 3. (i) Show that

1 1 1 1
= + + + ···
4 1·2·3 2·3·4 3·4·5
(ii) Let n ≥ 1. Show that
1 1 1 n
+ + ··· + = .
1·2 2·3 n(n + 1) n+1

Problem 4. Let n, m ∈ N and a, b, c, d ∈ R. Show that

n X m   
n m
X n m r n−r s m−s
(a + b) (c + d) = a b c d .
r=0 s=0
r s

Problem 5. Let ` ≥ 1. Study the Gauss sum

`−1
1 X 2πikr2 /`
G(k, `) = √ e .
` r=0

Problem 6. Show that if |x| < 1 we have the expansion

1
= 1 + 2x + 3x2 + 4x3 + · · · .
(1 − x)2
Let a > 0. Use this expansion to show that
2x 3x2 4x3
 
1 1
= 1 − + − + · · · .
(a + x)2 a2 a a2 a3
 Sums and Products 13

Problem 7. Apply mathematical induction to show that

1 + 3 + 5 + · · · + (2n − 1) = n2
1
12 + 42 + 72 + · · · + (3n − 2)2 = n(6n2 − 3n − 1)
2
13 + 33 + 53 + · · · + (2n − 1)3 = n2 (2n2 − 1).

Problem 8. Show that

1 1 1
ln(2) = + + + ···
1·2 3·4 5·6

Problem 9. Let x > 0. Show that a complicated way to calculate 1/x is

given by
1 1! 2!
+ + + ···
x + 1 (x + 1)(x + 2) (x + 1)(x + 2)(x + 3)

Problem
√ 10. Let y ∈ R and byc be the integer part of y. Let ω =
(1 + 5)/2 be the golden mean number (which is an irrational number).
We define
xj := j + ((j + 1)/ω)(λ − 1), j = 0, 1, . . .
where λ > 0 and

µk = xk − xk−1 , k = 1, 2, . . .

Show that the sequence µ1 , µ2 , . . . is non-periodic.

Problem 11. Let s > 0 and

Z ∞
ln(1 + st)
f (s) = dt.
t=0 1 + t2
Show that
πs2
 
ln(s) 1
f (s) = (1 − ln(s))s + + − s3 + O(s4 ).
4 3 9

Problem 12. Let n be an integer and n > 1. Show that

1 1 1
sn = 1 + + + ··· +
2 3 n
14 Problems and Solutions

cannot be integer for all n > 1.

Problem 13. Let x ∈ R. Show that the series

x x2 x3
+ + + ···
1 + x2 1 + x4 1 + x6
converges absolutely for all values of x, except for +1 and −1.

Problem 14. Show that the series

∞ ∞ ∞
X 1 X z X z(1 + z)
zj = , jz j = , j 2zj =
j=0
1 − z j=0
(1 − z)2 j=0
(1 − z)3

converge for all |z| < 1.

Problem 15. Let n be an integer with n ≥ 2. Find

 
X 1
 − 1 .

 2i j
3 
1≤i<j≤n

Problem 16. Let ` ≥ 1. Show that

`
X 1 `
=
j=1
4j 2 − 1 2` + 1

and
`  
X 1 1 2
= 1− .
j=2
j(j 2 − 1) 4 `(` + 1)

Problem 17. Let 0 ≤ r < 1. Calculate the sum

∞
X
rj cos(jφ)
j=1

utilizing the identity

1 ijφ
cos(jφ) ≡ (e + eijφ ).
2

Problem 18. Calculate the sum

∞
X 1
j=1
j2 + 1
 Sums and Products 15

with the knowledge that

∞ ∞ ∞ ∞
X 1 X 1 X 1 X 1 π2
2
≡ 2
− 2 2
, 2
= .
j=1
j + 1 j=1 j j=1
j (j + 1) j=1
j 6

Problem 19. Show that

r
1√ √ √
q q
1 1
cos(π/4) = 2, cos(π/8) = 2 + 2, cos(π/16) = 2+ 2+ 2
2 2 2
r
1√ √ √
q q
1 1
sin(π/4) = 2, sin(π/8) = 2 − 2, sin(π/16) = 2 − 2 + 2.
2 2 2

Problem 20. Let n ≥ 0 and x 6= 0. Find

n
X
Sn (x) = kxk .
k=0

Utilize that
n
X n
X n
X
Sn+1 (x) = Sn (x) + (n + 1)xn+1 = (k + 1)xk+1 = kxk+1 + xk+1
k=0 k=0 k=0
n
X
= xSn (x) + xk+1
k=0

and a case study with x = 1 and x 6= 1.

Problem 21. The Fibonacci numbers are given by xt+2 = xt+1 + xt with
t = 0, 1, . . . and x0 = x1 = 1. Consider
1 1 1 1 1 1 1 1 1 1 1 1
F = + + + + + +··· ≡ + + + + + +··· .
x0 x1 x2 x3 x4 x5 1 1 2 3 5 8
Show that the series converges. Is F < 4?

Problem 22. Calculate the sum

∞
X 2n + 1
S= 2
.
j=1
n (n + 1)

Problem 23. Show that the Cantor series

c2 c3 c4
c1 + + + + ···
2! 3! 4!
16 Problems and Solutions

with 0 ≤ cn ≤ n − 1 is convergent.

Problem 24. Let N ≥ 1 be an integer and k = −4N, −4N + 1, . . . , 4N .

We define
N   
1 X 1 kπ
f (k) = cos j− .
N j=1 2 N

Show that f (0) = 1, f (±2N ) = −1, f (±4N ) = 1 and 0 otherwise.

Problem 25. Show that

X 1
= π cot(πx).
(x − k)
k∈Z

Note that cot(0) = ∞ and cot(π/2) = 0.

Problem 26. Let c > 0. Apply the Poisson summation formula

+∞
X +∞
X Z +∞
f (n + a) = exp(2πika) f (x) exp(−2πikx)dx
n=−∞ k=−∞ −∞

for the function

x exp(−cx2 ) for x ≥ 0

f (x) :=
0 for x < 0

Problem 27. Let k, m = 1, 2, . . .. Write a C++ program that implements

the sum
m−1
X k + m − 1
1
E(k, m) = k+m−1
2 i=0
i

for a given k and m. The sum plays a role for the arithmetic triangle.

Problem 28. Let L ∈ N0 and κ be a nonnegative real number.

(i) Show that

L  
!2
X L −κLn
Z(κ) = e = (1 + e−κL )2L .
n=0
n

(ii) Show that for κ > 0 one has for large L that Z ≈ 1 + 2Le−κL . Show
that at κ = 0 one has Z(κ = 0) = 22L .
 Sums and Products 17

Problem 29. Let n > 2. Show that

n
X 1
≈ ln(n).
j=2
j

Problem 30. Let n ≥ 1 and ω be the primitive nth root of 1. Show that
n−1
X
ω j = 0.
j=0

Problem 31. Show that

∞
X x4n
ex + e−x + 2 cos(x) = 4 .
n=0
(4n)!

Problem 32. Show that

 
∞ ∞
!
k
X 1 X (−1)
  =1
j=0
j! k!
k=0

Problem 33. Let n ∈ N. Applying the principle of mathematical induc-

tion show that
n
X n(n + 1)(n + 2)
A(n) = k(k + 1) = .
3
k=0

Note that A(0) = 0.

Problem 34. Let N ≥ 2. Find the sum

N
X
(j1 j2 j3 + (j1 + j2 + j3 )).
j1 ,j2 ,j3 =1
j3 ≥j2 ≥j1

Problem 35. Let h, r > 0 and fixed and m, n ∈ N. Consider

s 
2
h 2
Sm,n = mnr sin(π/n) + (r − r cos(π/n))
m
s 4
m2 sin(π/n)

sin(π/n) 1
= πr h2 + π 4 r 2 4 .
π/n 4 n π/n
18 Problems and Solutions

(i) Consider limm,n→∞ Sm,n with m = n2 .

(ii) Consider limm,n→∞ Sm,n with m = n3 .
(iii) Consider limm,n→∞ Sm,n with limm→∞,n→∞ (m2 /n4 ) = 0 and show
that limm,n→∞ Sm,n = πrh which is a surface area of a half cylinder with
radius r and length h.

Problem 36. (i) Let a > 0. Show that

∞
X cos(kx) π cosh(a(π − x)) 1
= − .
k 2 + a2 2a sinh(aπ) 2a
k=1

(ii) Let a > 0. Show that

∞
X 1 π  aπ 
= tanh .
(2k + 1)2 + a2 2a 2
k=−∞

Problem 37. Let k ≥ 2. Consider

k
(−1)k (−1)j
 
1 1 1 X
f (k) = k! − + − ··· + ≡ k! .
2! 3! 4! k! j=2
j!

Thus f (2) = 1. Find f (3), f (4), f (5), f (6), ... . Which of these numbers
is a prime number?

Problem 38. Let M be an integer with M ≥ 2 and =(φ) < 0. Show that
M −1
X e(2πimk)/M 4kM e−2πikφ/M M 2 e−2πikφ/M
S(k) = 2 =− + .
m=1
sin (π(m + φ)/M ) 1−e −2πiφ
sin2 (πφ)

Note that
∞
1 X
= −4 pe−2πi(m+φ)p/M .
sin2 (π(m + φ)/M ) p=0

Problem 39. The q-exponential function is defined by

∞
X xn
exq :=
j=0
[j]!

where
q j − q −j
[j] := .
q − q −1
 Sums and Products 19

Find exq for x = 1 and q = 1/2.

Problem 40. Let a, b, c ∈ R. Factorize

(b − c)3 + (c − a)3 + (a − b)3 .

The following Maxima may be helpful

T: (b-c)^3 + (c-a)^3 + (a-b)^3;
T: expand(T);
T: ratsimp(T);

Problem 41. Let n ≥ 0. Starting from

n
X
cj xj = (1 + x)n
j=0

show that

c0 + c1 + · · · + cn = 2 n
c0 − c1 + c2 − c3 + · · · + (−1)n cn = 0
c1 + 2c2 + 3c3 + · · · ncn = n2n−1
(2n)!
c20 + c21 + · · · + c2n = .
(n!)2

Problem 42. Let n ≥ 1. Show that

n
X
ln(1 + 1/j) = ln(n + 1).
j=1

Problem 43. Let x ∈ (−1, 1). Show that

∞
X x
jxj = .
j=0
1 − x)2

Problem 44. (i) Assuming that

1 1 1 1 1 1 1
1− + − + − + − + · · · = ln(2).
2 3 4 5 6 7 8
20 Problems and Solutions

Show that
1 1 1 1 1 1 1 1 1
1− − + − − + − − + · · · = ln(2).
2 4 3 6 8 5 10 12 2
Note that the convergence of the first series is not absolute.
(ii) Show that
1 1 1 1
1 + + + ··· + − ln(n)
3 5 2n + 1 2
tends to a finite value as n → ∞. Find this value.

Problem 45. Let β > 0. Show that

∞
X 1
e−βn = .
n=0
1 − e−β

Problem 46. Let n be a positive integer and f (θ1 , . . . , θN ) be a periodic

function, i.e. periodic 2π for each θj (j = 1, . . . , N ). Show for large n we
have ! N n
YN Z 2π X
dθk  cos(θj ) f (θ1 , . . . , θN )
k=1 0 j=1

N Z
!
∞
Y 2
≈ Nn dθk e−n(θk ) /(2N )
f (θ1 , . . . , θN ).
k=1 −∞

Problem 47. Let A, B be finite sets and n(A), n(B) the numbers of
elements in A and B, respectively. Is

n(A) + n(B) = n(A ∪ B) + n(A ∩ B) ?

Prove or disprove.

Problem 48. Let f : [0, 1] → [0, 1] be a differentiable function, for

example the logistic map f (x) = 4x(1 − x). Let f (k) be the k-th iterate of
the function f . Let  
 d
bk := sup  f (k) (x) .

0≤x≤1 dx

Show that
lim (bk )1/k
k→∞

exist. Apply the chain rule.

 Sums and Products 21

Problem 49. Let n ≥ 2. Show that

n  
X n
k = n2n−1 .
k
k=1

Problem 50. Is the series

∞
X k
S=
1 + 2k 3
k=1

convergent?

Problem 51. Give a non-trival infinite sequence (x0 , x1 , x2 , . . .) such that

∞
X |xj |
j=0
1 + |xj |

is finite.

Problem 52. Let a > 0. Show that

∞
X 1 π
(−1)j = .
j=−∞
j2 +a2 a sinh(πa)

Problem 53. Let n = 0, 1, 2, . . .. The Fermat numbers are given by

n
Fn = 2(2 )
+ 1.

Show that the Fermat numbers satisfy the recurrence relation

Fn+1 = (Fn − 1)2 + 1

with F0 = 3

Problem 54. Let N ≥ 1. Consider the Hilbert space L2 ([−1, 1]). The
Chebyhev-Gauss-Lobatto points are given by

xj = cos(jπ/N ), j = 0, 1, . . . , N.

Find the four points for N = 3.

22 Problems and Solutions

Problem 55. Let |z| < 1. Show that

∞
X ∞
X
jz j = (1 − z)−1 zj .
j=1 j=1

Show that
∞
X ∞
X
j 2 z j = (1 + z)(1 − z)−2 zj
j=1 j=1

Problem 56. Show that

∞  n
1 2 2n 1X 2 1 1
+ + · · · + n+1 + · · · = = = 1.
3 9 3 3 n=0 3 3 1 − 2/3

Problem 57. (i) Let a > 0. Show that

∞
X 1 1
= .
n=0
(a + n)(a + n + 1) a

(ii) Show that

∞
X 1 1
= .
n=1
n(n + 1)(n + 2) 4
(iii) Let n ≥ 2. Show that
n−1
X 1 1
=1−
j=0
(j + 1)(j + 2) n+1

and thus show that

∞
X 1
= 1.
j=0
(j + 1)(j + 2)

Problem 58. Let n be a positive integer. Show that

sin(2n α)
= cos(α) cos(2α) cos(22 α) · · · cos(2n−1 α).
2n sin(α)

Problem 59. Let Γ(x) be the gamma function. Show that

n  
k n (α + 2k)Γ(α + k)
X
(−1) = δn,0 .
k Γ(α + k + n + 1)
k=0
 Sums and Products 23

Problem 60. Let a > 0 and a ∈/ N. Show that

 
2a sin(ax) 1 cos(x) cos(2x) cos(3x)
cos(ax) = − 2 + 2 − 2 + ···
π 2a2 a − 12 a − 22 a − 32

for all x ∈ [−π, π].

Problem 61. (i) Let n ∈ N. Show that

1
14 + 24 + · · · + n4 = n(n + 1)(2n + 1)(3n2 + 3n − 1).
30
(ii) Let n ≥ 1. Show that
n
X n5 n4 n3 n
k4 = + + − .
5 2 3 30
k=0

Problem 62. Let N ≥ 1. Find the sums

N −1    
X 2π
vk = exp jk − 1 , k = 0, 1, . . . , N − 1.
j=0
N

Problem 63. (i) Let ` be a positive integer and φ ∈ R. Show that

+`
X sin((` + 1/2)φ)
eimφ = .
sin(φ/2)
m=−`

(ii) The function

sin((n + 12 )θ)
Dn (θ) :=
sin( 21 θ)
is called the Dirichlet kernel. Show that
k=+n
X
Dn (θ) = eikx .
k=−n

Problem 64. Let k be a positive integer and k ≥ 2. Show that

 
k−1
X
1 + x − 2xk ≡ (1 − x) + 2x(1 − xk−1 ) ≡ (1 − x) 1 + 2 xj  .
j=1
24 Problems and Solutions

Problem 65. Let n ≥ 2. Let Sn be the standard n-simplex embedded in

Rn
 
 X n 
Sn := x ∈ Rn : xj = 1, for xk ≥ 0, for k = 1, . . . , n .
 
j=1

We denote by Sk,n−1 the kth face of the boundary of Sn . They are (n − 1)-
simplexes

Sk,n−1 := { x ∈ Rn : xk = 0, x ∈ Sn } for k = 1, . . . , n.

Show that the boundary ∂Sn of Sn is the union

∂Sn = ∪nj=1 Sj,n−1

of the faces.

Problem 66. Show that

∞ x
Y sin(x)
cos =
j=1
2j x

for x ∈ R.

Problem 67. The triple sum

∞ ∞ ∞
X X X
0 (−1)k1 +k2 +k3
p
k1 =−∞ k2 =−∞ k3 =−∞
k12 + k22 + k32

plays a role in solid state physics (Madelung constant). Here 0 indicates

that the term (k1 , k2 , k3 ) = (0, 0, 0) is omitted. Given a positive integer N .
Write a C++ program that implements the sum
N N N
X X X
0 (−1)k1 +k2 +k3
p .
k1 =−N k2 =−N k3 =−N
k12 + k22 + k32

Run the program for different N and compare with the exact value.

Problem 68. Find the sum

L  
X 2
SL = exp iπ n2
n=1
L

for L = 1, 2, 3.
 Sums and Products 25

Problem 69. Show that the series

∞
X sin(jx)
√
j=1
j

converges for all x ∈ R.

Problem 70. Let k ≥ 0 and s > 0. Show that

k
Y s+`+1 s+k+1
= .
s+` s
`=0

Problem 71. Let n be a positive integer. Give a C++ implementation

of the sum
n  
n X
X k
`
k=0 `=0

using templates so that the Verylong class of SymbolicC++ can be used.

Problem 72. Let n, m ≥ 1. Implement the sum

m   
(2m)! X k n m
Sn,m := 2
22m m! k k
k=0

using the Verylong class and Rational class of SymbolicC++.

Problem 73. Let n ∈ N and m1 , m2 , . . . , mn ∈ N0 . Consider the function

f (m1 , m2 , . . . , mn ) :=

 δm1 ,0 for n = 1
δm2 ,0 δm1 ,1 for n = 2
Qn−2 Pj
δmn ,0 δ(m1 +···+mn ),n−1 j=1 H(j − k=1 mn−k ) for n ≥ 3


where H is the step function


1 for x ≥ 0
H(x) :=
0 for x < 0

Give an implementation of this function in SymbolicC++ for a given n

using the Verylong class.
Chapter 2

Maps

2.1 Solved Problems

Problem 1. Newton’s sequence takes the form of a difference equation

f (xt )
xt+1 = xt −
f 0 (xt )

where t = 0, 1, 2, . . . and x0 is the initial value at t = 0. Let f : R → R be

given by
f (x) = x2 − 1
and x0 6= 0. Find the fixed points of f . Find the fixed points of the
difference equation. Let x0 = 1/2. Find x1 , x2 , x3 . Find Newton’s sequence
for this function. Obtain the exact solution of the difference equation.

Problem 2. (i) Solve the nonlinear recurrence relation

xn+1 = x2n , n = 0, 1, . . .

where x0 = 2.
(ii) Solve the linear recurrence relation

xn+1 = xn + xn−1 + xn−2 , n = 2, 3, . . .

and the initial values x0 = x1 = x2 = 1.

26
 Maps 27

(iii) Solve the linear recurrence relation

n−1
X
xn+1 = 1 + xj , x0 = 1.
j=0

Problem 3. Let x > 0 and p > 0. Consider the map

f (x) = xep−x .

(i) Find the fixed points. Study the stability of the fixed points.
(ii) Show that f has a least one periodic point x∗ with x∗ 6= 0 or p.

Problem 4. Let f : R → R+ be a positive, continously differentiable

function, defined for all real numbers and whose derivative is always neg-
ative. Show that for any real number x0 (initial value) the sequence (xk )
obtained by Newton’s method
f (xk )
xk+1 = xk −
f 0 (xk )
has always limit ∞.

Problem 5. Let f : R → R be a continuosly differentiable map. Let f (n)

be the n-th iterate of f . Calculate the derivative of f (n) at x0 .

Problem 6. (i) Solve the second order linear difference equation

xt+2 = xt+1 + xt t = 0, 1, 2, . . .

where x0 = 0 and x1 = 1.
(ii) Give the definition of the golden mean number and derive this number.
(iii) Calculate
xt+1
lim .
t→∞ xt

Problem 7. The recursion relation

Fn+2 = Fn + Fn+1

with the initial values F0 = F1 = 1 provides the Fibonacci sequence

1, 1, 2, 3, 5, 8, 13, 21, 34, 55, . . .. A generalization of the Fibonacci sequence
is the q-analogue of the sequence defined by the recursion relation

Fn+2 (q) = Fn (q) + qFn+1 (q)

28 Problems and Solutions

and the initial condition F0 (q) = 1 and F1 (q) = q. Here, q is a real or

complex number.
(i) Give the first five terms of the sequence.
(ii) Find a generating function of Fn (q).
(iii) Find an explicit expression for Fn (q).

Problem 8. (i) Find the linear map f : {0, 1} → {−1, 1} such that

f (0) = −1, f (1) = 1. (1)

(ii) Find a linear map g : {−1, 1} → {0, 1} such that

g(−1) = 0, g(1) = 1. (2)

This is obviously the inverse map of f .

Problem 9. Consider the differentiable function f : R → R

1 − cos(2πx)
f (x) =
x
where using L’Hospital f (0) = 0.
(i) Find the zeros of f .
(ii) Find the maxima and minima of f .

Problem 10. Given two manifolds M and N , a bijective map φ from M

to N is called a diffeomorphism if both φ : M → N and its inverse φ−1 are
differentiable. Let f : R → R be given by the analytic function

f (x) = 4x(1 − x)

and the analytic function g : R → R be given by

g(x) = 1 − 2x2 .

(i) Can one find a diffeomorphism φ : R → R such that

g = φ ◦ f ◦ φ−1 ?

(ii) Consider the diffeomorphism ψ : R → R

ψ(x) = sinh(x).

Calculate ψ ◦ g ◦ ψ −1 .
 Maps 29

Problem 11. Consider the polynomial p(x) = x3 − 3x + 3. Show that for

any positive integer N , there is an initial value x0 such that the sequence
x0 , x1 , x2 , . . . obtained from Newton’s method

p(xn ) 2x3n − 3
xn+1 = xn − 0
= , n = 0, 1, 2, . . .
p (xn ) 3(x2n − 1)

has period N .

Problem 12. Consider the linear recursion equation for F (n) with n =
0, 1, 2, . . .
(n + 1)F (n + 1) − nF (n − 1) + F (n) = 0
where F (0) = 0, F (1) = 1. We define
∞
X
f (z) := F (n)z n
n=1

where z is an undeterminate. Find the differential equation for f with the

initial condition. Solve the differential equation.

Problem 13. Let f : Rn → Rn be an analytic function. Consider the

map
xj = f (xj−1 ) = · · · = f (x0 ).
To study the evolution of phase-space distributions, we can introduce the
evolution operator U (x0 , x, j) such that any initial phase-space distribution
ρ(x, 0) evolves into
Z
0
ρ(x ; j) = U (x0 , x; j)ρ(x, 0)dx
Ω

where Ω is the phase space area. Find U (x0 , x; j).

Problem 14. Find the solution of the recursion relation

2 1
aj+2 = aj+1 + aj , j = 1, 2, . . .
3 3
with a1 = 5, a2 = 1. Calculate limj→∞ aj .

Problem 15. Solve the linear difference equation

xt+1 = 2xt − t, t = 0, 1, 2, . . .

with the initial value x0 = 1.

30 Problems and Solutions

Problem 16. Let

N0 := { 0, 1, 2, . . . }
Find an invertible map f : N0 × N0 × N0 → N0 with

f (0, 0, 0) = 0, f (0, 0, 1) = 1, f (0, 1, 0) = 2, f (0, 1, 1) = 3,

f (1, 0, 0) = 4, f (1, 0, 1) = 5, f (1, 1, 0) = 6, f (1, 1, 1) = 7

etc. Find the inverse function.

Problem 17. Consider the second order difference equation

xt+2 = xt+1 xt

with the initial values x0 = a, x1 = b. Give the solution. What are the
fixed points?

Problem 18. Let A be a set. Suppose that it is possible to define subsets

A1 , A2 , . . . of A which have the properties that
(i) the sets are pairwise disjoint; that is Ai ∩ Aj = ∅ for all i, j = 1, 2, . . .
and j 6= i

(ii) A1 ∪ A2 ∪ · · · = A.

Then the family of sets { A1 , . . . } is called a partition of A.

Let S = {1, 2, 3, . . . , 9} and A = {1, 4, 7}. B = {2, 3, 5, 6}, C = {7, 8, 9}. Is

{A, B, C} a partition of S?

Problem 19. Let S be a finite set. Let P(S) be the power set of S. Then
P(S) has 2|S| elements.
(i) Show that the composition

A ◦ B := (A ∪ B) ∩ ({A ∪ {B)

defines a group, where A, B ∈ P(S). Here { denotes the complement.

(ii) Show that the composition

A ◦ B := (A ∩ B) ∪ ({A ∩ {B)

defines a group, where A, B ∈ P(S). Here { denotes the complement.

Problem 20. Find a function f : [0, 1] → (0, 1] that is one-to-one.

 Maps 31

Problem 21. Let n ∈ N. Consider the map

 
1 n
ut+1 = ut + , t = 0, 1, 2, . . .
2 ut
given the initial value u0 with u0 > 0. Show that
√  √ 2
ut+1 − n ut − n
√ = √ .
ut+1 + n ut + n
√ √
Show that ut → n as t → ∞. Show that n is a fixed point.

Problem 22. Let 0 ≤ α < π/4. Consider the transformation

1
X(x, y, α) = p (x cos(α) + iy sin(α))
cos(2α)
1
Y (x, y, α) = p (−ix sin(α) + y cos(α)).
cos(2α)

(i) Show that X 2 + Y 2 = x2 + y 2 .

(ii) Do the matrices
 
1 cos(α) i sin(α)
cos(2α) −i sin(α) cos(α)
p

form a group under matrix multiplication?

Problem 23. A smooth function f : Rn → R is called homogeneous of

degree r if
f (x1 , . . . , xn ) = r f (x1 , . . . , xn ). (1)
Show that (Euler’s identity)
n
X ∂f
xj = rf.
j=1
∂xj

Problem 24. Find the invariance group of the function f : R → R

f (x) = sin(x).

Problem 25. Consider the analytic function f : R → R

f (x) = cos(sin(x)) − sin(cos(x)).

32 Problems and Solutions

Show that f admits at least one critical point. By calculating the second
order derivative find out whether this critical point refers to a maxima or
minima.

Problem 26. Let f : R2 → R, g : R2 → R be analytic functions. We

define the star product

f (x1 , x2 ) ? g(x1 , x2 ) :=
 
∂ ∂ ∂ ∂
lim exp − f (x1 , x2 )g(x01 , x02 ).
x01 →x1 ,x02 →x2 ∂x1 ∂x02 ∂x01 ∂x2
Let
f (x1 , x2 ) = sin(x1 + x2 ), g(x1 , x2 ) = sin(x1 − x2 ).
Find the star product.

Problem 27. Let N1 , N2 be given positive integers. Let n1 = 0, 1, . . . , N1 −

1, n2 = 0, 1, . . . , N2 − 1. There are N1 · N2 points. The points (n1 , n2 ) are
a subset of N0 × N0 and can be mapped one-to-one onto a subset of N0

j(n1 , n2 ) = n1 N2 + n2

where j = 0, 1, . . . , (N1 − 1)(N2 − 1). Find the inverse of this map.

Problem 28. Let a ∈ R. Consider the transformation

1 1 ax
t(t, x) = eax sinh(at),
e x
e(t, x) = (e cosh(at) − 1)
a a
with lima→0 e
t = t, lima→0 x
e = x. Find the inverse of the transformation.

Problem 29. Show that the map f : (0, 1) → R

x − 1/2
x 7→ f (x) =
x(x − 1)

is bijective. Note that f (1/2) = 0 and f (1/4) = 4/3.

Problem 30. Consider the map

f : R → { (x1 , x2 ) : x21 + (x2 − 1)2 = 1 ∧ x2 6= 2 }

defined by
2t2
 
4t
f (t) = 2
, 2 .
t +4 t +4
 Maps 33

Show that f is bijective. Show that f and f −1 are continuous.

Problem 31. Let N ≥ 2 and δ := 1/(N + 1). Solve the linear one-
dimensional linear difference equation

xj+1 − 2xj + xj−1 = −12δ 4 (j + 1)2 , j = 0, 1, . . . , N − 1

with the boundary conditions x−1 = xN = 0.

Problem 32. (i) Consider the analytic map f : R → R

f (x) = x(1 − x).

Find the fixed points of f . Are the fixed points stable? Prove or disprove.
(ii) Let x0 = 1/2 and iterate

f (x0 ), f (f (x0 )), f (f (f (x0 ))), . . .

Does this sequence tend to a fixed point? Prove or disprove.

(iii) Let x0 = 2 and iterate

f (x0 ), f (f (x0 )), f (f (f (x0 ))), . . .

Does this sequence tend to a fixed point? Prove or disprove.

(iv) Find the critical points of f . Then find the extrema of f .
(v) Find the roots of f , i.e. solve f (x) = 0.
(vi) Find the minima of the function g(x) := |f (x)|.

Problem 33. Let c > 0 and 0 ≤ v1 < c, 0 ≤ v2 < c. We define the

composition
v1 + v2
v1 ? v2 := .
1 + v1 v2 /c2
Is the composition associative?

Problem 34. Solve the initial value problem of the system of first order
difference equations

x1,t+1 = −2x1,t
8
x2,t+1 = x1,t − x2,t .
9
The first difference equation is independent of x2,t and we find the solution
of the initial value problem

x1,t+1 = (−2)t x1,0 .

34 Problems and Solutions

Problem 35. Let mA , mB , RA , RB be the masses and centre-of mass

coordinates of mass A and B, respectively. We set m = mA + mB . Find
the inverse of the transformation
1
r(RA , RB ) = RA − RB , R(RA , RB ) = (mA RA + mB RB ).
m

Problem 36. Consider the map f : N0 × N0 → N0 × N0

f1 (n1 , n2 ) = |n1 − n2 |, f2 (n1 , n2 ) = n1 + n2 .

Is the map invertible?

Problem 37. Let n, m ≥ 0. Consider the differential operators

n m
X dj X dj
Kn := uj (x) , Lm := vj (x)
j=0
dxj j=0
dxj

where the uj (x)’s and vj (x)’s are smooth functions. If the two differential
operators Kn and Lm commute, then there is a nonzero polynomial R(z, w)
such that R(Kn , Lm ) = 0. The curve Γ defined by R(z, w) = 0 is called the
spectral curve. If we consider the eigenvalue problem

Kn ψ = zψ, Lm ψ = wψ

then (z, w) ∈ Γ.
(i) Let (α ∈ R)
2
d2

3
K= + x + α + 2x,
dx2
 2 3
d 3 d2 d
L= 2
+ x + α + 3x 2
+3 + 3x(x2 + α).
dx dx dx
Show that K and L commute with w2 = z 3 − α.
(ii) Show that
 3 2
d 2 d
K= 3
+ x + α +2
dx dx
and 3
d3 d4

d
L= + x2 + α +3 + 3(x2 + α) + 3x
dx3 dx 4 dx
commute.

Problem 38. Let A, B, C be nonempty sets. Let f : A → B and

g : B → C are invertible maps. Then the composition of the functions
 Maps 35

g ◦ f : A → C in invertible and (g ◦ f )−1 = f −1 ◦ g −1 . Function composition

is associative.

Problem 39. Consider the map f : (−1, 1) → R

x
f (x) = .
1 − |x|

Show that the map f −1 : R → (−1, 1) is given by

x
f −1 (x) = .
1 + |x|

Problem 40. Let τ ∈ R. Show that the curves

       
x1 (τ ) 1 y1 (τ ) τ
x(τ ) = = , y(τ ) = =
x2 (τ ) τ y2 (τ ) τ2

are linearly independent.

Problem 41. Let r, k be positive integers with r ≤ k and M =

{x1 , . . . , xr } and N = {y1 , . . . , yk }. Then there
k!
(k − r)!
surjective maps f : M → N . Find all the maps for r = k = 2.

Problem 42. Consider the first order difference equation

xτ +1 xτ + 2xτ + xτ +1 = 0, τ = 0, 1, 2, . . .

with x0 6= 0. Show that this difference equation can be linearized with

xτ = 1/yτ where x0 6= 0. First show that x∗ = 0 and x∗ = −3 are solutions
of the nonlinear difference equation.

2.2 Supplementary Problems

Problem 1. Let x, y ∈ R3 and × be the vector product
 
x2 y3 − x3 y2
x × y =  x3 y1 − x1 y3  .
x1 y2 − x2 y1
36 Problems and Solutions

Hence the vector product is a map from R3 × R3 into R3 . Show that the
map is differentiable.

Problem 2. Study the map (Legendre map) f : R2 → R2

f1 (x1 , x2 ) = ex1 cos(x2 ), f2 (x1 , x2 ) = −ex1 sin(x2 ).

Find the functional determinant.

Problem 3. Consider the four symbols {A, B, C, D}. The Rudin-Shapiro

substitution is

A 7→ AC, B 7→ DC, C 7→ AB, D 7→ DB.

Find the sequence starting of with B.

Problem 4. Solve the difference equation

xt+1 = txt + t2 mod 2

with t = 0, 1, 2, . . . and x0 = 1.

Problem 5. Study the map f : Z × Z × Z → Z × Z × Z

f1 (x1 , x2 , x3 ) = x1 − x2 x3 , f2 (x1 , x2 , x3 ) = −x2 + x1 x3 ,

f3 (x1 , x2 , x3 ) = x3 − x1 x2 .
(i) Show that the fixed points are

(0, 0, 0), (1, 0, 0), (0, 0, 1), (−1, 0, 0), (0, 0, −1).

(ii) Show that (x1 = 2, x2 = 2, x3 = 2) provides a periodic orbit.

(iii) Show that (x1 = 1, x2 = 2, x3 = 1) provides an eventually periodic
orbit.
(iv) Is the map invertible?

Problem 6. Consider the map f : [0, 1]3 → [0, 1]3 given by

1 1
f1 (x1 , x2 , x3 ) = 2x1 , f2 (x1 , x2 , x3 ) = x2 , f3 (x1 , x2 , x3 ) = x3
2 2
for 0 ≤ x1 ≤ 1/2 and
1 1
f1 (x1 , x2 , x3 ) = 2x1 −1, f2 (x1 , x2 , x3 ) = (x2 +1), f3 (x1 , x2 , x3 ) = (x3 +1)
2 2
 Maps 37

for 1/2 < le ≤ 1. Find the orbit for x1 = x2 = x3 = 1/3.

Problem 7. Consider the map f : R2 → R2

f1 (x1 , x2 ) = sinh(x1 ) sin(x2 ), f2 (x1 , x2 ) = cosh(x1 ) cos(x2 ).

Find the fixed points of the map and study their stability.

Problem 8. Study the system of difference equations

x0,t+1 = 2x0,t − x1,t − x2,t − x3,t − x4,t

1
x1,t+1 = − x1,t + x0,t
2
1
x2,t+1 = − x2,t + x0,t
2
1
x3,t+1 = − x3,t + x0,t
2
1
x4,t+1 = − x4,t + x0,t
2
with the initial conditions

x0,0 = 1, x1,0 = x2,0 = x3,0 = x4,0 = 0.

Problem 9. Consider the difference equation

xt+1 = 2xt − 2, t = 0, 1, 2, . . .

where x0 = 1. Find the general solution. Show that x∗ = 2 is a particular

solution (the fixed point).

Problem 10. The Fibonacci numbers can be defined by

    
xt+2 1 1 xt+1
=
xt+1 1 0 xt

with x0 = x1 = 1. Find the eigenvalues of the 2 × 2 matrix. Study the

sequence
xt+4 xt+3
   
 xt+3   xt+2 
 = (A ⊗ A) 
xt+1 xt+1
 
xt xt
with x0 = x1 = x2 = x3 = 1 and ⊗ denotes the Kronecker product.
38 Problems and Solutions

Problem 11. Let x0 ∈ R+ and r ∈ R+ . Show that

 
1 r
xj+1 = xj + , j = 0, 1, 2, . . .
2 xj
√
tends to r.

Problem 12. Let A = (aj k) be a 3 × 3 matrix with det(A) 6= 0. Consider

the map f : R2 → R2 given by
a11 x1 + a12 x2 + a13 a21 x1 + a22 x2 + a23
f1 (x1 , x2 ) = , f2 (x1 , x2 ) = ,
a31 x1 + a32 x2 + a33 a31 x1 + a32 x2 + a33
Find the inverse of the map.

Problem 13. Consider the map f : R2 → R2

f1 (x1 , x2 ) = x1 + x2 , f2 (x1 , x2 ) = x1 x2

or written as difference equation

x1,τ +1 = x1,τ + x2,τ , x2,τ +1 = x1,τ x2,τ .

Find the fixed points of the map. Then solve the difference equation for
x1,0 = 1/4, x2,0 = 1/5.

Problem 14. Show that

 2  2
∂F (x, y) ∂F (x, y)
+
∂x ∂y
expressed in polar coordinates x = r cos(θ), y = r sin(θ) takes the form
 2  2
F (x(r, θ), y(r, θ) 1 ∂F (x(r, θ), y(r, θ))
+ .
∂r r ∂θ

Problem 15. Let r > 1 and τ = 0, 1, 2, . . .. Study the system of difference

equation
1
x1,τ +1 = ex2,τ , x2,τ +1 = re−x1,τ
r
with the initial conditions x1,0 = x2,0 = 0. Are there fixed points?

Problem 16. Let m ∈ N. Solve the recurrence relation

f (n) = 1 + f (n/2) for n ≥ 2, n = 2m

 Maps 39

with the initial condition f (1) = 1.

Problem 17. Study the initial value problem second order difference
equation
xτ +2 = xτ +1 + xτ − xτ +1 xτ
with x0 = 1, x1 = 1/2.

Problem 18. A set A is equilvalent to the set B (one writes A ∼ B) if

there is a mapping f : A → B which is (1, 1) and onto. Show that R2 ∼ R.
Utilize that
(·a1 a2 a3 . . . , ·b1 b2 b3 . . .) ↔ ·a1 b1 a2 b2 . . .
defines a (1, 1) map between pairs of decimal expansions and single expres-
sions of numbers in the interval (0, 1).

Problem 19. Let Ω = [0, 1), T (x) = 2x (mod 1), and µ the Lebesgue
measure. Show that T is ergodic.

Problem 20. (i) Let c > 0. Study the difference equation

xτ
xτ +1 = , τ = 0, 1, . . .
c + xτ
with x0 > 0. First find the fixed points.
(ii) Study the difference equation
6xτ
xτ +1 =
(1 + xτ )2
with x0 > 0. First find the fixed points.

Problem 21. Let x ∈ [0, 1] and bac be the integer part of a. Show that
the sequence xt (t = 0, 1, . . .) given by
1
x0 = 0, and xt+1 = xt + b2t+1 (x − xt )c
2t+1
converges to x.

Problem 22. The Kustaanheimo-Stiefel transformation is defined by the

map from R4 (coordinates u1 , u2 , u3 , u4 ) to R3 (coordinates x1 , x2 , x3 )

x1 (u1 , u2 , u3 , u4 ) = 2(u1 u3 − u2 u4 )
x2 (u1 , u2 , u3 , u4 ) = 2(u1 u4 + u2 u3 )
x3 (u1 , u2 , u3 , u4 ) = u21 + u22 − u23 − u24
40 Problems and Solutions

together with the constraint

u2 du1 − u1 du2 − u4 du3 + u3 du4 = 0.
(i) Show that
r2 = x21 + x22 + x23 = u21 + u22 + u23 + u24 .
(ii) Show that
1 1
∆3 = ∆4 − 2 V 2
4r 4r
where
∂2 ∂2 ∂2 ∂2 ∂2 ∂2 ∂2
∆3 = 2 + 2 + , ∆4 = + + +
∂x1 ∂x2 ∂x23 2
∂u1 2
∂u2 2
∂u3 ∂u24
and V is the vector field
∂ ∂ ∂ ∂
V = u2 − u1 − u4 + u3
∂u1 ∂u2 ∂u3 ∂u4
(iii) Consider the differential one form
α = u2 du1 − u1 du2 − u4 du3 + u3 du4 .
Find dα. Find LV α, where LV (.) denotes the Lie derivative.
(iv) Let g(x1 (u1 , u2 , u3 , u4 ), x2 (u1 , u2 , u3 , u4 ), x3 (u1 , u2 , u3 , u4 )) be a smooth
function. Show that LV g = 0.

Problem 23. The Kustaanheimo-Stiefel map R4 → R3 ((u1 , u2 , u3 , u4 ) →

(x1 , x2 , x3 )) is given by
x1 = 2(u1 u3 − u2 u4 ), x2 = 2(u1 u4 + u2 u3 ), x3 = u21 + u22 − u23 − u24
together with the constraint
α ≡ u2 du1 − u1 du2 − u4 du3 + u3 du4 = 0.
Show that applying the Kustaanheimo-Stiefel map the Laplacian operator
∆3 in R3 can be written as
1 1
∆3 = ∆4 − 2 V 2
4r 4r
where
∂ ∂ ∂ ∂
r = (x21 +x22 +x23 )1/2 = u21 +u22 +u23 +u24 , V = u2 −u1 −u4 +u3 .
∂u1 ∂u2 ∂u3 ∂u4
Chapter 3

Functions

3.1 Solved Problems

Problem 1. Let fj : Rn → R, j = 1, 2, . . . , n be real-valued functions
with continuous second-order partial derivatives everywhere on Rn . Sup-
pose that there are constants cij such that
∂fi ∂fj
− = cij
∂xj ∂xi
for all i and j, 1 ≤ i ≤ n. Prove that there is a function g : Rn → R
such that fi + ∂g/∂xi is linear for all i = 1, 2, . . . , n. A linear function
p : Rn → R is of the form
p(x) = a0 + a1 x + a2 x2 + · · · + an xn .

Problem 2. Let f : Rn → R be a differentiable function and x ∈ Rn .

Consider the invertible n × n matrix R and the transformation x0 = Rx.
The transformation operator P̂R associated with the invertible matrix R is
defined by
P̂R f (x) := f (R−1 x).
Note that the operator acts upon the coordinates x and not on the argument
of f . This means P̂R P̂S f (x) = f (S −1 R−1 x), where S is another invertible
n × n matrix. Show that
P̂R P̂S f (x) = P̂RS f (x).

41
42 Problems and Solutions

Problem 3. A ratio list is a finite list of positive numbers, (r1 , r2 , . . . , rn ).

An iterated function system realizing a ratio list (r1 , r2 , . . . , rn ) in a metric
space S is a list (f1 , f2 , . . . , fn ), where fj : S → S is a similarity with
ratio rj . A nonempty compact set K ⊆ S is an invariant set for the
iterated function system (f1 , f2 , . . . , fn ) iff K = f1 (K)∪f2 (K)∪· · ·∪fn (K).
The triadic Cantor set is an invariant set for an iterated function system
realizing the ratio list (1/3, 1/3). The Sierpinski gasket is an invariant set
for an iterated system realizing the ratio list (1/2, 1/2, 1/2). The dimension
associated with a ratio list (r1 , r2 , . . . , rn ) is the positive number s such that
r1s + r2s + · · · + rns = 1. Let (r1 , r2 , . . . , rn ) be a ratio list. Suppose each
rj < 1. Show that there is a unique nonnegative number s satisfying
n
X n
X
rjs ≡ es ln(rj ) = 1.
j=1 j=1

The number s is 0 iff n = 1.

Problem 4. The arc length of the equilateral hyperbola

p
h(t) = t2 − 1, t≥1

starting at t = 1 is given by
x
r
2t2 − 1
Z
Lh (x) = dt
1 t2 − 1

as a function of the terminal point t = x. The tangent line to the hyperbola

at t = x is p x
Th (t) = x2 − 1 + √ (t − x)
2
x −1
whose intersection with the t-axis is t = 1/x (t ∈ (0, 1)). The line
√
x2 − 1
Nh (t) = − t
x
is perpendicular to Th passing through the origin.
(i) Find the point Ph where the lines Th and Nh intersect.
(ii) Calculate the distance from (x, h(x)) to the common point Ph .

Problem 5. Given a differentiable function f . The logarithmic derivative

of f is defined as
1 df
.
f dx
 Functions 43

When f is a real function of real variables and takes strictly positive values
then the chain rule provides

d 1 df
ln(f (x)) = .
dx f dx

The logarithmic derivative has the following properties.

1. The logarithmic derivative of the product of functions is the sum of their
logarithmic derivatives

1 d 1 df 1 dg
(f g) = +
f g dx f dx g dx

2. The logarithmic derivative of the quotient of functions is the difference

of their logarithmic derivatives

1 d 1 df 1 dg
(f /g) = −
f /g dx f dx g dx

3. The logarithmic derivative of the α-th power of a function f is α times

the logarithmic derivative of the function

1 d α 1 df
(f ) = α
f α dx f dx

4. The logarithmic derivative of the exponential of a function equals the

derivative of a function
1 d f df
e =
ef dx dx
(i) Find the logarithmic derivative of f : R → R, f (x) = cosh(x).
(ii) Let f be a meromorphic function in the open and conncted set D ⊆ C.
Let G ⊆ D be a region such that its closure G ⊆ D and its boundary ∂G is
a continuous curve not containing a zero or pole of f . Let N be the number
of zeros of f lying inside G and P be the number of poles of f lying inside
G. Then Z
1 1 df (z)
N −P = dz
2πi ∂G f (z) dz
where ∂G is an oriented boundary of G. Calculate the left and right-hand
side for the function f (z) = 1/z and G = { (x, y) : x2 + y 2 ≤ 1 }.

Problem 6. Let f be an analytic function. Let p ∈ N and α ∈ R. Show

that p
αx2 dp
     2 
d αx
+ αx f ≡ exp − exp f .
dx 2 dxp 2
44 Problems and Solutions

Problem 7. Consider

1 if 0 ≤ t ≤ s
K(t, s) = .
0 if s ≤ t ≤ 1

Show that this kernel satisfies the functional equation

 
t−a
K(cu + a, t) = K u,
c

where c > 0.

Problem 8. Find a continuous differentiable function f : R → R which

has no fixed points and no critical points.

Problem 9. Let a, b, p, q, r ∈ R, b > a, a < p, p < r, r < q, q < b and

p − a = r − p, q − r = b − q.

In fuzzy logic the following membership function plays an important role



 0 x≤a
 m−1 m


 2 ((x − a)/(r − a)) a<x≤p
1 − 2m−1 ((r − x)/(r − a))m p < x ≤ r

f (x; a, b, r, p, q) =

 1 − 2m−1 ((x − r)/(b − r))m r < x ≤ q
2m−1 ((b − x)/(b − r))m q<x<b




0 x≥b


where m is the fuzzifier. Im most cases m = 2. Where are the crossover

points? What is the value at the centre?

Problem 10. Let f : R2 → R2 be the mapping of (x, y) → (u, v) given

by
u(x, y) = ex cos(y), v(x, y) = ex sin(y).

What is the range of f ? Show that the Jacobian determinant is not zero at
any point of R2 . Thus every pont of R2 has a neighbourhood in which f is
one-to-one. Nevertheless f is not one-to-one on R2 . What are the images
under f of lines parallel to the coordinate axes?

Problem 11. Let x, y ∈ R. Calculate the square root

p
x2 + y 2 − 2xy.
 Functions 45

Problem 12. A criterion for linearly independence of functions f0 , f1 , . . . , fn ∈

C n [a, b] is that the Wronski determinant is nonzero
 f f ... f 
0 1 n
 f00 f10 ... fn0 
det 
 .. .. .. =
..  6 0.
. . . .
(n) (n) (n)
f0 f1 ... fn
Here 0 denotes derivative. Apply this criterion to the functions

f0 (x) = cos(x), f1 (x) = sin(x).

Problem 13. For a sphere of radius r and mass density ρ the mass that
must be concentrated at its centre is (λ ≥ 0)
4πrρ
M (λ) = (cosh(λr) − sinh(λr)/(λr)).
λ2
Find limλ→0 M (λ).

Problem 14. Let −1 < a < 1. Find the inverse of the transformation
z−a
λ(z) = .
1 − az

Problem 15. Consider the Jacobi elliptic functions

sn(x, k), cn(x, k), dn(x, k)

where x ∈ R and k 2 ∈ [0, 1]. We have

sn(x, 0) = sin(x), cn(x, 0) = cos(x), dn(x, 0) = 1

and
ex − e−x 2
sn(x, 1) = = tanh(x), cn(x, 1) = dn(x, 1) = ≡ sech(x).
ex + e−x ex + e−x
(i) Find an expression using Jacobi elliptic functions that interpolates be-
tween sin(x) for k = 0 and sinh(x) for k = 1.
(ii) Find an expression using Jacobi elliptic functions that interpolates be-
tween cos(x) for k = 0 and cosh(x) for k = 1.
(iii) Use this result to interpolate between the matrices
   
cos(x) sin(x) cosh(x) sinh(x)
and .
− sin(x) cos(x) sinh(x) cosh(x)
46 Problems and Solutions

Problem 16. Let f1 , f2 be differentiable functions and f1 (x) > 0, f2 (x) >
0 for all x. Let f (x) = f1 (x)f2 (x). Find

d
(ln(f (x)).
dx

Problem 17. Let x ∈ R. The sequence of functions { fk (x) } is defined

by f1 (x) = cos(x/2) and for k > 1 by

fk (x) = fk−1 (x) cos(x/2k ).

Thus
fk (x) = cos(x/2) cos(x/22 ) · · · cos(x/2k ).
Obviously, we have fk (0) = 1 for every k. Calculate limk→∞ fk (x) as a
function of x for x 6= 0.

Problem 18. (i) Consider the transformation in R3

x0 (a, θ1 ) = cosh(a)
x1 (a, θ1 ) = sinh(a) sin(θ1 )
x2 (a, θ1 ) = sinh(a) cos(θ1 )

where a ≥ 0 and 0 ≤ θ1 < 2π. Find

x20 − x21 − x22 .

(ii) Consider the transformation in R4

x0 (a, θ1 , θ2 ) = cosh(a)
x1 (a, θ1 , θ2 ) = sinh(a) sin(θ2 ) sin(θ1 )
x2 (a, θ1 , θ2 ) = sinh(a) sin(θ2 ) cos(θ1 )
x3 (a, θ1 , θ2 ) = sinh(a) cos(θ2 )

where a ≥ 0, 0 ≤ θ1 < 2π and 0 ≤ θ2 ≤ π. Find

x20 − x21 − x22 − x23 .

Extend the transformation to Rn .

Problem 19. A fixed charge Q is located on the z-axis with coordinates

ra = (0, 0, d/2), where d is interfocal distance of the prolate spheroidal
 Functions 47

coordinates
1
x(η, ξ, φ) = d((1 − η 2 )(ξ 2 − 1))1/2 cos φ
2
1
y(η, ξ, φ) = d((1 − η 2 )(ξ 2 − 1))1/2 sin(φ)
2
1
z(η, ξ, φ) = dηξ
2
where −1 ≤ η ≤ +1, 1 ≤ ξ ≤ ∞, 0 ≤ φ ≤ 2π. Express the Coulomb
potential
Q
V =
|r − ra |
in prolate spheroidal coordinates.

Problem 20. Toroidal coordinates are defined by

c sinh(α) cos(φ)
x1 (α, β, φ) =
cosh(α) − cos(β)
c sinh(α) sin(φ)
x2 (α, β, φ) =
cosh(α) − cos(β)
c sin(β)
x3 (α, β, φ) = .
cosh(α) − cos(β)

Use the L’Hosptial rule to find x1 , x2 , x3 for α → 0 and β → 0.

Problem 21. Let c, θ ∈ R and

f (θ) = c(eiθ + e−iθ ).

Calculate exp(f (θ)).

Problem 22. A continuous function f : R2 → R is called an alternating

function if
f (x, y) = −f (y, x).
Give an example of an analytic alternating function. Find the minima and
maxima of the function.

Problem 23. Consider the functions

1
j1 (x) = (sin(x) − x cos(x))
x2
1
j2 (x) = 3 ((3 − x2 ) sin(x) − 3x cos(x)).
x
48 Problems and Solutions

Use the L’Hospital rule to find j1 (0) and j2 (0).

Problem 24. Find the invariance group of the function f : R → R

f (x) = sin(x).

Problem 25. Let a > 0. Consider the transformation

a sin(2ax) a sinh(2ay)
u(x, y) = , v(x, y) = .
2(sin2 (ax) + sinh2 (ay)) 2(sin2 (ax) + sinh2 (ay))
Find the inverse transformation.

Problem 26. (i) Find an analytic function f : R → R that has no fixed

point and no critical point. Draw the function.
(ii) Find an analytic function f : R → R that has no fixed point and exactly
one critical point at x = 0. Draw the function.

Problem 27. Let f : R → R be an analytic function. Calculate the

commutator  
d d
cos(x) , sin(x) f.
dx dx

Problem 28. (i) Consider the analytic function f : R2 → R2

f1 (x1 , x2 ) = sinh(x2 ), f2 (x1 , x2 ) = sinh(x1 ).

Show that this function admits the (only) fixed point (0, 0). Find the
functional matrix at the fixed point
 
∂f1 /∂x1 ∂f1 /∂x2 
.
∂f2 /∂x1 ∂f2 /∂x2  (0,0)

(ii) Consider the analytic function g : R2 → R2

g1 (x1 , x2 ) = sinh(x1 ), g2 (x1 , x2 ) = − sinh(x2 ).

Show that this function admits the (only) fixed point (0, 0). Find the
functional matrix at the fixed point
 
∂g1 /∂x1 ∂g1 /∂x2 
.
∂g2 /∂x1 ∂g2 /∂x2 (0,0)

(iii) Multiply the two matrices found in (i) and (ii).

 Functions 49

(iv) Find the composite function h : R2 → R2

h(x) = (f ◦ g)(x) = f (g(x)).
Show that this function also admits the fixed point (0, 0). Find the func-
tional matrix at this fixed point
 
∂h1 /∂x1 ∂h1 /∂x2 
.
∂h2 /∂x1 ∂h2 /∂x2 (0,0)

Compare this matrix with the matrix found in (iii).

Problem 29. An approximation of e−x (x ≥ 0) as a rational polynomial

using a 3rd order Padé approximation is given by
1 − x/2 + x2 /10 − x3 /120
e−x ≈ .
1 + x/2 + x2 /10 + x3 /120
Note that e−x ≥ 0 for all x. For which value of x > 0 does the right-hand
side takes negative values?

Problem 30. Let x ∈ R. Consider the function

x−i
f (x) = .
x+i
Find f (0), f (1), f (−1) and f (x → ∞). Is there an inverse?

Problem 31. Consider the function

f (x) = x3 − x2 .
Find an approximation of the derivative of f by using
f (x − 2h) − 8f (x − h) + 8f (x + h) − f (x + 2h) 1
f 0 (x) ≈ − h4 f (5) (ξ)
12h 30
−3f (x − h) − 10f (x) + 18f (x + h) − 6f (x + 2h) + f (x + 3h) 1 4 (5)
f 0 (x) ≈ + h f (ξ)
12h 20
−25f (x) + 48f (x + h) − 36f (x + 2h) + 16f (x + 3h) − 3f (x + 4h) 1
f 0 (x) ≈ − h4 f (5) (ξ)
12h 5
where x ≤ ξ ≤ x + h.

Problem 32. Let z ∈ C. The Airy functions Ai(z) and Bi(z) are defined
as sums of the power series
Ai(z) = c1 f (z) − c2 g(z)
√
Bi(z) = 3(c1 f (z) + c2 g(z))
50 Problems and Solutions

where
1 3 1·4 6 1·4·7 9
f (z) = 1 + z + z + z + ···
3! 6! 9!
2 2 · 5 7 2 · 5 · 8 10
g(z) = z + z 4 + z + z + ···
4! 7! 10!
with    
1 1 1 2
c1 = Γ 3−1/6 , c2 = Γ 31/6 .
2π 3 2π 3
Show that the radius of convergence of these infinite series is infinite.

Problem 33. Find the inverse of the function y = tanh(x). Note that
y ∈ (−1, 1).

Problem 34. Consider the analytic function f : R → R

∞
X xj ex − 1
f (x) = ≡ .
j=0
(j + 1)! x

Find the fixed points and critical points of the function. Note that f (0) = 1.

Problem 35. Find non-negative analytic functions f : R → R such that

f (0) = 0, f (1) = 1, f (2) = 0.

Problem 36. The Euler dilogarithm function Li2 (x) is defined for x ∈
(0, 1) as
∞ x
xj ln(1 − t)
X Z
Li2 (x) = =− dt.
j=1
j2 0 t

It can be analytically continued to the complex plane with the branch cut
from 1 to ∞ along the real axis using the integral. Calculate Li2 (1/2).

Problem 37. Let Ai be the Airy function and Jn , In are the Bessel and
modified Bessel functions of order n. Show that
d2 Ai
= xAi(x)
dx2
1
Ai(x) = x1/2 (I−1/3 (z) − I1/3 (z))
3
1
Ai(−x) = x1/2 (J1/3 (z) + J−1/3 (z))
3
 Functions 51

where z := 2x3/2 /3.

Problem 38. Find a continuous function f : R → R such that

f (x) = f (−x), f (x) = f (x + 2π), f (0) = 0.

Problem 39. Let N be an integer and N ≥ 2. The generalized hyperbolic

function for a given N is defined as
∞
(N )
X xj+kN
fj (x) := , j = 0, 1, . . . , N − 1.
(j + kN )!
k=0

(N )
The functions fj are analytic.
(i) Show that
(N ) (N )
fj (x) = fj+N (x)
(N )
i.e. fj is periodic with respect to the index j.
(ii) Show that
d (N ) (N )
f (x) = fj−1 (x).
dx j
(iii) Let ω N = 1, i.e. ω is the N -th primitive root of unity. Show that
N −1
(N ) 1 X −jk
fj (x) = ω exp(ω k x).
N
k=0

Problem 40. Let n = 2, 3, 4, . . .. Find

Γ(n/2)
Γ(n)
where Γ denotes the gamma function.

Problem 41. Consider the hyperspherical coordinates

x1 (θ, φ, ψ) = cos(θ)
x2 (θ, φ, ψ) = sin(θ) cos(φ)
x3 (θ, φ, ψ) = sin(θ) sin(φ) cos(ψ)
x4 (θ, φ, ψ) = sin(θ) sin(φ) sin(ψ)

with x21 +x22 +x23 +x24 = 1. Show that the angular distance can be calculated
as
cos(djk ) =
52 Problems and Solutions

cos(θj ) cos(θk )+sin(θj ) sin(θk )(cos(φj ) cos(φk )+sin(φj ) sin(φk ) cos(ψj −ψk )).

Problem 42. Consider the function

f (x) = 1 + x + 2x2 + 3x3 .

Find the second order derivative of f at a = 1 applying

f (a + ) − 2f (a) + f (a − )
f 00 (a) = lim .
→0 2

Problem 43. Let −1 < x < 1 and

∞
X 1
aj xj = √ .
j=0
1−x

Find the expansion coefficients aj .

Problem 44. Show that the function

f (x) = cos(x) cosh(x) + 1

has infinitely many roots, i.e. solutions of f (x) = 0. What happens for x
large?

Problem 45. The Jacobi elliptic functions sn(x, k), cn(x, k), dn(x, k)
with k ∈ [0, 1] and k 2 + k 02 = 1 have the properties

sn(x, 0) = sin(x), cn(x, 0) = cos(x), dn(x, 0) = 1

and
ex − e−x 2
sn(x, 1) = , cn(x, 1) = = dn(x, 1).
ex + e−x ex + e−x
We define

u1 (x, y, k, k 0 ) = sn(x, k)dn(y, k 0 )

u2 (x, y, k, k 0 ) = cn(x, k)cn(y, k 0 )
u3 (x, y, k, k 0 ) = dn(x, k)sn(y, k 0 ).

(i) Find u1 (x, y, 0, 1), u2 (x, y, 0, 1), u3 (x, y, 0, 1) and calculate u21 (x, y, 0, 1)+
u22 (x, y, 0, 1) + u23 (x, y, 0, 1).
(ii) Find u1 (x, y, 1, 0), u2 (x, y, 1, 0), u3 (x, y, 1, 0) and calculate u21 (x, y, 1, 0)+
u22 (x, y, 1, 0) + u23 (x, y, 1, 0).
 Functions 53

Problem 46. Let m be an integer with m ≥ 1. Then the gamma function

Γ is given by
Γ(m) = (m − 1)!.
Thus Γ(1) = Γ(2) = 1, Γ(3) = 2, Γ(4) = 6, Γ(5) = 24. Furthermore

1 1 · 3 · 5 · · · · · (2m − 1) √
Γ(m + ) = π.
2 2m
√ √ √
Thus
√ Γ(3/2) = π/2, Γ(5/2) = 3 π/4, Γ(7/2) = 15 π/8, Γ(9/2) =
105 π/16. Calculate
Γ n2 − 21

Γ n2

for m = 5, m = 10, m = 20.

Problem 47. Let c1 > 0, c2 > 0. Consider the function f : R → R

c1 x 2
f (x) = x2 + .
1 + c2 x2
Find the minima and maxima. Find the fixed points.

Problem 48. Let c > 0. Consider the function f : R → R

4  cx 
fc (x) = 2 sin2 .
x 2
Find fc (0). Show that fc has a maximum at x = 0.

Problem 49. (i) Find the minima and maxima of the analytic function
f :R→R
f (x) = cosh(x) cos(x).
Find the fixed points.
(ii) Find the minima and maxima of the analytic function f : R → R

f (x) = sinh(x) sin(x).

Find the fixed points.

Problem 50. Let S1 and S2 be two finite sets with n1 and n2 elements,
respectively. The number of functions f : S1 → S2 is given by

nn2 1 .

(i) Let n1 = n2 = 2. Find all possible functions.

54 Problems and Solutions

(ii) Let n1 = n2 = n. Then there are n! one-to-one functions f : S1 → S2 .

Let n = 3. Find all one-to-one functions.

Problem 51. Let µ > 0 and

x01
   
x1
R =  x2  , R0 =  x02  .
x3 x03
(i) Show that
exp(−|R − R0 |/µ)
=
|R − R0 |
∞ Z ∞
X 1 p
n cos(n(φ−φ0 )) Jn (kr)J(kr0 ) exp(− k 2 + 1/µ2 |x3 −x03 |)kdk
n=0 0 k2 + 1/µ
where εn = 1 for n = 0 and εn = 2 for n > 0 and Jm (kr) is ordinary Bessel
function of order m.
(ii) Consider the functions
√
k
fs,k,n (R) = Jn (kr)einφ+isx3
2π
where 0 ≤ k < ∞, −∞ < s < ∞, and n = 0, ±1, ±2, . . .. Show that
Z 2π Z +∞ Z ∞
dx3 fs,k,n (R)f¯s0 ,k0 ,n0 (R)rdr = δnn0 δ(s − s0 )δ(k − k 0 ).
0 −∞ 0

Problem 52. Let f1 , f2 : R → R be analytic functions and

1
g1 (x) = (f1 (x) + f2 (x) − |f1 (x) − f2 (x)|
2
1
g2 (x) = (f1 (x) + f2 (x) + |f1 (x) − f2 (x)|.
2
Let f1 (x) = sin(x) and f2 (x) = cos(x). Find the maxima and minima of
g1 . Find the maxima and minima of g2 .

Problem 53. Let the function f be continuous on a closed interval [a, b]

and is continuous differentiable in the open interval (a, b). Then there exists
a point c in (a, b) such that (mean value theorem)
f (b) − f (a) = f 0 (c)(b − a).
Apply the theorem to the function
f (x) = 1 + 2x + 3x2
 Functions 55

and the interval [−1, 1].

Problem 54. The plane {(x1 , x2 , x3 ) : x1 + 21 x2 + 13 x3 = 1} and the

cylinder {(x1 , x2 , x3 ) : x21 + x22 = 4} intersect. Find the shortest distance
from (0, 0, 0) to this curve.

Problem 55. Consider the plane {(x1 , x2 , x3 ) : 2x1 + 3x2 − x3 = 5} in

the three dimensional Euclidean space E3 . Find the shortest distance from
(0, 0, 0) to this plane.

Problem 56. We denotes by Q, R and C the rational, real and complex

fields, respectively. Let B1 and B2 be Banach spaces. A map f : B1 → B2
is said to be additive if and only if

f (x1 + x2 ) = f (x1 ) + f (x2 )

for all x1 , x2 ∈ B1 . Show that f is additive implies that f (qx) = qf (x) for
all q ∈ Q. Show that f is additive and continuous implies that f (rx) =
rf (x) for all r ∈ R.

Problem 57. Let a1 d1 − b1 c1 6= 0, a2 d2 − b2 c2 6= 0 and the functions

f1 : C → C
a1 z + b1 a2 z + b2
f1 (z) = , f2 (z) = .
c1 z + d1 c2 z + d2
Find f2 ◦ f1 , i.e. the function compositon.

Problem 58. Let x1 , x2 ≥ 0. Consider

√ ∞ X ∞
1 + x1 X
f (x1 , x2 ) = √ √ = F (j1 , j2 )xj11 xj22 .
1 + x1 1 + x2 j =0 j =0
1 2

Find the expansion coefficients.

Problem 59. Consider the function

3
f (x) = (sin(x) − x cos(x)).
x3
Find f (0) using the L’Hospital rule.

Problem 60. Let

sin(x)
f (x) = .
sinh(x)
56 Problems and Solutions

Find f (0).

Problem 61. Consider the function f : R → R

f (x) = x3 .
Show that f is not convex. Show that f is convex if we restrict to the
domain x ≥ 0.

Problem 62. Consider the convex functions f : R → R and g : R → R.

Show that f + g is convex. Show that max{ f, g } is convex.

Problem 63. Definition (Convex Set). A subset C of Rn is said to

be convex if for any a and b in C and any θ in R, 0 ≤ θ ≤ 1, the n-tuple
θa + (1 − θ)b also belongs to C. In other words, if a and b are in C then
{ θa + (1 − θ)b : 0 ≤ θ ≤ 1 } ⊂ C.
Definition (Convex Polyhedron). Let a1 , . . . , ap be p points in Rn . The
n-tuple
Xp Xp
θj aj , θj ≥ 0, j = 1, . . . , p, θj = 1
j=1 j=1

is called a convex combination (or a convex sum) of a1 , . . . , ap . If X ⊂ Rn

then the set of all (finite) convex combination of points of X is called the
convex hull of X and is denoted by H(X). If X is finite, X = {a1 , . . . , ap },
then H(X) is called the convex polyhedron spanned by a1 , . . . , ap and is
also denoted by H(a1 , . . . , ap ).

Defintion. Let S be a nonempty convex set in Rn , where Rn is the n-

dimensional Euclidean space. The function f : S → R is said to be convex
if
f (λx1 + (1 − λ)x2 ) ≤ λf (x1 ) + (1 − λ)f (x2 )
for each x1 , x2 ∈ S and for each λ ∈ [0, 1]. The function is said to be
strictly convex if the above inequality holds as a strict inequality for each
distinct x1 , x2 ∈ S and for each λ ∈ (0, 1).

Let
a1 = (1, −1), a2 = (2, 2), a3 = (3, 1)
2
be points of R . Find the convex polyhedron.

Problem 64. Show that the function φ : [0, ∞) → R defined by


0 if x = 0
φ(x) = (1)
x ln x if x 6= 0
 Functions 57

is strictly convex, i.e.,

φ(αx + βy) ≤ αφ(x) + βφ(y) (2)

if x, y ∈ [0, ∞), αβ ≥ 0, α + β = 1 with equality only when x = y or α = 0

or β = 0. By induction we get
k
! k
X X
φ α i xi ≤ αi φ(xi ) (3)
i=1 i=1

if
k
X
xi ∈ [0, ∞), αi ≥ 0, αi = 1. (4)
i=1

Equality holds only when all the xi , corresponding to non-zero αi , are equal.

Problem 65. Let x ≥ 0 and

f (x) = x ln(x) − x + 1. (1)

(i) Show that

f (x) ≥ 0. (2)
(ii) Show that from L’Hospital rule we find that f (0) = 1.
(iii) Show that the function is convex.

Problem 66. Show that f : R → R

f (x) = |x| (1)

is convex.

Problem 67. Let f : R2 → R be defined as

f (x, y) = ax2 + by 2 + 2cxy + d

where a, b, c, d ∈ R. For what values of a, b, c, d is f concave?

Problem 68. Let r1 , r2 ∈ R. Consider the map f : R2 → R2

f1 (x1 , x2 ) = r1 x1 (1 − x1 − x2 ), f2 (x1 , x2 ) = r2 x1 x2 .

Find df1 , df2 , df1 ∧ df2 .

√
Problem 69. Find x = arctan(−1/ 3).
58 Problems and Solutions

Problem 70. Let f, g : R → R be analytic functions. Find

d2 −f (x)
ef (x) (e g(x)).
dx2

Problem 71. Consider the map f : R2 → R2

f1 (x1 , x2 ) = sin(x21 + x22 ), f2 (x1 , x2 ) = cos(x21 + x22 ).
Find the Jacobian matrix and the Jacobian determinant.

Problem 72. Consider the map f : R → R

1 1
f (x) = − x2 − x + .
2 2
Then f (1) = −1, f (−1) = 1. Thus we have a periodic orbit with period 2.
Is the orbit attracting or repelling? We set x1 = 1, x2 = −1. We have to
test
|(df (x = x1 )/dx)(df (x = x2 )/dx)| < 1
for attracting and
|(df (x = x1 )/dx)(df (x = x2 )/dx)| > 1
for repelling.

Problem 73. Let α ∈ R and f (x) = eαx . Find

1
(f (x + h) − 2f (x) + f (x − h)).
h2
Then consider h → 0.

Problem 74. Consider the analytic function f : R → R, f (x) = cos(x).

The equation cos(x) = x has one solution, i.e. we have one fixed point
for f . Consider f (f (x)) = cos(cos(x)). Does the equation cos(cos(x)) = x
admits other solutions besides the one of cos(x) = x.

Problem 75. Let a, b ∈ N. Consider the 2 × 2 matrix

 
1 a
C=
b 1 + ab
with det(C) = 1. Consider the map (generalized Arnold cat map)
   
x1,τ +1 x1,τ
=C mod 1
x2,τ +1 x2,τ
 Functions 59

where τ = 0, 1, 2, . . .. Find the eigenvalues of C which relates to the one-

dimensional Liapunov exponents.

Problem 76. Consider the functions f : R → R, g : R → R

f (x) = 2x, g(x) = 3x2 − 1.

Find (f ◦ g)(x) = f (g(x)), (g ◦ f )(x) = g(f (x)) and the extrema of these
functions.

Problem 77. Let a ∈ R. Given the map f : R2 → R2

f (x1 , x2 ) = (x1 + x2 , 2x1 + ax2 ).

(i) Find the fixed points of the map.

(ii) Find the matrix Df (x1 , x2 ) and show that the matrix Df (x1 , x2 ) is
invertible if and only if a 6= 2.

Problem 78. Let A, B be non-empty sets. If f : A → B is (1,1) and

onto (i.e. bijective) one can define the inverse function f −1 : B → A as the
unique function from B to A such that

(f ◦ f −1 )(y) = y for all y ∈ B

(f −1 ◦ f )(x) = x for all x ∈ A.

The function f : (0, 1) → R

x − 1/2
x 7→ f (x) =
x(x − 1)

is bijective. Find f −1 . Note that f (1/2) = 0, f (1/4) = 4/3, f (3/4) = −4/3.

Problem 79. Give non-trivial functions f : R → R which have no critical

points and satisfiy f (0) = 0. Trivial means f (x) = cx, where c is a constant.

Programming Problems

Problem 1. Let f1 , f2 , f3 : R3 → R be continuously differentiable func-

tion. Find the determinant of the 3 × 3 matrix A = (ajk )

∂fj ∂fk
ajk := − .
∂xk ∂xj
60 Problems and Solutions

Apply computer algebra.

Problem 2. Consider the map f : R → R given by

f (x) = x2 + x + 1.

Find the minima and maxima of

g(x) = f (f (x)) − f (x)f (x).

Apply computer algebra.

Problem 3. Consider the polynomials

f (x) = x2 + 2x + 1, g(x) = x + 4;

Find (f ◦ g)(x) = f (g(x)), (g ◦ f )(x) = g(f (x)) and f (g(x)) − g(f (x)).

Problem 4. Let f : R → R be an analatic function. The Schwarzian

derivative is defined by
2
f 000 (x) 3 f 00 (x)

S(f (x)) := 0 −
f (x) 2 f 0 (x)

where 0 denotes the derivative with respect to x. Let a, b, c, d ∈ R with

ad − bc 6= 0 and
af (x) + b
g(x) = .
cf (x) + d
Show that S(g(x)) = S(f (x)). Apply computer algebra.

3.2 Supplementary Problems

Problem 1. Find √ √
1+x− 1−x
lim .
x→0 x

Problem 2. Consider the analytic function f : R2 → R

xex (ey − 1) − yey (ex − 1)

f (x, y) = .
xy(ex − ey )
 Functions 61

1
(i) Show that f (0, 0) = 2 applying the L’Hospital rule.
(ii) Show that
ex − x − 1
f (x, x) = .
x2
(iii) Show that
tanh(x/2)
f (x, −x) = .
x

Problem 3. Consider the map from R4 to the vector space of 2 × 2

matrices over R
 
x1 + x4 x2 + x3
f (x1 , x2 , x3 , x4 ) = .
−x2 + x3 x1 − x4

Find the inverse of the map. Note that the determinant of the matrix is
given by x21 + x22 − x23 − x24 .

Problem 4. Let

D = { (x1 , x2 , x3 ) : x21 + x22 + x23 = 1, x1 ≥ 0, x2 ≥ 0, x3 ≥ 0 }.

Let v be a normalized vector in R3 with nonnegative entries and A be a 3×3

matrix over R with strictly positive entries. Show that the map f : D → D
Av
f (v) =
kAvk

has a fixed point, i.e. there is a normalized vector v0 such that

Av0
= v0 .
kAv0 k

Problem 5. Let f, g : R → R be analytic functions. Find

d2 −f (x)
(ef (x) e )x.
dx2

Problem 6. Let Z be the set of integers and N0 be the set of natural

numbers including 0. Consider the one-to-one map f : Z → N0

 0 if j=0
f (j) = 2j − 1 if j is positive .
−2j if j is negative

62 Problems and Solutions

Give a C++ implementation with the class Verylong of the function f and
its inverse.

Problem 7. Let { f1 , f2 , . . . , fn } be a set of convex functions from Rn →

R. Show that the nonnegative linear combination
f (x) = α1 f1 (x) + α2 f2 (x) + · · · + αn fn (x), α1 , α2 , . . . , αn ≥ 0
is convex.

Problem 8. Study the map f : R2 → R2

p p
f1 (x1 , x2 ) = 1 − sin(x2 ), f2 (x1 , x2 ) = 1 + sin(x1 ).
First find the fixed points if there are any.

Problem 9. Let k ∈ N and x ≥ 0. Find the fixed points of the map

√
fk (x) = k + x
and study their stability.

Problem 10. Consider the continuous function f : R+ → R+


2x x<1
f (x) =
max{4 − 2x, 14 } x ≥ 1
Find the fixed points and study their stability.

Problem 11. Let

D = {(x1 , x2 , x3 ) : x21 + x22 + x23 = 1, x1 ≥ 0, x2 ≥ 0, x3 ≥ 0}.
Let v be a normalized vector in R3 with nonnegative entries and A be a
3 × 3 matrix over R with strictely positive entries. Show that the map
f :D→D
Av
f (v) =
kAvk
has fixed point, i.e. there is a v0 such that
Av0
= v0 .
kAv0 k

Problem 12. Let f, g : R → R be two analytic functions and S be the

Schwarzian derivative. Show that if Sf < 0 and Sg < 0, then
S(f ◦ g) < 0.
 Functions 63

Problem 13. Let x, y ∈ R+ . Consider the function f : (R+ )2 → R

√
1+x
f (x, y) = √ √ .
1+x+ 1+x

Then f (0, 0) = 1/2 and f (x, y) + f (y, x) = 1. Set

√ ∞ X ∞
1+x X
√ √ = F (m.n)xm y n .
1 + x + 1 + x m=0 n=0

Show that F (m, n) = −F (n, m) unless m = n = 0 and F (m, m) = 0 for

m ≥ 1. Show that

F (0, 0) = 1/2, F (0, 1) = −F (1, 0) = −1/8, F (0, 2) = −F (2, 0) = 1/16,

F (0, 3) = −F (3, 0) = −5/128, F (1, 2) = −F (2, 1) = −1/128.

Problem 14. Consider the map f : R2 → R2

f1 (x1 , x2 ) = ex1 cos(x2 ), f2 (x1 , x2 ) = ex1 sin(x2 ).

Find the fixed points of the map.

Problem 15. Consider the function f : R3 → R3

   
f1 (x1 , x2 , x3 ) x1 cos(x2 ) sin(x3 )
f (x) =  f2 (x1 , x2 , x3 )  =  x1 sin(x2 ) sin(x3 )  .
f3 (x1 , x2 , x3 ) x1 cos(x3 )

Find the fixed points and study their stability.

Problem 16. Consider the function f : R2 → R2

   2 
f1 (x1 , x2 ) x1 − x22
f (x) = = .
f2 (x1 , x2 ) 2x1 x2

The Jacobian matrix is

   
∂f1 /∂x1 ∂f1 /∂x2 2x1 −2x2
=
∂f2 /∂x1 ∂f2 /∂x2 2x2 2x1

with determinant equal to 4(x21 + x22 ). Find the fixed points and study their
stability.
64 Problems and Solutions

Problem 17. Consider the differential operators

d2 c x2
D1 = + +
dx2 x2 16
d2 x2
  
c i d i
D2 = sinh(α) + − + cosh(α) − x −
dx2 x2 16 2 dx 4
 2 2
  
d c x i d i
D3 = cosh(α) + − + sinh(α) − x − .
dx2 x2 16 2 dx 4
(i) Show the differential operator satisfy (Lie algebra su(1, 1))
[D1 , D2 ] = −iD3 , [D2 , D3 ] = iD1 , [D3 , D1 ] = iD2 .
For the actual calculations let f : R → R be a smooth function and one
shows that [D1 , D2 ]f (x) = −iD3 f (x) etc.
(ii) Show that
3 c
D32 − D12 − D22 = − − .
16 4

Problem 18. Let

   
v1 (τ ) w1 (τ )
v(τ ) =  v2 (τ )  , w(τ ) =  w2 (τ ) 
v3 (τ ) w3 (τ )
where vj (τ ), wj (τ ) (j = 1, 2, 3) are smooth functions of τ . Calculate
d
(v × (v × w)).
dτ

Problem 19. Show that the (2, 2) Padé approximant of the cosine func-
tion is given by
12 − 5x2
cos(x) ≈ .
12 + x2

Problem 20. Let f (x) = x2 and g(x) = x ln(x). Show that

g(x)
lim = 0.
x→∞ f (x)

Problem 21. Let c > 0. Find

sinh(cx) x
lim , lim .
x→0 x x→0 sinh(cx)
 Functions 65

Problem 22. Show that the function f : (0, 1) → R

x − 1/2
f (x) =
x(x − 1)

is bijective.

Problem 23. Show that the function

π
f (x) =
sin(πx)

has poles at 0, ±1, ±2, . . ..

Problem 24. For any real number x one defines bxc (floor of x) as the
largest integer less than or equal to x. One defines dxe (ceiling of x) as
the smallest integer greater than or equal to x. Find the minima and the
maxima of the function
f (x) = 2x − bxc.

Problem 25. Let 0 ≤ t < 2. Show that

   
2+t 1 + t/2
ln = ln = 2arctanh(t/2).
2−t 1 − t/2

Problem 26. Consider the continuous function f : R → R

 
1 x
f (x) = +1 .
2 |x| + 1

Find the fixed points and study their stability.

Problem 27. Consider the function f : R2 → R2

1 2
f1 (x1 , x2 ) = (x − x22 ), f2 (x1 , x2 ) = x1 x2 .
2 1

Find the fixed points and study their stability. Is their an inverse f −1 .

Problem 28. Explain

r r
1 − cos(α) 1 + cos(α)
sin(α/2) = ± , cos(α/2) = ± .
2 2
66 Problems and Solutions

Problem 29. Consider the function f : R → R defined by

 2
e−1/x for x 6= 0
f (x) =
0 for x = 0

Show that the function is not analytic. Note that f (x) = f (−x) and

d2 f
 
df 2 4 6
= 3 f (x), = − 4 f (x)
dx x dx2 x6 x
Furthermore limx→±∞ f (x) = 1.

Problem 30. Let N > 0. Show that

lim −N e−(1/) = 0.

→0

Problem 31. Consider the map f : R2 → R2 given by

   
f1 (x1 , x2 ) x1 cos(x2 )
= .
f2 (x1 , x2 ) x1 sin(x2 )

Find the fixed points and study their stability. Show that the functional
matrix is given by
   
∂f1 /∂x1 ∂f1 /∂x2 cos(x2 ) −x1 sin(x2 )
= .
∂f2 /∂x1 ∂f2 /∂x1 sin(x2 ) x1 sin(x2 )

Is the map f invertible?

Problem 32. Let a > 0. Show that

sinh(2ax)
lim = 2a.
x→0 sinh(x)

Problem 33. (i) Write down the first four terms of the Taylor expansion
of cos(x). Then find (2, 2)[x] of the Padé approximant. Discuss the case
x → ∞ for both functions.
(ii) Write down the first four terms of the Taylor expansion of cosh(x).
Then find (2, 2)[2] of the Padé approximant. Discuss the case x → ∞ for
both functions.

Problem 34. Consider the Hénon map (a > 0, b > 0)

x1,t+1 = a − bx2,t + x21,t , x2,t+1 = x1,t , t = 0, 1, 2, . . . .

 Functions 67

Let |x1,0 | < R, |x2,0 | < R, where R is the larger root of λ2 −(|b|+1)λ−a = 0.
Show that all points (x1,0 , x2,0 ) outside of this domain tend to ∞ or −∞
for t → ∞ or t → −∞.

Problem 35. Consider the one-dimensional map (t = 0, 1, 2, . . .)

xt + 2
xt+1 = , x0 ≥ 0.
xt + 1

Find the fixed points of the map. Let x0 = 0. Does limt→∞ xt tend to a
fixed point?

Problem 36. Consider the analytic functions f : R → R, g : R → R

f (x) = sinh(x), g(x) = 2x3 .

Find f −1 ◦ g ◦ f , g −1 ◦ f ◦ g.

Problem 37. Let R > r > 0. Consider the function f : R2 → R2 (battle

of the sexes)
  
R 0 x2
f1 (x1 , x2 ) = ( x1 1 − x1 ) = ((R+r)x2 −r)x1 +r(1−x2 )
0 r 1 − x2
  
r 0 x2
f2 (x1 , x2 ) = ( x1 1 − x1 ) = ((R+r)x2 −R)x2 +R(1−x2 ).
0 R 1 − x2

Find the fixed points and study their stability.

Problem 38. Let c1 , c2 ∈ R and f : R → R be an analytic function.

Show that
   
d d
exp c1 exp(c2 x)f (x) ≡ exp(c2 x) exp c1 exp(c1 c2 )f (x).
dx dx

Problem 39. Let n be a non-negative integer. Let k = 0, 1, . . . , n. Find

the derivative
∂n
(x + y)n .
∂xk ∂xn−k
Show that
∂n
(x + y)n = n!.
∂xk ∂xn−k
68 Problems and Solutions

Problem 40. The content (n-dimensional volume) bounded by a hyper-

sphere of radius r is known to be
2rn π n/2
Vn =
nΓ(n/2)
where Γ is the gamma function. Let r = 1. Show that

lim Vn = 0.
n→∞

Problem 41. Let n ≥ 1. Show that the vector space spanned by

xn , yxn−1 , . . . , y n−1 x, y n

is n + 1 dimensional.

Problem 42. Let , x ∈ R. Show that

sinh(x) sin(x)
lim = x, lim = x.
→0 sinh() →0 sin()

Problem 43. The Hurwitz zeta-function ζH (s, a) is defined by

∞
X
ζH (s, a) := (n + a)−s , 0 < a ≤ 1, <(s) > 1.
k=0

(i) Show that the Hurwitz zeta-function can be written in the form
∞ Z
1 X ∞ s−1 −(n+a)t
ζH (s, a) = a−s + t e dt
Γ(s) n=1 0

where Γ(s) is the Gamma function.

(ii) Show that
∂
ζH (s, a) = −sζH (s + 1, a)
∂a 
∂  1
ζH (s, a) = ln(Γ(a)) − ln(2π).
∂s s=0 2

Problem 44. Let f : R → R be an analytic function. Show that (differ-

ential identity)
dn n

−n d

n
f (x) = x n
f (x)  .
dx d 
=1
 Functions 69

Problem 45. Find an analytic function f : R → R such that

f (∞) = 1, f (−∞) = 0

and
f (x1 ) ≤ f (x2 ) whenever x1 ≤ x2 .
Find the derivative of the function f and determine the maxima of the
function df /dx.

Problem 46. Let


1 for x>0
Θ(x) =
0 otherwise

Show that

 1 for x1 > x3 > x2
Θ(x1 − x3 ) − Θ(x2 − x3 ) = −1 for x2 > x3 > x1
0 otherwise


Problem 47. (i) Let

π


cos 2 cos(θ)
f (θ) = .
sin(θ)

Find f (θ) for θ = nπ (n ∈ Z) using the L’Hospital rule. Plot f (θ) as a

function of θ.
(ii) Apply the L’Hospital rule to show that

ex − 1
lim = 1.
x→0 x
(iii) Show that

sinh(x) x
lim = 1, lim = 1.
x→0 x x→0 sinh(x)

(iv) Consider the function f : R → R

x cosh(x) − sinh(x)
f (x) = .
2x sinh2 (x)

Find
lim f (x), lim f (x).
x→0 x→∞
70 Problems and Solutions

Problem 48. Consider the functions

x sinh(x)
f (x) = , g(x) = .
sinh(x) x
Find
lim f (x), lim g(x).
x→0 x→0

Problem 49. Let x ∈ R. Show that

1 − e−x
lim = 1.
x→∞ x

Problem 50. (i) Show that

1 − cos(x)
lim = 0.
x→0 x
(ii) Show that
x−1
lim = 1.
x→1 ln(x)
(iii) Show that
cosh(x) − cosh(2x)
lim = 0.
x→0 sinh(x)
(iv) Let x > 0. Show that
ln(1 + x)
lim = 1.
x→0 x

Problem 51. Let α ∈ R. Find

sin(α) cos(α)
lim .
α→0 α

Problem 52. Let k > 0 and dimension 1/length. Find the extrema of
the functions
f1 (x) = ln(cosh(kx))
f2 (x) = cosh(kx) + α sinh(kx), −1 ≤ α ≤ +1
1
f3 (x) = Asech(kx) ≡ A
cosh(kx)
sinh2 (kx)
f4 (x) = asech2 (kx) ≡ a .
cosh2 (kx)
 Functions 71

Problem 53. Consider the analytic function f : R → R,

f (x) = cos(x) + sin(x).

(i) Find the critical points of f and the minima and maxima of the function.
(ii) Find the roots of f , i.e. solve f (x) = 0.
(iii) Find the fixed points of f , i.e. solve f (x) = x.
(iv) Find the differential equation together with the initial conditions the
function f satisfies.

Problem 54. Consider the polynomial f : R → R

3 3
f (x) = x3 − x2 + x − .
2 2
Show that there is a root, i.e. a solution of f (x) = 0 in the interval [1, 2].
Apply the Newton method (say 3 steps) to find an approximate solution
for the root. Apply the two initial condition x0 = 1 and x0 = 2. After
finding this root how would you proceed to find the other two roots?

Problem 55. A real valued function f defined on an open subset G of

a Banach space E is said to be Fréchet differentiable at a point x ∈ G if
there is f 0 (x) ∈ E ∗ such that

|f (x + v) − f (v) − hf 0 (x), vi|

lim = 0.
v→0 kvk

Then f 0 (x) is called the Fréchet derivative of f at x. Let E = R. Let

f (x) = x3 + 2x2 + 3x + 4. Find the Fréchet derivative.

Problem 56. Consider the function

Rx
exp(−x2 /2) 0
exp(τ 2 /2)dτ
f (x) = .
x
Show that
lim f (x) = 1.
x→0

Problem 57. Consider the functions f : R → [−1, 1], g : R → [−1, 1]

f (x) = sin(x), g(x) = cos(x).

(i) Find the minima and maxima of the function

h1 (x) = max(f (x), g(x)).

72 Problems and Solutions

Is the function h1 differentiable?

(ii) Find the minima and maxima of the function

h2 (x) = min(f (x), g(x)).

Is the function h2 differentiable?

Problem 58. Let a ∈ R. Show that

lim (1 + a/x)x = ea .
x→∞

Problem 59. Let k ∈ R, x, y ∈ {+1, −1} such that xy = ±1. Show that

ekxy = cosh(k) + xy sinh(k).

Problem 60. Consider the coordinate transformation

x1 (r, α, β, γ) = r sin(β/2) sin((γ − α)/2)

x2 (r, α, β, γ) = r sin(β/2) cos((γ − α)/2)
x3 (r, α, β, γ) = r cos(β/2) sin((γ + α)/2)
x4 (r, α, β, γ) = r cos(β/2) sin((γ + α)/2).

Show that x21 + x22 + x23 + x24 = r2 .

Chapter 4

Polynomial

4.1 Solved Problems

Problem 1. The Chebyshev polynomials are defined by

Tk (x) = cos(k arccos(x)), k = 0, 1, . . . x ∈ [−1, 1]

Thus the first six polynomials are

T0 (x) = 1
T1 (x) = x
T2 (x) = 2x2 − 1
T3 (x) = 4x3 − 3x
T4 (x) = 8x4 − 8x2 + 1
T5 (x) = 16x5 − 20x3 + 5x.

Find 1, x, x2 , x3 , x4 , x5 as functions of T0 , T1 , T2 , T3 , T4 , T5 .

Problem 2. Let Ln (x), Hn (x) be the Laguerre polynomials and Hermite

polynomials, where n = 0, 1, . . .. Let

x−α ex dn −x n+α
L(α)
n (x) := (e x )
n! dxn

73
74 Problems and Solutions

be the associated Laguerre polynomials with α > −1 and n = 0, 1, . . .. The

Laguerre polynomials are recovered by setting α = 0. We have

H2n (x) = (−4)n n!Ln(−1/2) (x2 ) (1)

and the following addition formula for the associated Laguerre polynomials
Lα
n (x)
n
(α) (β)
X
L(α+β+1)
n (x + y) = Ln−k (x)Lk (y) . (2)
k=0

(i) Find a new sum rule by inserting (2) into (1).

(ii) Consider the sum rule
n
1 √ √ X (−1/2) (−1/2)
Hn ( 2x)Hn ( 2y) = (−1)k Ln−k ((x+y)2 )Lk ((x−y)2 ). (3)
n!2n
k=1

Insert (1) into (3) to find a sum rule for Hermite polynomials.

Problem 3. The Hermite polynomial of degree n can be written as

[n/2]
X (−1)k n!
Hn (x) = (2x)n−2k .
k!(n − 2k)!
k=0

Express xn using the Hermite polynomials.

Problem 4. Given the differentiable function f : R → R with

f (x) = 4x(1 − x).

(i) Find the fixed points of f (x) and f (f (x)).

(ii) Find the critical points of f (x) and f (f (x)).

Problem 5. Let P (z) be a polynomial of degree n ≥ 2 with distinct zeros

ζ1 , . . . , ζn . Show that
n
X 1
0 (ζ )
=0
j=1
P j

where 0 denotes the derivative, i.e. P 0 (ζ) ≡ dP (z = ζ)/dz.

Problem 6. Given a set of N real numbers x1 , x2 , . . . , xN . It is often

useful to express the sum of the j powers

sj = xj1 + xj2 + · · · + xjN , j = 0, 1, 2, . . .

 Polynomial 75

in terms of the elementary symmetric functions

N
X
σ1 = xi
i=1
N
X
σ2 = xi xj
i<j
N
X
σ3 = xi xj xk
i<j<k
.. ..
. .
σN = x1 x2 · · · xN .
Consider the special case with three numbers x1 , x2 , x3 . Then the elemen-
tary symmetric functions are given by
σ1 = x1 + x2 + x3 , σ2 = x1 x2 + x1 x3 + x2 x3 , σ3 = x1 x2 x3 .
We know that the elementary symmetric functions are the coefficients (up
to sign) of the polynomial with the roots x1 , x2 , x3 . In other words the
values of x1 , x2 , x3 each satisfy the polynomial equation
x3 − σ1 x2 + σ2 x − σ3 = 0. (1)
Find a recursion relation for
sj := xj1 + xj2 + xj3 , j = 0, 1, 2, . . .
and give the initial values s0 , s1 , s2 . Calculate s3 and s4 .

Problem 7. The dominant tidal potential at position (r, φ, λ) due to the

moon or sun is given by
GM ∗ r2 0
U (r) = P2 (cos(ψ))
r∗3
where M ∗ is the mass of the moon or sun located at (r∗ , φ∗ , λ∗ ). Moreover,
ψ is the angle between mass M ∗ and the observation point at (r, φ, λ),
where φ is the latitude and λ is the longitude. By the spherical cosine
theorem we have
cos(ψ) = sin(φ) sin(φ∗ ) + cos(φ) cos(φ∗ ) cos(λ − λ∗ ).
The Legendre polynomials are defined as
1 dn 2
Pn (x) := (x − 1)n
2n n! dxn
76 Problems and Solutions

with n = 0, 1, 2, . . . and the associated Legendre polynomials are defined as

dm (1 − x2 )m/2 dm+n 2
Pnm (x) := (1 − x2 )m/2 P n (x) = (x − 1)n
dxm 2n n! dxm+n
with Pn0 (x) = Pn (x) and Pnm = 0 if m > n.
(i) Show that U (r) can be written as

GM ∗ r2

1
U (r) = P20 (sin φ)P20 (sin φ∗ ) + P21 (sin φ)P21 (sin φ∗ ) cos(λ − λ∗ )
r∗3 3

1
+ P22 (sin φ)P22 (sin φ∗ ) cos(2(λ − λ∗ )) .
12

(ii) Give an interpretation (maxima and nodes) of the terms in the paren-
thesis.

Problem 8. Consider the cubic equation

y 3 + py + q = 0, p, q ∈ R, pq 6= 0. (1)

Show that applying the nonlinear transformation

p
y(z) := z −
3z
equation (1) can be reduced to

p3
z 6 + qz 3 − =0
27
and with u = z 3 to a quadratic equation.

Problem 9. Find all integers c for which the cubic equation

x3 − x + c = 0

has three integer roots.

Problem 10. Let Φ be an endomorphism of the space Cn [X] of polyno-

mials of degree n with complex coefficients, which maps a polynomial p(X)
to the polynomial p(X + 1). Let Ψ be an endomorphism
  of the space Cn [X]
X
which maps a polynomial p(X) to (1 − X)n p 1−X , which of course is also
a polynomial. Show that we have a braid-like relation

Φ ◦ Ψ ◦ Φ = Ψ ◦ Φ ◦ Ψ.
 Polynomial 77

Problem 11. Let a ∈ R. Let r and s be the roots of the quadratic

equation
a2 − 1
x2 + ax + = 0.
2
Find r3 + s3 in terms of a, and express it as a polynomial in a with rational
coefficients.

Problem 12. Consider the polynomial p(x) = x3 − x2 + x − 2. Does

there exist a nontrivial polynomial q(x) with real coefficients such that the
degree of every term of the product p(x)q(x) is a multiple of 3? If so, find
one. If not, show there is none.

Problem 13. (i) Let n be a positive integer. Let f : [a, b] → R be

a continuous function. The Bernstein polynomials of degree n associated
with the continuous function f are given by
n  
1 X n
Bn (f (x), x) := (x − a)j (b − x)n−j f (xj )
(b − a)n j=0 j

where
b−a
xj = a + j , j = 0, 1, . . . , n.
n
Consider the function f : [0, 1] → R given by f (x) = sin(4x). Show that
B2 (f, x) is not a “good approximation” for f . Consider x = π/8.
(ii) The Bernstein basis polynomials are defined as
 
n j
Bn,j (x) := x (1 − x)n−j , j = 0, 1, 2, . . . , n, x ∈ [0, 1].
j

Show that
n
X
Bn,j (x) = 1.
j=0

Show that Bn,j satisfies the recursion relations

Bn,j (x) = (1 − x)Bn−1,j (x) + xBn−1,j−1 (x), j = 1, 2, . . . , n − 1

Bn,0 (x) = (1 − x)Bn−1,0 (x)
Bn,n (x) = xBn−1,n−1 (x).

Problem 14. A one-dimensional map f is called an invariant of a two-

dimensional map g if
g(x, f (x)) = f (f (x)).
78 Problems and Solutions

Let
f (x) = 2x2 − 1.
Show that f is an invariant for

g(x, y) = y − 2x2 + 2y 2 + d(1 + y − 2x2 ).

Problem 15. Consider the functions f (z) = z 3 and h(z) = z + 1/z. Find
a function p such that
h(f (z)) = p(h(z)). (1)

Problem 16. Consider the map fc (z) = z 2 + c, where c ∈ C. Find all

complex c-values where the map fc has a fixed point z ∗ with fc0 (z ∗ ) = −1.

Problem 17. Let p(x, y) be a real polynomial. Show that if p(x, y) = 0

for infinitely many (x, y) on the unit circle x2 + y 2 = 1, then p(x, y) = 0 on
the unit circle.

Problem 18. Show that the equation

zn = a (1)

where n is a positive integer and a is any nonzero complex number, has

exactly n roots. Hint. Set

a = ρ(cos(φ) + i sin(φ)). (2)

Problem 19. Let A be a 2 × 2 matrix over the real numbers R. The

trace is defined as
tr(A) = a11 + a22 .
It can be proved that the trace is the sum of the eigenvalues of A, i.e.

tr(A) = λ1 + λ2 .

Thus we have
tr(A2 ) = λ21 + λ22 .
Let  
1 1
A= .
1 1
Find the eigenvalues using the two equations.
 Polynomial 79

Problem 20. Find the zeros of the cubic polynomial

α(x) := a0 + a1 x + a2 x2 + x3 (1)

over C, where a0 , a1 , a2 ∈ R and a0 6= 0.

Problem 21. Show that the zeros of

z 3 − 6z 2 + 11z − 6 = 0 (1)

are given by 1, 2, and 3.

Problem 22. Find the zeros of the quartic polynomial

α(x) = a0 + a1 x + a2 x2 + a3 x3 + x4 (1)

over C when a0 6= 0.

Problem 23. Find the zeros of

α(x) = 35 − 16x − 4x3 + x4 . (1)

Problem 24. The variational equation of the Lorenz model

dX
= −σX + σY (1a)
dt
dY
= −XZ + τ X − Y (1b)
dt
dX
= XY − bZ (1c)
dt
is given by
    
dx0 /dt −σ σ 0 x0
 dy0 /dt  =  (τ − Z) −1 −X   y0  . (2)
dz0 /dt Y X −b z0

(i) Show that Lorenz model possess the steady-state solution X = Y =

Z = 0, representing the state of no convection.
(ii) Show that with this basic solution, the characteristic equation of the
variational matrix is

(λ + b)(λ2 + (σ + 1)λ + σ(1 − τ )) = 0. (3)

80 Problems and Solutions

(iii) Show that this equation has three real roots when τ > 0; all are negative
when τ < 1, but one is positive when τ > 1. The criterion for the onset of
convection is therefore τ = 1.
(iv) Show that when τ > 1, system (1) possess two additional steady state
solutions p
X = Y = ± b(τ − 1), Z = τ − 1. (4)
(v) Show that for either of these solutions, the characteristic equation of
the matrix in (2) is
λ3 + (σ + b + 1)λ2 + (τ + σ)bλ + 2σb(τ − 1) = 0. (5)
(vi) Show that this equation possesses one real negative root and two com-
plex conjugate roots when τ > 1. Show that the complex conjugate roots
are pure imaginary if the product of the coefficients of λ2 and λ equals the
constant term, or
τ = σ(σ + b + 3)(σ − b − 1)−1 . (6)

Problem 25. The variational equation of the Lotka Volterra model

du1 du2
= u1 − u1 u2 , = −u2 + u1 u2 (1)
dt dt
is given by     
dv1 /dt 1 − u2 −u1 v1
= (2)
dv2 /dt u2 −1 + u1 v2
where u1 > 0 and u2 > 0.
(i) Show that (1) possess the steady-state solution u1 = u2 = 1.
(ii) Show that with this basic solution, the characteristic equation of the
matrix in (2) is
λ2 + 1 = 0. (3)
(iii) Find the solution of the characteristic equation and discuss.

Problem 26. Show the following. Let P (z) be a polynomial. Then either

1. P (z) has a fixed point q with P 0 (q) = 1,

2. P (z) has a fixed point q with |P 0 (q)| > 1.

Problem 27. Let Sn be the symmetric group. The symmetric group

Sn acts naturally on polynomials in n variables. For a polynomial p in n
variables, define the symmetrized polynomial associated to p, Symn (p), by
X
Symn (p) := σ(p).
σ∈Sn
 Polynomial 81

Let n = 2 and
p(x1 , x2 ) = x21 x2 + 2x2 .
Find Sym2 (p).

Problem 28. Consider the system of equations

z1 + z2 + · · · + zn−1 + zn = 0
z1 z2 + z2 z3 + · · · + zn−1 zn + zn z1 = 0
..
.
z1 z2 · · · zn−1 + z2 z3 · · · zn + · · · + zn−1 zn · · · zn−3 + zn z1 · · · zn−2 = 0
z1 z2 · · · zn = 1.

Find the solutions for the case n = 2 and n = 3. This system of equations
arise as follows. Let p be a prime number. A vector x = (x0 , x1 , . . . , xp−1 ) ∈
Cp viewed as a function Z/p → C has discrete Fourier transform x̂ =
(x̂0 , x̂1 , . . . , x̂p−1 ), where

p−1
X
x̂j = ω jk xk , ω := e2πi/p .
k=0

The vector x is called equimodular if all its coordinates have the same abso-
lute value, and x is called bi-equimodular if both x and x̂ are equimodular.
The question is: Which vectors are bi-equimodular?

Problem 29. Let α ∈ R. Find the roots of the characteristic equation

λ6 − 2 cos(3α)λ3 + 1 = 0.

Problem 30. Show that the n-th order polynomial

p(x) = a0 + a1 x + a2 x2 + · · · + an xn

which goes exactly through n + 1 data points is unique.

Problem 31. Let p : R → R be a polynomial that satisfies

p(1 − x) + 2p(x) = 3x

for all x ∈ R. Find the values of p(0) and p(1). Give an example of a
polynomial that satisfies this condition.
82 Problems and Solutions

Problem 32. (i) Let x1 , x2 ∈ R and x1 6= x2 . Find the solutions of the

system of equations

1 1
x1 − = 0, x2 + = 0.
(x1 − x2 )2 (x1 − x2 )2

(ii) Let x1 , x2 , x3 ∈ R and x1 6= x2 , x1 6= x3 , x2 6= x3 . Find the solutions

of the system of equations

1 1
x1 − − =0
(x1 − x2 )2 (x1 − x3 )2
1 1
x2 + 2
− =0
(x1 − x2 ) (x2 − x3 )2
1 1
x3 + + = 0.
(x1 − x3 )2 (x2 − x3 )2

Problem 33. Let z, w ∈ C. Find the solution of the system of equations

|z|2 + |w|2 = 1, z 2 + w3 = 0.

Problem 34. Consider the two polynomials

p1 (x) = a0 + a1 x + · · · + an xn , p2 (x) = b0 + b1 x + · · · + bm xm

where n = deg(p1 ) and m = deg(p2 ). Assume that n > m. Let r(x) =

p2 (x)/p1 (x). We expand r(x) in powers of 1/x, i.e.
c1 c2
r(x) = + 2 + ···
x x
From the coefficients c1 , c2 , . . . , c2n−1 we can form an n × n Hankel matrix
c
1 c2 ··· cn 
 c2 c3 ··· cn+1 
Hn = 
 .. .. .. .
.. 
. . . .
cn cn+1 ··· c2n−1

The determinant of this matrix is proportional to the resultant of the two

polynomials. If the resultant vanishes, then the two polynomials have a
non-trivial greates common divisor. Apply this theorem to the polynomials

p1 (x) = x3 + 6x2 + 11x + 6, p2 (x) = x2 + 4x + 3.

 Polynomial 83

Problem 35. Let p1 and p2 two polynomials with n = degree(p1 ), m =

degree(p2 ) and n > m. We expand the rational function
p2 (x)
r(x) =
p1 (x)
with respect to powers of 1/x, i.e.,

r(x) = d1 x−1 + d2 x−2 + · · ·

The coefficients d1 , d2 , . . . , d2n−1 are inserted into the n × n Hankel matrix

d d2 ... dn 
1
 d2 d3 . . . dn+1 
Hn =   .. .. .. .
.. 
. . . .
dn dn+1 . . . d2n−1
A Hankel matrix is a diagonal matrix in which all the elements are the
same along any diagonal that slopes from northeast to southwest. If the
determinant of this matrix is zero, then the two polynomials have a non-
trival common divisor. Apply this algorithm to the polynomials

p1 (x) = x3 + 6x2 + 11x + 6, p2 (x) = x2 + 6x + 8

to test whether they have a common non-trivial divisor.

Problem 36. Let p be a polynomial with coefficients in R. If the equation

p(x) = 0 has repeated roots, then p(x) and dp(x)/dx have a highest common
factor. Apply this to the polynomial

p(x) = 32x4 − 64x3 + 24x2 + 8x − 3.

Problem 37. Let p be a polynomial with real coefficients. The equation

p(x) = 0 cannot have more positive roots than there are changes of sign
from + to − and from − to + in the coefficients of the polynomial p(x),
or more negative roots than there are changes of sign in p(−x) (Descartes
rule of signs). Apply the rule to the polynomiol

p(x) = x6 + 7x3 + x − 2 = 0.

Problem 38. The Liouville-Riemann definition for the fractional integral

operator Dx−q is given by
Z x
−q 1
Dx f (x) := (x − y)q−1 f (y)dy, q > 0.
Γ(q) 0
84 Problems and Solutions

The fractional differential operator Dxν for ν > 0 is given by the definition
dn
Dxν f (x) := (Dν−n f (x)), ν − n < 0.
dxn x
Let f (x) = x2 . Find Dx−q f (x).

Problem 39. Let n ∈ N. Show by induction that xn − y n is divisible

without remainder by x − y for all values of n. We have
xn+1 − y n+1 ≡ x(xn − y n ) + y n (x − y).

Problem 40. Consider the polynomial

n
X
Pn (x) = (−1)j aj xn−j , a0 = 1.
j=0

The homogeneous product sums symmetric functions hk (x1 , . . . , xn ) of the

zeros of this polynomial are defined as follows
n
Y
(1 − xj x) = (1 + h1 x + h2 x2 + h3 x3 + · · ·)−1 .
j=1

Show that the first few sums are given explicitly by

n
X
h1 (x1 , . . . , xn ) = xj
j=1
Xn n
X
h2 (x1 , . . . , xn ) = x2j + xj xk
j=1 j<k
n
X Xn n
X
h3 (x1 , . . . , xn ) = x3j + + xj xk x` .
j=1 j6=k j<k<`

Problem 41. Consider the quintic equation

p(x) = x5 − 5x3 + 5x − 5 = 0.
Is p irreducible over the rational numbers? What is the Galois group of p?
Look for solutions of the form r + 1/r.

Problem 42. The Bernoulli polynomials Bn (x) (n = 0, 1, . . .) can be

defined recursively by
dBn (x)
= nBn−1 , n = 1, 2, . . .
dx
 Polynomial 85

with B0 (x) = 1 and the condition

Z 1
Bn (x)dx = 0, n ≥ 1.
0

The Bernoulli numbers Bn are defined by Bn := Bn (x = 0).

(i) Find the first four Bernoulli polynomials.
(ii) Show that
m n  
X 1 X n
Bm (x) = (−1)k (x + k)m .
n=0
n + 1 k
k=0

(iii) Show that

Bn (1 − x) = (−1)n Bn (x).

Problem 43. Consider the operators

d d
D1 = x, D2 = , D3 = x
dx dx
which apply to the function of the Bargmann space
xn
B := { f (n) = √ , n ∈ N0 , x ∈ R }.
n!
Show that
√ √
D1 f (n) = n + 1f (n + 1), D2 f (n) = nf (n − 1), D3 f (n) = nf (n).

Find the commutators [D1 , D2 ], [D1 , D3 ], [D2 , D3 ].

Problem 44. Express the polynomial

p(x) = 7x21 + 6x22 + 5x23 − 4x1 x2 − 4x2 x3 + 14x1 − 8x2 + 10x3 + 6

in matrix form
   
x1 x1
p(x) = ( x1 x2 x3 ) A  x2  + vT  x2  + c
x3 x3

Find the eigenvalues and normalized eigenvectors of the 3 × 3 matrix.

Problem 45. Find the polynomials generated by

dpj+1 (x)
= (j + 1)pj (x), j = 0, 1, 2, . . .
dx
86 Problems and Solutions

with p0 (x) = 1. The constants of integration we set to 0.

Problem 46. The complete Bell polynomials Bj (x1 , x2 , . . . , xj ) are given

by the exponential generating function
 
∞ j ∞
X t X tn
exp  xj  = Bn (x1 , . . . , xn ) . (1)
j=1
j! n=0
n!

Taking the n-th derivative with respect to t we obtain

 
n ∞ j

d X t 
exp  x j = Bn (x1 , . . . , xn ) (2)
dtn j=1
j! 
t=0

with B0 = 1. Find the first four Bell polynomials.

Programming Problem

Problem 1. Consider the polynomials f1 : R → R, f2 : R → R

f1 (x) = x2 + 2x + 1, f2 (x) = 2x.
Apply Maxima to find
h1 (x) = f1 (f2 (x)), h2 (x) = f2 (f1 (x))
and C = f1 − f2 .

4.2 Supplementary Problems

Problem 1. Construct a polynomial

p(x) = x2 + ax + b
that admits the roots
1 √ 1 √
( 3 + 3), − ( 3 − 3)
6 6
and the fixed points
1 √ 1 √
( 30 + 6), − ( 30 − 6).
6 6
 Polynomial 87

Problem 2. Let a, b ∈ R. Consider the quartic equation

(λ − 1)2 (λ − b2 )2 − λ2 a2 (1 − b2 )2 = 0

Show that the roots are given by

1
λ1 (a, b) = (1 − a + (1 + a)b2 + (T+ )1/2 )
2
1
λ2 (a, b) = (1 − a + (1 + a)b2 − (T+ )1/2 )
2
1
λ3 (a, b) = (1 + a + (1 − a)b2 + (T− )1/2 )
2
1
λ4 (a, b) = (1 + a + (1 − a)b2 − (T− )1/2 )
2
where T± = (1 + a2 )(1 − b2 )2 ± 2a(b4 − 1).

Problem 3. Let n be a positive integer. Show that xn − y 2 has x − y as

a factor for all n.

Problem 4. Let s = 1/2, 1, 3/2, 2, . . . be the spin values. The Brillouin

function is defined as
   
2s + 1 2s + 1 1 1
Bs (x) := coth x − coth x
2s 2s 2s 2s
Find Bs (x = 0) applying L’Hospital.

Problem 5. Let n ∈ N0 and x, m ∈ R. The Sonine polynomials are

defined by
n
n
X Γ(m + n + 1)xj
Sm (x) = (−1)j
j=0
Γ(m + j + 1)(n − j)!j!

where Γ denotes the gamma function. Show that

0 1
Sm (x) = 1, Sm (x) = m + 1 − x.

Show that the Sonine polynomials satisfy

Z ∞
Γ(m + j + 1)
e−x xm Sm
j k
(x)Sm (x)dx = = δj,k .
0 j!

Problem 6. (i) Show that the cubic roots of unity z 3 = 1 are

√ √
1 3i 2 1 3i
1, w=− + , w =− − . (2)
2 2 2 2
88 Problems and Solutions

(ii) Show that the zeros of z 3 + 1 = 0 are given by

1 i√ 1 i√
−1, + 3, − 3. (1)
2 2 2 2
(iii) Show that the set

1 i√ 1 i√
S = { ω1 = − + 3, ω2 = − − 3, ω3 = 1 }
2 2 2 2
of the cubic roots of 1, forms an abelian group with respect to multiplication
on the set of complex numbers C, i.e. show that

ω1 ω2 = ω3 , ω2 ω1 = ω3 , ω1 ω3 = ω1 ,

ω3 ω1 = ω1 , ω2 ω3 = ω2 , ω3 ω2 = ω2 . (1)
Obviously, the neutral element is ω3 . From (1) we see that each element
has an inverse (which is unique), i.e.

(ω1 )−1 = ω2 , (ω2 )−1 = ω1 , (ω3 )−1 = ω3 .

The associative law is true for all complex numbers.

Problem 7. Let a, b ∈ R and z be denote a root of the quadratic equation

z 2 + az + b = 0. (1)

Show that the sequence of powers of z, {z n }, n ≥ 2, satisfies the linear

difference equation

wn + awn−1 + bwn−2 = 0, n ≥ 2. (2)

Set wn = xn . Then wn−1 = xn−1 .

Problem 8. A real number is said to be algebraic if it is a zero of a

polynomial
cn xn + cn−1 xn−1 + · · · + c0
where the coefficients cj are integers. The height of such polynomials is
defined as the positive integer number

h = n + |cn | + |cn−1 | + · · · + |c0 |.

Show that there are only finitely many polynomials of height h. Show that
the set of all algebraic real numbers is enumerable.
 Polynomial 89

Problem 9. Let p0 = 1, p1 , p2 , . . . be polynomials which satisfy the

recursion relation
n  2
1 X 2k n!
pn = 2 xk pn−k , n ∈ N.
2n (n − k)!
k=1

Giva a SymbolicC++ and Maxima implementation of this recursion rela-

tion. For example p1 = 2x1 .

Problem 10. Show that every polynomial α(x) ∈ C[x] of degree m ≥ 1

has precisely m zeros over C, where any zero of multiplicity is to be counted
as n of the m zeros.

Problem 11. Show that if r ∈ C is a zero of any polynomial α(x) with

real coefficients, then r̄ is also a zero of α(x), where r̄ denotes the complex
conjugate of r.

Problem 12. Let

P = {fi (x), i = 1, . . . , k}, x = (x1 . . . xn ) (1)
be a set of multivariate polynomials in the ring Q[x1 . . . xn ], with a solution
set
S = S(fi . . . fk ) = {x | fi (x) = 0, ∀i = 1, . . . , k} (2)
All polynomials X
u(x) = gj (x)fj (x) (3)
j

for arbitrary polynomials gj will vanish in all points of S. The set of all u
establishes a polynomial ideal
 
X 
I = I(f1 · · · fk ) = gj (x)fj (x) (4)
 
j

and classical algebra tells that S is invariant if we replace the set {f1 , · · · , fk }
by any other basis for the ideal I(f1 , · · · fk ).

The Buchberger algorithm allows one to transform the set of polynomials

into a canonical basis of the same ideal, the Gröbner basis GB = GB(I).
For the purpose of the equation solving especially Gröbner bases computed
under lexicographical term ordering are important. They allow one to de-
termine the set S directly. If I has dimension zero (S is a finite set of
isolated points, GB has in most cases the form
g1 (x1 , xk ) = x1 + c1,m−1 xm−1
k + c1,m−2 xkm−2 + · · · + c1,0
90 Problems and Solutions

g2 (x2 , xk ) = x2 + c2,m−1 xm−1

k + c2,m−2 xkm−2 + · · · + c2,0
···
gk−1 (xk−1 , xk ) = xk−1 + ck−1,m−1 xkm−1 + ck−1,m−2 xm−2
k + · · · + ck−1,0
m−1
gk (xk ) = xm
k + ck,m−1 xk + ck,m−2 xm−2
k + · · · + ck,0 .
A basis in this form has the elimination property: the variable dependency
has been reduced to a triangular form, just as with a Gaussian elimination
in the linear case. The last polynomial is univariate in xk . It can be
solved with usual algebraic or numeric techniques; its zero x̄k then are
propagated into the remaining polynomials, which then immediately allow
one to determine the corresponding coordinates (x̄1 , . . . , x̄k−1 ).
Consider the system
{y 2 − 6y, xy, 2x2 − 3y − 6x + 18, 6z − y + 2x}. (4)
(i) Show that that this system has for {x, y, z} the lexicographical Gröbner
basis
{g1 (x, z) = x−z 2 +2z−1, g2 (y, z) = y−2z 2 −2z−2, g3 (z) = z 3 −1}. (5)
(iii) Show that the roots of the third polynomial are given by
( √ √ )
1 − 3i 3i − 1
z = 1, z = , z= . (6)
2 2

If we propagate one of them into the basis, we generate univariate polyno-

√
mials of degree one which can be solved immediately; e.g. Selecting 1−2 3i
for z the basis reduces to
( √ )
3 3i + 3
x= , y, 0 (7)
2

such that the final solution for this branch is

( √ √ )
3 3i + 3 1 − 3i
x= , y = 0, z = . (8)
2 2

For zero dimensional problems the last polynomial will always be univariate.
However, in degenerate cases the other polynomials can contain their lead-
ing variable in a higher degree, then containing more mixed terms with the
following variables. And there can be additional polynomials with mixed
leading terms (of lower degree) imposing some restrictions. But the varible
dependency pattern will remain triangular (evenutally with more than k
rows). There is a special algorithm for decomposing such ideals using ideal
quotients.
Chapter 5

Equations

5.1 Solved Problems

Problem 1. Solve the equation
1 1 1
= + .
3 x 4

Problem 2. Let x1 , x2 be positive real numbers. Consider the equation

x2 sin(θ) = x1 cos(θ).

Find sin(θ).

Problem 3. Let L be a given positive real number. Solve the system of

two coupled nonlinear equations

1 = x2 L + 2x2 y + Lx2 y 2
0 = −x2 + 2x2 y + Lx2 y 2 .

Problem 4. Let c, y ∈ R. Solve the quadratic equation

x = cy + yx − x2

91
92 Problems and Solutions

with respect to x.

Problem 5. Consider a triangle in the plane. Consider a point within

the triangle. We draw lines from this point to the three vertices, thereby
dividing the triangle into three triangles of area A1 , A2 , A3 . The sides of
the triangle are designated by the same number as the opposite vertex. The
areas are identified by the number of the adjacent side. The quantities Lj
(j = 1, 2, 3)
A1 A2 A3
L1 = , L2 = , L3 =
A A A
where A is the area of the original triangle, are defined to be the triangular
coordinates. Show that
L1 + L2 + L3 = 1.
The relationship between the Cartesian coordinates x, y which are the
coordinates of the points in the elements, and the triangular coordinates
L1 , L2 , L3 are
x = L1 x1 + L2 x2 + L3 x3 , y = L1 y1 + L2 y2 + L3 y3
where xj , yj (j = 1, 2, 3) are the coordinates of the nodes. The triangular
coordinates can be expressed in terms of the known locations of the vertices,
i.e.     
1 1 1 1 L1
 x  =  x1 x2 x3   L2  .
y y1 y2 y3 L3
Find L1 , L2 , L3 as function of xj , yj , x, y.

Problem 6. Let x, y 6= 0. Find all solutions of the equation

4x2 y 2
= 1.
(x2+ y 2 )2

Problem 7. Solve the equation

71x2 = 133 (mod 11).

Problem 8. Let x, y ∈ N. Find all solutions of

16x + 7y = 601.

Problem 9. Let x, y ∈ Z. Find all solutions of

18x + 12y = 4.
 Equations 93

Problem 10. Consider the two hyperplane (n ≥ 1)

x1 + x2 + · · · + xn = 2, x1 + x2 + · · · + xn = −2.

The hyperplanes do no intersect. Find the shortest distance between the

hyperplanes. First consider the cases n = 1 and n = 2 and then the general
case. What happens if n → ∞?

Problem 11. Let d be a positive distance, v, V be velocities v 6= V and

T1 , T2 time-intervals. Assume that
d d
= T1 , = T2 .
V +v V −v
Find d/V .

Problem 12. Solve the quadratic equation

x2 − ix − (1 + i) = 0.

Problem 13. Let z ∈ C. Solve

 
z+1
= i.
z−1

Problem 14. Solve the system of nonlinear equations

x2 − (y − z)2 = a2
y 2 − (z − x)2 = b2
z 2 − (x − y)2 = c2 .

Problem 15. Show that the quartic equation

x4 + qx2 + rx + s = 0

can be written as

(x2 − ex + f )(x2 + ex + g) = 0

where e2 is the root of the cubic equation

z 3 + 2qz 2 + (q 2 − 4s)z − r2 = 0.
94 Problems and Solutions

Problem 16. Let

   
z1 w1
z= , w=
z2 w2

be elements of C2 . Solve the equation z∗ w = w∗ z.

Problem 17. A special set of coordinates on S n called spheroconical (or

elliptic spherical) coordinates are defined as follows: For a given set of real
numbers α1 < α2 < · · · < αn+1 and nonzero x1 , . . ., xn+1 the coordinates
λj (j = 1, . . . , n) are the solutions of the equation
n+1
X x2j
.
j=1
λ − αj

Find the solutions for n = 2.

Problem 18. Find all solutions of the system of equations

1 cos(α) cos(β)
   
1  1   cos(α) sin(β) 
 = .
2 1 sin(α) cos(β)
1 sin(α) sin(β)

Problem 19. Let φ ∈ [0, 2π). Solve the cubic equation

4x3 − 3x − cos(φ) = 0

over the real numbers.

Problem 20. Let  > 0. Find the solution of the coupled two non-linear
equations
x2 + x − y − 1 = 0, y 2 + x − y − 1 = 0.
Study  → 0 for these solutions.

Problem 21. Consider the six 3 × 3 matrices

 0
b012
  
a12 b12 0 a12 0
0
X12 =  c12 d12 0  , X12 =  c012 d012 0,
0 0 1 0 0 1
   0 0

a13 0 b13 a13 0 b13
0
X13 = 0 1 0  , X13 = 0 1 0 ,
c13 0 d13 c013 0 d013
 Equations 95
   
1 0 0 1 0 0
0
X23 = 0 a23 b23  , X23 = 0 a023 b023  .
0 c23 d23 0 c023 d023
Find the 9 conditions on the entries such that (local Yang-Baxter equation)
0 0 0
X12 X13 X23 = X23 X13 X12 .

Find solutions of these 9 equations for the 24 unkowns a12 , . . . , d023 .

Problem 22. Solve the equations

2 arcsin(x) + arcsin(2x) − π/2 = 0

2 arcsin(x) − arccos(3x) = 0
arccos(x) − arctan(x) = 0
arccos(2x2 − 4x − 2) − 2 arcsin(x) = 0
 
2x
2 arctan(x) − arctan = 0.
1 − x2

Problem 23. Let x ∈ R. Find all values of x that satisfy

|5x + 7| = 3.

Then find the smallest and largest values.

Problem 24. Let θ ∈ [0, 2π). Can one find x ∈ R such that

e−x/2
sin(θ) = √ ?
1 + e−x

Problem 25. Let x < 1. Find a solution of the equation

f 2 (x) − 2f (x) + x = 0.

Problem 26. What are the conditions on c1 > 0 and c2 > 0 such that
the system of equations

x1 + x2 = c1 , x1 x2 = c2

has real solutions?

96 Problems and Solutions

Problem 27. Let m, n be positive integers. Find all solution of the

system of equations

2mn + (m2 + n2 ) = (m2 − n2 )2 , m − n = 1.

Problem 28. Let x ∈ R.

(i) Find all solutions of 2x = |x| + 1.
(ii) Find all solutions of 2x = −|x| + 1.

Problem 29. Let x ∈ Z. Solve the equation

x2 − 2x + 2 = 0 mod 5.

Problem 30. Let a > b > 0. Find the points of intersections of the two
ellipses
x21 x22 x21 x22
+ = 1, + = 1.
a2 b2 b2 a2

Problem 31. Let r1 ≥ 0, r2 ≥ 0, r3 ≥ 0. Find the solutions of the

system of equations

r1 + r2 + r3 = 1, r12 + r22 + r32 = 1.

Problem 32. Consider the cubic equation

p(x) = x3 − 6x2 + 11x − 6 = 0.

Let x1 , x2 , x3 be the roots. Given the equations

x1 + x2 + x3 = −a1 , x1 x2 + x2 x3 + x3 x1 = a2 , x1 x2 x3 = −a3

Find a1 , a2 , a3 .

Problem 33. Let z ∈ C. Find all solutions of

z + z ∗ + zz ∗ = 0.

Set z = reiφ with r ≥ 0 and φ ∈ [0, 2π). An alternative would be setting

z = x + iy with x, y ∈ R.

Problem 34. We want to find the positive root of f (x) = 0 with

f (x) = x3 − x2 − x − 1.
 Equations 97

We write x3 − x2 − x − 1 = 0 as x = 1 + 1/x + 1/x2 and set

xt+1 = 1 + 1/xt + 1/x2t , t = 0, 1, . . .

with x0 = 1. Do we find the positive root?

Problem 35. Find all 2 × 2 invertible matrices S over R with det(S) = 1

such that
       
0 1 0 1 0 1 0 1
S = S S S −1 = .
0 1 0 1 0 1 0 1

Thus we have to solve the three equations

s21 = 0, s11 + s12 = s22 , s11 s22 = 1.

Problem 36. Let z ∈ C. Find all solutions of

z + z ∗ + zz ∗ = 0.

Set z = reiφ with r ≥ 0.

Problem 37. Let a be a real constant. Solve the two equations

−a3 + x3 − 3ax2 + 3a2 x1 = 0, x2 − 2ax1 + a2 = 0

with respect to x1 , x2 , x3 .

Programming Problems

Problem 1. Let  ∈ R. Consider the quadratic equation

x2 − x +  − 1 = 0.

Find the roots x1 (), x2 (). Then find the mimimum of x21 +x22 with respect
to .

A SymbolicC++ program to solve this problem is:

// quadratic.cpp

#include <iostream>
#include "symbolicc++.h"
using namespace std;
98 Problems and Solutions

int main(void)
{
Symbolic x("x"), eps("eps"), f = 0;
Equations soln = solve((x^2)-eps*x+eps-1==0,x);
cout << "Solutions: " << endl << soln << endl;
Equations::iterator i;
for(i=soln.begin();i!=soln.end();i++) f += (i->rhs^2);
cout << "f(eps) = " << f << endl;
Equations min = solve(df(f,eps)==0,eps);
for(i=min.begin();i!=min.end();i++)
if(double(df(f,eps,2)[*i]) > 0)
cout << "Minimum at " << *i << endl;
return 0;
}
/*
Solutions:
[ x == eps-1,
x == 1 ]
f(eps) = eps^(2)-2*eps+2
Minimum at eps == 1
*/

5.2 Supplementary Problems

Problem 1. Let x, y ∈ R. Find all solutions of

e−x−y = e−x + e−y .

Problem 2. Solve the system of nonlinear equations

x1 x2 = 1, x2 x3 = 2, x3 x1 = 3.

Extend to the general case

x1 x2 = c1 , x2 x3 = c2 , . . . , xn−1 xn = cn−1 , xn x1 = cn

where cj (j = 1, . . . , n) are positive constants.

Problem 3. Find the solution of the equation

√
e−x − 1 = 2( 2 + 1).
 Equations 99

Problem 4. Let n be a positive integer. Solve the nonlinear equation

1 − ne−x x
−x
=
1 + ne 2
for n = 1, 2, 3, 4, 5.

Problem 5. Let c1 > 0 and c2 > 0. What is the condition on c1 , c2 such

that
x1 + x2 = c1 , x1 x2 = c2
has a real solution?

Problem 6. Let n be a positive integer. Solve

1 − ne−x x
=
1 + ne−x 2
for n = 1, 2, 3, 4, 5.

Problem 7. Let x > 0. Show that solution of the equation

x(1 + x1/2 ) = (1 − x1/2 )

is given by
1 √ √
x= ((17 + 3 33)1/3 + (17 − 3 33)1/3 − 1)2 ≈ 0.2955977 . . .
9

Problem 8. Let x1 , x2 , x3 ∈ R. Find solutions of the equation

x21 x22 + x22 x23 + x23 x21 = x1 x2 x3

with x1 6= 0, x2 6= 0, x3 6= 0. Note that

x1 = x2 = x3 = 1/3, x1 = 1/3, x2 = x3 = −1/3, x1 = x2 = −1/3, x3 = 1/3, x1 = x3 = −1/3, x2 = 1/3

are solutions.

Problem 9. (i) Consider the function

f (z, z̄) = z + z̄ + z z̄.

Find all solutions of f (z, z̄) = 0.

(ii) Let A be an n × n matrix over C. Find all solutions of

A + A∗ + AA∗ = 0n .
100 Problems and Solutions

Find all 2 × 2 matrices A such that

A + A∗ + AA∗ = I2 .

Problem 10. (i) Let φ, θ ∈ [0, 2π). Consider the equation

2η
eiφ sin(θ) = .
|η|2 + 1

Let η = 0, 1, −1, i, −i. Find φ and θ.

(ii) Let θ ∈ [0, 2π). Consider the equation

|η|2 − 1
cos(θ) = .
|η|2 + 1

Let η = 0, 1, −1, i, −i. Find φ and θ.

Problem 11. Show that roots of the polynomial f : R → R

f (x) = x4 − x3 − 10x2 − x + 1

are given by
1 √ 1 √ √ √
x1 = ( 5 − 3), x2 = (− 5 − 3), x3 = 2 + 3, x4 = 2 − 3.
2 2

Problem 12. Find the real root of the polynomial f : R → R

f (x) = x3 + x − 1.

The real root lie in the interval [0, 1]. Note that f (0) = −1 and f (1) = 1.

Problem 13. Find solution solutions of the equation (Scherk’s surface)

cos(x2 )ex3 = cos(x1 ).

For example (x1 , x2 , x3 ) = (0, 0, 0) is a solution and (x1 , x2 , x3 ) = (π/3, π/3, 0)

is a solution.

Problem 14. Let x > 0. Find all solutions of

√
q
x = 1 + 2 + x.

First show that x4 − 2x2 − x − 1 = 0.

 Equations 101

Problem 15. Show that the system of equations

3x + y − z + u2 = 0
x − y + 2z + u = 0
2x + 2y − 3z − 2u = 0

can be solved for x, y, u in terms of z; for x, z, u in terms of y; for y, z, u

in terms of x; but not for x, y, z in terms of u.

Problem 16. Are there solutions of the equation (z ∈ C)

sin(z) = ez and cos(z) = ez ?

Problem 17. (i) Show that cos(π/7) is a real root of

8x4 + 4x3 − 8x2 − 3x + 1 = 0.

(ii) Show that

cos(π/7) cos(2π/7) cos(4π/7) = −1/8.

Problem 18. Draw the curve in the plane given by

sinh(x) = exp(−2y).

Note that if y = 0, then sinh(2x) = 1 and therefore 2x = arcsinh(1).

Problem 19. Solve the system of integral equations

Z 2π Z 2π
(f (φ))3 cos(φ)dφ = 0, (f (φ))3 sin(φ)dφ = 0.
0 0

Problem 20. Consider the two circles in the plane

(x1 + 4)2 + (x2 + 5)2 = 194, (x1 − 3)2 + (x2 − 2)2 = 40.

Show that the two circles intersect and find the points of intersection. Show
that x1 + x2 = 9.
Chapter 6

Normed Spaces

Let C be the field of complex numbers. Let V be a vector space (linear

space) over C. Then V is a normed linear space if for every f ∈ V there is
a real number kf k such that (c ∈ C)

(i) kf k ≥ 0

(ii) kf k = 0 if and only if f = 0

(iii) kcf k = |c| kf kfor every c ∈ C
(iv) kf + gk ≤ kf k + kgk.
A normed linear space V which does have the property that all Cauchy
sequences are convergent is said to be complete. A complete normed linear
space is called a Banach space.

102
 Normed Spaces 103

6.1 Solved Problems

Problem 1. Consider the Banach space R4 and the four vectors
1 0 0 0
       
 x   1   0  0
v1 =  2  , v2 =   , v3 =   . v4 =  
x /2 x 1 0
x3 /3 x2 2x 2
where x ∈ R. Show that the vectors are linearly independent. Find the
distances

kv1 − v2 k, kv2 − v3 k, kv3 − v4 k, kv4 − v1 k.

Problem 2. Consider the Hilbert space C2 and the vectors

   
cos(i) − sin(i)
v1 = , v2 = .
sin(i) cos(i)

Find the distance kv1 − v2 k.

Problem 3. (i) Let a, b ∈ R. Show that

d(a, b) := | arctan(a) − arctan(b)|

defines a distance in R.
(ii) Show that xn = arctan(n), (n ∈ N) is a Cauchy sequence. Is the metric
space { R , d } complete?

Problem 4. Let n ≥ 1, 0 ≤ a < b and p ∈ Rn . Show that there exists

a map k : C ∞ (Rn , R) such that k(x) = 0 for kx − pk ≥ b, k(x) = 1 for
kx − pk ≤ a, and 0 < k(x) ≤ 1 for kx − pk ≤ b.

Problem 5. Let x ∈ R2 . Is

kxk := |x1 x2 |1/2

a norm on R2 ?

Problem 6. Show that if a function f : R → R satisfies

|f (x) − f (y)| ≤ M (|x − y|)a

for some fixed M > 0 and a > 1, then f is a constant function, i.e., f is
identically equal to some real number b for all x ∈ R.
104 Problems and Solutions

Problem 7. Let f , g be continuously differentiable functions on the

interval [0, 1]. One defines
Z 1 
df dg
hf, gi = f (x)g(x) + dx.
0 dx dx

Show that this satisfies the properties of an inner product. Calculate hf, gi
for f (x) = sin(x), g(x) = cos(x). Extend it to
 
Z 1 X n j j g(x)
d f (x) d
hf, gi =   dx.
0 j=0
dxj dxj

Problem 8. The p-norm of a vector (x1 , . . . , xn ) ∈ Rn is defined as

 1/p
Xn
kxkp :=  |xj |p 
j=1

where p ∈ R+ . Find the norm for p → ∞.

Problem 9. Let x ∈ Rn and kxk be the Euclidean norm of x. If B ⊂ Rm

is a nonempty compect set and x ∈ Rn , then we define the distance

dist(x, B) := min{ kx − bk : b ∈ B }.

If A, B ⊂ Rn are nonempty compact sets then we define the distance

dist(A, B) := max{ dist(a, B) : a ∈ A }.

Show that
dist(A, B) = 0 ⇔ A ⊂ B
and dist(A, B) <  means that A ⊂ N (B) where

N (B) := { x : dist(x, B) <  }

is the -neighborhood of B.

Problem 10. Let v and w be two normalized column vectors in Cn .

Does s !
(v∗ w)(w∗ v)
D(v, w) := 2 arccos
(v∗ v)(w∗ w)
provide a distance measure between v and w.
 Normed Spaces 105

Problem 11. Let z1 , z2 ∈ C. One can define a distance measure via

|z1 − z2 |
ρ(z1 , z2 ) = p
(1 + |z1 |2 )(1 + |z2 |2 )

Let z1 = eiφ1 and z2 = eiφ2 . Find ρ(z1 , z2 ).

Problem 12. The chordal distance between to complex numbers z1 , z2

is defined as
|z1 − z2 |
d(z1 , z2 ) := p p .
1 + |z1 |2 1 + |z2 |2
Let z1 = eiπ/2 and z2 = e−iπ/2 . Find d(z1 , z2 ).

Problem 13. Find n × n matrices A, B such that

k[A, B] − In k → min

where k.k denotes the norm and [, ] denotes the commutator.

106 Problems and Solutions

6.2 Supplementary Problems

Problem 1. Let x, y ∈ R. Is

|x − y|
d(x, y) =
2 + |x − y|

a metric on R? Note that d(x, x) = 0 and d(x, y) ≥ 0.

Problem 2. Let R be the set of real numbers and x, y ∈ R. Show that

d(x, y) = |x − y|

provides a distance function. Let x = (x1 , . . . , xn ) ∈ Rn and y = (y1 , . . . , yn ) ∈

Rn Show that v
u 
u X n
d(x, y) = t (xj − yj )2 
u
j=1

provides a distance function. Of course one always assume the positive

square root.

Problem 3. Consider the functions f : R → R, g : R → R

f (x) = exp(−x2 ), g(x) = exp(−|x|).

Find the distance d(f, g) between f and g given by

|f (x) − g(x)|
Z
d(f, g) = dx.
R 1 + |f (x) − g(x)|
Chapter 7

Complex Numbers and

Complex Functions

7.1 Solved Problems

Problem 1. Let z ∈ C. Solve the nonlinear equation

z 3 = z|z|2 .

Problem 2. (i) Find the complex numbers z satisfying z 2 = z̄.

(ii) Find the complex numbers z satisfying z 3 = z̄.

Problem 3. Solve the cubic equation z 3 = −1. Do the solutions form

a group under multiplication? If not, what numbers have to be added to
form a group. Find them by multiplication of the solutions of the cubic
equation.

Problem 4. Find the square root of z = 4 exp(iπ/4).

Problem 5. Let φ ∈ [0, 2π). Consider the complex number z = 1 − eiφ .

Find the condition on φ such that |z| = 1.

107
108 Problems and Solutions

Problem 6. Let x, y ∈ R and z = x + iy. Find the real and imaginary

part of i/(πz).

Problem 7. Let φ ∈ [0, 2π) and θ ∈ (−π, π). Can any complex number
be represented by  
iφ 1
z = e tan θ ?
2

Problem 8. Find

zdz̄ − z̄dz, (zdz̄ − z̄dz) ⊗ (zdz̄ − z̄dz).

Problem 9. Let x, c ∈ R and c 6= 0. Find the real and imaginary part of

the function
x − ic
fc (x) = .
x + ic

Problem 10. Let x, y ∈ R. Solve

x−iy−i xy − i
= .
x+iy+i xy + i

√
Problem 11. Find i.

Problem 12. (i) Calculate

p(α, β, θ, φ) = | cos(α) cos(β) sin(θ)eiφ + sin(α) sin(β) cos(θ)|2 .

(ii) Show that p(α, β, θ, φ) ≤ 1.

(iii) Simplify the result from (i) for θ = π/4 and φ = 0.

Problem 13. Let z1 , z2 ∈ C. Consider the distance measures

|z1 − z2 |
d1 (z1 , z2 ) := |z1 − z2 |, d2 (z1 , z2 ) := ,
1 + |z1 − z2 |

|z1 − z2 |
d3 (z1 , z2 ) := p .
(1 + |z1 |2 )(1 + |z2 |2 )
Here d3 is the chordal distance. Let z1 = eiφ and z2 = eiφ2 . Find d1 , d2 ,
d3 .
 Complex Numbers and Complex Functions 109

Problem 14. Let f (z1 ) and g(z2 ) be a pair of analytic functions of z1

and z2 , respectively. We define
Z 2π
1
f (z1 ) ◦ g(z2 ) := f (z1 eiθ )g(z2 e−iθ )dθ.
2π 0
Let
∞
X ∞
X
f (z1 ) = ak z1k , g(z2 ) = bk z2k
k=0 k=0

for |zj | < R, j = 1, 2. Find f (z1 ) ◦ g(z2 ).

Problem 15. Let z = x + iy, where x, y ∈ R. Find

∂ ∂ ∂ ∂
, , , .
∂z ∂ z̄ ∂x ∂y

Problem 16. Study the behaviour (stability) of the fixed points of the
complex map f : C → C, f (z) = z 2 .

Problem 17. Let z = x + iy (x, y ∈ R) be a nonzero complex number.

We define the principal argument by z = |z| exp(iarg(z)), where arg(z) ∈
(−π, π] and we define the imaginary remainder Imr(z) and the imaginary
quotient Imq(z) by
=(z) = Imr(z) + 2πImq(z)
where Imr(z) ∈ (−π, π] and Imq(z) ∈ Z. Show that

ln(ez ) = <(z) + iImr(z)

and in particular

ln(ez ) = z iff =(z) ∈ (−π, π].

Problem 18. Let z ∈ C and consider the analytic map

f (z) = exp(z).

Find the solutions of the equation

z = f (z).

This are the fixed points of f . We set z = x + iy (x, y ∈ R. Then

x + iy = exp(x + iy) ≡ ex eiy = ex (cos(y) + i sin(y)).

110 Problems and Solutions

Thus we have to solve

ex cos(y) − x = 0, ex sin(y) − y = 0.

Problem 19. Let A be an n×n matrix. Suppose f is an analytic function

inside on a closed contour Γ which encircles λ(A), where λ(A) denotes the
eigenvalues of A. We define f (A) to be the n × n matrix
I
1
f (A) = f (z)(zIn − A)−1 dz.
2πi Γ
This is a matrix version of the Cauchy integral theorem. The integral is
defined on an element-by-element basis f (A) = (fjk ), where
I
1
fjk = f (z)eTj (zIn − A)−1 ed z.
2πi Γ
Let f (z) = z 2 and  
0 1
A= .
1 0
Calculate f (A).

Problem 20. Let x1 , y1 , x2 , y2 ∈ R. Consider the two complex numbers

z1 = x1 + iy1 , z2 = x2 + iy2 .

In polar form we have

z1 = r1 eiφ1 = r1 (cos(φ1 ) + i sin(φ1 )), z2 = r2 eiφ2 = r2 (cos(φ2 ) + i sin(φ2 ))

where r12 = x21 + y12 , r22 = x22 + y22 . Now

z = z1 + z2 = (x1 + x2 ) + i(y1 + y2 ).

Find z in polar form z = reiφ .

Problem 21. Let a, b, φ be fixed real numbers. Consider the function

w(z) = z + az 2 + beiφ z 3 .

Write it as real and imginary part, with w = u + iv and z = x + iy.

Problem 22. Consider the z-transform

∞ I
X
−n 1
x(z) = x(n)z , x(n) = x(z)z n−1 dz.
n=0
2πi
 Complex Numbers and Complex Functions 111

Let
N
X
S(N ) := x(n).
n=1
Then
N I N
X 1 X
S(N ) = x(n) = x(z) z n−1 dz.
n=1
2πi n=1

It follows that (geometric series)

x(z)(z N − 1)
I
1
S(N ) = dz.
2πi z−1
Apply this expression and the residue theorem to calculate
N
X
S(N ) = n3 .
n=1

Problem 23. (i) Consider the complex number z = eiφ . Let n ∈ N. Find
√
n
z.
(ii) Let z = reiφ and w = x + iy (x, y ∈ R). Find z w .

Problem 24. Consider the complex numbers z1 = 0.4 + 0.3i, z2 = 5 + 2i.

Calculate z1z2 . Hint: Set z1 = r1 eiφ1 .

Problem 25. Consider the complex numbers

z1 = x1 + iy1 = 1 + 4i, z2 = x2 + iy2 = 3 − 2i.
Calculate logz2 (z1 ).

Problem 26. Let A > 0 and B ≥ 0. Consider the quadratic conformal

map in the complex w-plane of the unit disc in the complex z-plane
w(z) = Az + Bz 2 .
Let θ be the polar angle in the complex z-plane.
(i) Show that with the notation w := u + iv, z := x + iy (with u, v, x, y
real) one has the parametric equation of the boundary
u(θ) = A cos(θ) + B cos(2θ), v(θ) = A sin(θ) + B sin(2θ).
(ii) Let C := B/A. Show that for C = 0 (i.e. B = 0) one obtains a circular
disc. Show that for C = 1/4 the curvature vanishes at θ = π. Show that
for C = 1/2 the derivative dw/dz vanishes at the boundary, i.e. at θ = π.
112 Problems and Solutions

Problem 27. (i) Solve the equation

!4
i
z+ 2
i
= 1.
z− 2

(ii) Solve the system of equations

!4 !4
i i
z1 + 2 z1 − z2 + i z2 + 2 z2 − z1 + i
i
= , i
= .
z1 − 2
z1 − z2 − i z2 − 2
z2 − z1 − i

These equations play a role for the Bethe ansatz for spin systems.

Problem 28. Let z be a complex number such that |z| < ∞ and <(z) > 0.
Consider Z ∞
Z(z) = exp(−zx − x2 )dx.
0

Show that
∞
1X Γ( 1 + 1 k)
Z(z) = (−1)k 2 2 z k .
2 k!
k=0

Problem 29. Let z 6= 0. Show that the function

ln(z)
f (z) =
z−1
is analytic near z = 1 and admits the Taylor expansion
∞
X (−1)n
f (z) = (z − 1)n for |z − 1| < 1.
n=0
n + 1

Problem 30. Is the function f : C → C

f (z) = z + |z|

continuous? Find the fixed points of f .

Problem 31. Let x1 , x2 , y1 , y2 ∈ R and z1 = x1 + iy1 , z2 = x2 + iy2 with

z1 z̄1 6= 0, z2 z̄2 6= 0. Find the conditions such that
 
z2
( z̄1 z̄2 ) = 0.
−z1
 Complex Numbers and Complex Functions 113

Problem 32. Let z = ρeiθ and ζ = Reiω with with R > 0 and ρ > 0.
Show that
ρ2 − R 2
 
z+ζ
= < .
ρ2 − 2Rρ cos(ω − θ) + R2 z−ζ

7.2 Supplementary Problems

Problem 1. Show that

1 1
(−1 + 2i) = √ ei(π−arctan(2)) .
5 5

Problem 2. Let z1 = r1 eiφ1 , z2 = r2 eiφ2 with r1 , r2 ≥ 0. Show that

r2 + r22
 
z1 + z2
< = 1 .
z1 − z2 |z1 − z2 |

Problem 3. Let z = x + iy with x, y ∈ R. Show that

<(z 2 ) = x2 − y 2 .

Problem 4. Let 0 < r < 1 and φ ∈ [0, 2π). Consider the map

f (z) = reiφ + (1 − reiφ )z.

Find the fixed points. Find f (0), f (f (0)), f (f (f (0))).

Problem 5. Let x, y ∈ R. Find real and imaginary part of

s
x − iy iφ
z= e .
x + iy

Problem 6. Let z ∈ C. Show that

∞ 
z2
Y 
sin(z) = z 1− 2 2 .
k π
k=1
114 Problems and Solutions

Problem 7. Show that the Bessel function

∞
X (−1)n  z 2n+ν
Jν (z) =
n=0
n!Γ(ν + n + 1) 2

is an entire function of z for ν = 0, 1, . . .. Show that

1
Jν−1 (z) + Jν+1 (z) = 2ν Jν (z).
z

Problem 8. (i) Show that the radius of convergence of the function

∞
X
f (z1 , z2 ) = z1k z2
k=0

is given by { (z1 , z2 ) : |z1 | < 1 ∨ z2 = 0 }.

(ii) Show that the radius of convergence of the function
∞
X
f (z1 , z2 ) = (z1 z2 )k
k=0

is given by { (z1 , z2 ) : |z1 | · |z2 | < 1 }.

(iii) Show that the radius of convergence of the function
∞ X
k1  
X k1 k2 k1 −k2
f (z1 , z2 ) = z z
k2 1 2
k1 =0 k2 =0

is given by { (z1 , z2 ) : |z1 | + |z2 | < 1 }.

Problem 9. (i) Show that the sum

∞
X
k!z k
k=0

only converges for z = 0.

(ii) Show that the sum
∞
X zk
k2
k=0

converges for |z| ≤ 1 and diverges for |z| > 1.

(iii) Show that the sum
X∞
(z1 z2 )k
k=0
 Complex Numbers and Complex Functions 115

converges for { (z1 , z2 ) : |z1 | · |z2 | < 1 }.

(iv) Show that the sum
∞ X
k1  
X k1 k2 k1 −k2
z z
k2 1 2
k1 =0 k2 =0

converges for { (z1 , z2 ) : |z1 |+|z2 | < 1 }. The sum appears at the expansion
of 1/(1 − (z1 + z2 )).

√
Problem 10. Let x ∈ R+ and f (x) = x. Show that real function
can be extended with a power series expansion around 1 into the complex
domain
∞  
p X 1/2
f (z) = 1 + (z − 1) = (z − 1)k
k
k=0
iφ
√
and for z = re with −π/2 < φ < π/2 we have f (z) = reiφ/2 .

Problem 11. (i) Is the complex function

f (z) = z = x1 − ix2

holomorph?
(ii) Is the complex function

f (z) = z 2 − z 2 = 4ix1 x2

holomorph?

Problem 12. (i) Show that if n is a positive integer then

(r(sin(φ) + i sin(φ)))n ≡ rn (cos(nφ) + i sin(nφ)). (1)

Hint. Apply exp(iφ) ≡ cos(φ) + i sin(φ).

(ii) Show that the n-th roots of unity are
   
2π 2π
ρ = cos + i sin , ρ2 , ρ3 , . . . , ρn−1 , ρn ≡ 1. (1)
n n

Problem 13. Let  ∈ R and z ∈ C. Consider the product

∞
Y 1 − e−k
f (z) = .
k=1
1 − e−(k+z)
116 Problems and Solutions

Find f (0). Show that

f (z) = (1 − e−z )f (z − 1).

Show that
m
Y
f (m) = (1 − e−k ), m = 1, 2, . . .
k=1

Show that f is periodic with period 2πi/. Show that f has simple poles
at z = −m (m = 1, 2, . . .). Show that the residues are given by

(−1)m+1 exp(− 21 m2 + 12 m)

lim (z + m)f (z) = .
z→−m f (m − 1)

Problem 14. Show that

(1 + i)1/2 = ±21/4 (cos(π/8) + i sin(π/8)) .

Problem 15. Let n = 0, 1, . . . and x ∈ R. Find the real and imaginary

part of the functions

1 (ix − 1)n
fn (x) = √ .
π (ix + 1)n+1

Problem 16. Show that

et
I
1 1
= dt
(n − k)! 2πi tn−k+1

where the integration contour is a small circle around the origin in the
complex plane.

Problem 17. We know that 0 ≤ | sin(x)| ≤ 1 and 0 ≤ | cos(x)| ≤ 1 for

x ∈ R. This is no longer true for sin(z) and cos(z) with z ∈ C.
(i) Let a > 0. Show that
q q
| sin(az)| = sinh2 (ay) + sin2 (ax), | cos(az)| = sinh2 (ay) + cos2 (ax).

(ii) Find all solutions of | sin(z)| = 2 and | cos(z)| = 2.

(iii) Find all solutions of cos(z) = i.
 Complex Numbers and Complex Functions 117

(iv) Let n ∈ N. Show that (x, y ∈ R)

q
| sin(n(x + iy))| = sin2 (nx) cosh2 (ny) + cos2 (nx) sinh2 (ny)
> | sinh2 (ny)| → ∞

as n → ∞ for any y 6= 0.
Chapter 8

Integration

8.1 Solved Problems

Problem 1. The time-average of a continuous function f is
Z T
1
hf i := lim f (t)dt.
T →∞ 2T −T

Find the time-average of the functions

f1 (t) = cos(ωt) sin(ωt), f2 (t) = cos2 (ωt), f3 (t) = sin2 (ωt).

Problem 2. Let j, k = 1, 2, . . .. Consider the function

Z 1
f (j, k) = xj−1 (1 − x)k−1 dx, j, k = 1, 2, . . . .
0

Thus f (1, 1) = 1. Is
k
f (j − 1, k + 1) = f (j, k), j≥2
j−1
and
f (j + 1, k) = f (j, k) − f (j, k + 1) ?
Prove or disprove.

118
 Integration 119

Problem 3. Let

1 for 0≤τ ≤1
x(τ ) =
0 otherwise

and 
1 for 0≤τ ≤1
h(τ ) = .
0 otherwise
Find the convolution integral
Z ∞
y(t) = x(τ )h(t − τ )dτ.
−∞

Problem 4. Let T > 0 and ω = 2π/T . Let m, n ∈ N and α, β ∈ R.

Calculate
Z T
1
I(α, β) = cm sin(mωt + φm − α)cn sin(nωt + φn − β)dt.
T 0

Problem 5. A cubic B-spline with uniform knot spacing, centered at the

orign, is given by
 1 3
 6 (2 − |x|) if 1 ≤ |x| < 2
1
B(x) = 6 ((2 − |x|) − 4(1 − |x|)3 )
3
if 0 ≤ |x| < 1
0 otherwise


Find the integral Z ∞

B(x)dx.
−∞

Problem 6. Find the normalization constant K from the condition

Z 2π Z π Z π
1 = 4K sin2 (ω/2)dω sin(θ)dθ dφ.
0 0 −π

Problem 7. Let  ∈ R. Calculate the integral

Z 2π
f () = exp(eiφ )dφ
0

by finding an ordinary differential equation for f together with the initial

conditions. Obviously f (0) = 2π.
120 Problems and Solutions

Problem 8. Let b > a. Find the mean and variance of random variable
x with uniform probabilty density function p
 1
if a≤x≤b
p(x) = b−a
0 otherwise

Problem 9. Consider a one-dimensional lattice (chain) with lattice con-

stant a. Let k be the sum over the first Brioullin zone we have
Z π/a
1 X a
F ((k)) → F ((k))dk = G
N 2π −π/a
k∈1.BZ

where
(k) = 0 − 21 cos(ka).
Using the identity
Z ∞
δ(E − (k))F (E)dE ≡ F ((k))
−∞

we can write
!
Z ∞ Z π/a
a
G= F (E) δ(E − (k))dk dE.
2π −∞ −π/a

Calculate Z π/a
g(E) = δ(E − (k))dk
−π/a

where g(E) is called the density of states.

Problem 10. Let

i1 (t) = I 2 sin2 (ωt), i2 (t) = I 2 sin2 (ωt + φ).

Calculate Z T
1
hi1 (t)i2 (t)i := lim i1 (t)i2 (t)dt
T →∞ T 0
and
hi1 (t)i2 (t)i − hi1 (t)ihi2 (t)i.

Problem 11. Let α > 0. Find

Z ∞
1
exp(−αx)dx = .
0 α
 Integration 121

Problem 12. Let f (t) be a continuous function. Show that

Z x Z ζ Z x
dζ f (t)dt = (x − t)f (t)dt.
0 0 0

Problem 13. (i) Calculate

Z ∞
sech(t)tanh(t) cos(t + t0 )dt.
−∞

(ii) Calculate Z ∞
sech2 (t)tanh2 (t)dt.
−∞

Problem 14. Let λ > 0. Calculate

Z ∞
sin(λt)
dt.
−∞ λ sinh(t)

Problem 15. Calculate

Z ∞
exp(−x2 + 2ixy)dx.
−∞

Problem 16. Calculate the integral

Z ∞
et
I(λ) = t 2
cos(t + λ)dt
−∞ (1 + e )

using the residue technique.

Problem 17. Let m be an non-negative integer. Find

Z π
dθ| cos2m+1 (θ)|.
0

Problem 18. Let  > 0. Calculate

Z ∞
k 2 dk
.
0 ek2 − 1
122 Problems and Solutions

Problem 19. The mother Haar wavelet is given by


 −1 for 0 ≤ t < 1/2
f (t) = +1 for 1/2 ≤ t < 1
0 otherwise


Find the Fourier transform

Z ∞
fˆ(ω) = e−iωt f (t)dt.
−∞

Problem 20. The Poisson wavelet is given by

 
d
f (t) = t + 1 P (t)
dt

where
1 1
P (t) = .
π 1 + t2
Find the Fourier transform of f .

Problem 21. Let m ∈ Z. Calculate

Z 2π
1
am = eimφ 2 cos(φ)dφ.
2π 0

Problem 22. Let 0 < α < 1. Find the integral

Z ∞ α−1
x
dx.
0 1+x

Problem 23. The linear one-dimensional diffusion equation is given by

∂u ∂2u
= D 2, t ≥ 0, −∞ < x < ∞
∂t ∂x
where u(x, t) denotes the concentration at time t and position x ∈ R. D
is the diffusion constant which is assumed to be independent of x and t.
Given the initial condition c(x, 0) = f (x), x ∈ R the solution of the one-
dimensional diffusion equation is given by
Z ∞
u(x, t) = G(x, t|x0 , 0)f (x0 )dx0
−∞
 Integration 123

where
(x − x0 )2
 
0 0 1
G(x, t|x , t ) = p exp − .
4πD(t − t0 ) 4D(t − t0 )

Here G(x, t|x0 , t0 ) is called the fundamental solution of the diffusion equation
obtained for the initial data δ(x − x0 ) at t = t0 , where δ denotes the Dirac
delta function.
(i) Let u(x, 0) = f (x) = exp(−x2 /(2σ)). Find u(x, t).
(ii) Let u(x, 0) = f (x) = exp(−|x|/σ). Find u(x, t).

Problem 24. Let ω0 > 0 be a fixed frequency and t the time. Calculate
Z +∞
1
fˆ(ω) = √ e−|ωt| e−iωt dt.
2π −∞

Problem 25. Let ω0 > 0 be a fixed frequency and t the time. Calculate
Z ∞
ˆ 1
f (ω) = √ e−|ω0 t| e−iωt dt.
2π −∞

Problem 26. The sum

n   
1X kπ
lim exp 2 cos
n→∞ n n+1
k=1

can be cast into the integral

Z n
1
lim exp(2 cos(πx))dx. (1)
n→∞ n 0

Calculate this integral.

Problem 27. Let f : R → R be an analytic function. The Dirichlet

integral identity is given by
Z u Z u−w2
f (u − w1 − w2 )w1µ1 −1 w2µ2 −1 dw1 dw2 =
0 0
Z u
Γ(µ1 )Γ(µ2 )
f (u − w)wµ1 +µ2 −1 dw.
Γ(µ1 + µ2 ) 0
−x
Let f (x) = e . Calculate the left and right-hand side of this identity.
124 Problems and Solutions

Problem 28. Let r1 > 0, r2 > 0. Find the integral

Z 1
dx
I(r1 , r2 ) = p .
−1 r12 + r22 − 2r1 r2 x

Problem 29. Show that

√
Z  2 
2 x
ea = dx exp − + 2ax .
R 2

Problem 30. Let u ≥ 0 and

1 √
ρ(u) = exp(− u).
2
Find Z −∞
ρn = un ρ(u)du.
0

Problem 31. Calculate the integral

Z Z p
I= x2 + 4y 2 dydx
D

where D is the domain bounded by the positive x-axis, the positive y-axis
and the parabola y 2 = 1 − x.

Problem 32. Calculate the definite integral

Z 1
sin(x2 )dx.
0

Problem 33. (i) Consider the wavelet (ω0 > 0)

2 2
ψ0 (t) = (eiω0 t − e−ω0 /2 )e−t /2
.

Show that Z ∞
ψ0 (t)dt = 0.
−∞

Hint: Use (a > 0)

∞
Z r
2 π (b2 −4ac)/4a
e−(ax +bx+c)
dx = e .
−∞ a
 Integration 125

(ii) We define
d
ψn (t) := ψn−1 (t), n = 1, 2, . . .
dt
Show that Z ∞
tk ψn (t)dt = 0, 0 ≤ k ≤ n.
−∞

Problem 34. Let b > a. Consider the integral

Z y=b
f (y)dy
y=a

where f is a continuous function in [a, b]. Apply the transformation

1
y(x) = ((b − a)x + a + b)
2
so that the integration range is between −1 and +1. Then the Gauss
quadrature can be applied which extends over the interval [−1, +1].

Problem 35. Let  > 0. Find f of the equation

Z t
exp(−t) = 1 − f (s)ds.
0

Problem 36. (i) Find the area of the set

S2 := { (x1 , x2 ) : 1 ≥ x1 ≥ x2 ≥ 0 }.
(ii) Find the volume of the set
S3 := { (x1 , x2 , x3 ) : 1 ≥ x1 ≥ x2 ≥ x3 ≥ 0 }.
Extend the n-dimensions.

Problem 37. Let ω1 , ω2 be real and positive. Find

Z ∞
1 1
J(ω1 , ω2 ) = √ exp(− x2 + iω1 x + iω2 x2 )dx.
2π −∞ 2

√
Problem 38. Let e be the eccentricity of an ellipse, i.e. 1 − e2 = b/a.
Let n ≥ 1. Calculate the integral
Z 2π
dx
1

n
0 e − cos(x)
126 Problems and Solutions

by making the substitution z = exp(ix). Apply the residue theorem.

Hint. We have
Z 2π
i(−1)n+1 2n z n−1 dz
Z Z
dx dz
1 n = 1 1 1 n =
n .
z 2 − 2e z + 1




0 e − cos(x) |z|=1 e − 2 z + z iz |z|=1

Problem 39. Let q 2 > 0 and (p − q)2 > 0. Find

Z ∞ Z ∞
2
exp(−xq )dx, exp(−x(p − q)2 )dx.
0 0

Problem 40. Find α > 0 such that

Z +∞
exp(−α|x|)dx = 1.
−∞

Afterwards calculate
Z ∞ Z ∞
x exp(−α|x|)dx, x2 exp(−α|x|)dx.
−∞ −∞

Problem 41. Let c > 0. Show that

Z ∞
√
   
1 2 1 1
exp cy = √ dx exp − x2 + cyx .
2 2π −∞ 2

Problem 42. The Gauss invariant for two given closed loops Cα and Cβ
in R3 parametrized by rα (s), rβ is defined by

drα (s) drβ (s0 ) rα (s) − rβ (s0 )

I I
1
G(Cα , Cβ ) := ds ds0 ×
4π Cα Cβ ds ds0 |rα (s) − rβ (s0 )|3

where × denotes the vector product. Find G(Cα , Cβ ) for the two curves

Cα : x21 + x23 = 1, Cβ : (x1 − 1)2 + x22 = 1.

Problem 43. Let x ∈ R. We define [x] as the integer part of x and

{x} := x − [x]. Calculate the integrals
Z 7/2 Z 7/2
[x]dx, {x}dx.
0 0
 Integration 127

Problem 44. Let x ∈ R. Consider the integral

Z ∞ iy
e
f (x) = dy.
x y
(i) Show that f (−|x|) = f (|x|) − iπ.
(ii) Show that for large x
 
ix i 1
f (x) = e + + ··· .
x x2

Problem 45. Let n = 0, 1, 2, . . .. We define

Z π/2 Z π/2
an := cos2n (x)dx, bn := x2 cos2n (x)dx.
0 0

Then a0 = π/2 and b0 = π 3 /24. Show that using integration by parts

an = (2n − 1)(an−1 − an ).
Show that for n ≥ 1 we have
an = (2n − 1)nbn−1 − 2n2 bn .
Show that (n ≥ 1)
n
π2 X 1 bn π2
0≤ − 2
=2 ≤ .
6 k an 4(n + 1)
k=1

Problem 46. Calculate

Z 10
I= (x − int(x))dx
0

where int(x) defines the largest integer less than x.

Problem 47. Let 0 < a < b < 1. Find

Z b
ln(1 − x)
dx.
a x

Problem 48. Show that

Z π/2
cos10 (x) cos8 (2x) cos6 (4x) cos4 (6x) cos2 (8x)dx =
0
128 Problems and Solutions

5166673π 2966549762512816
+ .
536870912 98120709987525225

Problem 49. Let z ∈ C. Let ln(1 + z) be the branch of the logarithm

defined on C \ (−∞, −1]. Calculate
Z
In (r) = P v z n−1 ln(1 + z)dz, r > 0, n ∈ Z
|z|=r

where P v is the principal value.

Problem 50. Let b > a and a, b be finite. Consider the integral

Z b
f (x)dx.
a

(i) Apply the transformation

 
1 2x − a − b
x = (a + b + (b − a) tanh(y) ⇔ y = tanh−1
2 b−a

to the integral.
(ii) Apply the transformation

1 π 
x= a + b + (b − a) tanh sinh(y)
2 2
with
dx (b − a)π cosh(y)/4
=
dy cosh2 (π sinh(y)/2)
to the integral.

Problem 51. Calculate the integral

Z π
1 + sin(x)
dx
0 1 + cos(x)

utilizing the transformation t = tan(x/2) with −π < x < π. From this

transformation it follows that
2
x = 2 arctan(t), dx = dt
1 + t2

and x = 0 → t = 0, x = π/2 → t = 1.
 Integration 129

Problem 52. Simplify the integral

Z 1
cos(x)
I= √ dx
0 x
for numerical calculation.

Problem 53. Let a > 0. Let f : R → R, g : R → R be continuous

functions with f (−x) = f (x) and g(−x) = −g(x). Show that
Z a
f (x)g(x)dx = 0.
−a

Problem 54. Show that

Z ∞ Z ∞
e−x x2 dx = 2!, e−x x3 dx = 3!.
0 0

Let n = 4, 5, . . .. Show that

Z ∞
e−x xn dx = n!.
0

Problem 55. Let k1 > 0, k2 > 0, k3 > 0. Find the integral

Z ∞
dr
sin(k1 r) sin(k2 r) sin(k3 r) .
0 r

Problem 56. Calculate

Z
F (s, t) = e−ixt−|x−s| dx.
R

Problem 57. Let x ∈ (0, 1). Calculate

Z 1
ym
P dy
0 (y − x)

for m = 0, m = 1, m = 2, m = 3.

Problem 58. Let a, b > 0 and b > |a|. Show that

Z
dx π
2 + 2ax + b2
=√ .
R x b − a2
2
130 Problems and Solutions

Problem 59. Let a, b > 0. Show that

Z ∞ −ax
− e−bx
 
e b
dx = ln .
0 x a

Problem 60. Let x > 0. Show that

Z ∞
1
e−tx dt = .
0 x

Problem 61. a, b ∈ R. Let

f (x) = a exp(bx).

Find the conditions on a and b such that

Z 1 Z 1
1
f (x)dx = 1, xf (x)dx = .
0 0 2

Problem 62. Dawson’s integral is given by

Z x
2 2
f (x) = et −x dt, x ≥ 0.
0

(i) Show that for all complex values z the function f satisfies the linear
differential equation
df (z)
+ zf (z) = 1.
dz
(ii) Let j = 1, 2, . . .. Show that

f (j+1) (z) + 2zf (j) + 2jf (j−1) (z) = 0, j = 1, 2, . . .

where f (j) indicates the jth derivative.

Problem 63. Let a, b, c ∈ R and a + b cos(θ) + c sin(θ) 6= 0. Show that

Z +π
b sin(θ) − c cos(θ)
dθ = 0.
−π a + b cos(θ) + c sin(θ)

Problem 64. Let a > 0. Show that

Z ∞
cos(ax) π
2
dx = e−a .
0 1 + x 2
 Integration 131

Problem 65. Show that

Z ∞
ln(cosh(x)) π
3 dx = (ln(2) − 1/2).
0 cosh (x) 4

Problem 66. Let m, n = 0, 1, 2, . . .. Find the integral

Z t
fmn (t) = (t − τ )m τ n dτ.
0

Problem 67. Let a, b > 0. Find the integral

Z ∞
cos(at) − cos(bt)
dt.
0 t

Problem 68. Let q 2 > 0. Calculate the integral

Z ∞
exp(−q 2 x)dx.
0

Problem 69. Let x > 0. Assume that f : R → R is integrable over [0, x]

for all x > 0 and limx→∞ f (x) = a. Show that

1 x
Z
lim f (s)ds = a.
x→∞ x 0

Utilize
 Z x   Z x   Z x
1  1  1

x f (s)ds − a = 
  (f (s) − a)ds ≤   |f (s) − a|ds.
0 x 0 x 0

Problem 70. Consider the function f : [0, 1] → [0, 1]


1/x − int(1/x) x 6= 0
f (x) =
0 x=0

where int(y) denotes the integer part of y. The function is integrable.

(i) Let k ≥ 1 be a positive integer. Find
Z 1/k
f (x)dx.
1/(k+1)
132 Problems and Solutions

(ii) Find
Z 1
f (x)dx.
1/k

(iii) Find
Z 1
f (x)dx.
0

Problem 71. Show that

Z π/2 p
sin(x) π
p p dx = .
0 sin(x) + cos(x) 4

Problem 72. Calculate

Z 3
1
(x − bxc + )dx
0 2

where bxc denotes the greatest integer less than or equal to x.

Hilbert Transform

Problem 73. The Hilbert transform H and its inverse is given by

Z
1 f (x)
g(y) = H(f (x)) = P dx
π R x −y
Z
1 g(y)
f (x) = H −1 (g(y)) = P dy
π R y−x

where the Cauchy principal value is defined by

Z Z +R
P f (x)dx := lim f (x)dx.
R R→∞ −R

The Hilbert transform relates parts of the function in the same domain.
Let k > 0. Find the Hilbert transform of f (x) = cos(kx).

Problem 74. Consider the Hilbert transform

Z ∞
1 f (x0 ) 0
H[f ] = P 0
dx .
π −∞ x − x
 Integration 133

Find H 2 .

Problem 75. The Hilbert transform acts on a one-dimensional function

g(s) by a convolution with the kernel 1/(πs). The singularity at s = 0 is
handled in the Cauchy principal value sense. The Fourier transform of the
Hilbert kernel is −i sgnσ. Thus the Hilbert transform of g is
Z ∞ Z ∞
g(s − s0 ) 0
Hg(s) = ds = (−i sgnσ)G(σ)ei2πsσ dσ
−∞ πs0 −∞

where G(σ) is the Fourier transform of g, i.e.

Z ∞
G(σ) = g(s)e−i2πσs ds.
−∞

Suppose g is a function whose support is strictly less than radius R, i.e.

g(s) = 0 for all |s| > R −  for some small positive . Find an inversion
formula.

Problem 76. The Hilbert transform of a function f ∈ L2 (R) is defined

as Z ∞
1 f (x)
H(f )(y) = P V dx
π −∞ x −y
where P V stands for principal value. Calculate the Hilbert transform of
f (x) = 1/(1 + x4 ).

Problem 77. The Radon transform for a function f (x, y) is given by the
integral transform
Z +∞
P (r, θ) = Rf (x, y) = f (r cos(θ) − s sin(θ), r sin(θ) + s cos(θ))ds.
−∞

The function P (r, θ) desribes the values of points on projections. Show that
the inverse Radon transform can be given by
1
f (x, y) = R−1 P (r, θ) = − BHDP (r, θ)
2π
where D is the partial differential operator Dg(r, θ) = ∂g/∂r with respect
to r, H is the Hilbert transform operator
Z
1 g(u, θ)
Hg(r, θ) = − du
π r−u
and B is the backprojection operator
Z π
Bg(r, θ) = g(x cos(θ) + y sin(θ), θ)dθ.
0
134 Problems and Solutions

Problem 78. The Hilbert transform of a function f ∈ L2 (R) is defined

as Z ∞
1 f (x)
H(f )(y) = P V dx
π −∞ x −y
where P V stands for principal value. Calculate the Hilbert transform of

f (x) = exp(−x2 /2).

Problem 79. Let f : R → R

x
f (x) = .
sinh(x)
Then f (0) = 1. Let c > 0. Find
Z c
f (x)dx.
0

Problem 80. Let p1 , p2 , p3 > 0. Calculate

4 ∞
Z
dr
f (p1 , p2 , p3 ) = sin(p1 r) sin(p2 r) sin(p3 r) .
π 0 r

Problem 81. Find the constant K (normalization) from the condition

Z 2π Z π Z π
2
1 = 4K sin (ω/2)dω sin(θ)dθ dφ.
0 0 −π

Problem 82. Consider

Z π Z π Z π
1
I= dα sin(β)dβ F (α, β, γ)dγ.
8π 2 −π 0 −π

(i) Find I for F (α, β, γ) = 1.

(ii) Find I for F (α, β, γ) = α + β + γ.
(iii) Find I for F (α, β, γ) = αβγ.

Problem 83. Consider the circle around (0, 0, 0) in the x1 − x2 plane

   
x1,1 (t) cos(t)
r1 (t) =  x1,2 (t)  =  sin(t)  , t ∈ [0, 2π]
x1,3 (t) 0
 Integration 135

and the circle around (0, 1, 0) in the x2 − x3 plane

   
x2,1 (s) 0
r2 (s) =  x2,2 (s)  =  1 + cos(s)  .
x2,3 (s) sin(s)

Then the derivatives are

   
− sin(t) 0
dr1 (t)  dr2 (t) 
= cos(t)  , = − sin(s)  .
dt dt
0 cos(s)

Calculate (Gauss formula)

 
r1 (t) − r2 (s)
I I
1 dr1 (t) dr2 (s)
dtds × ·
4π dt ds |r1 (t) − r2 (s)|3

where × denotes the vector product, · denotes the scalar product and con-
tour integrations run from 0 to 2π.

Problem 84. Show that

Z ∞ Z ∞
e−x x2 dx = 2!, e−x x3 dx = 3!.
0 0

Problem 85. (i) Find the area of the set

S2 := { (x1 , x2 ) : 1 ≥ x1 ≥ x2 ≥ 0 }.

(ii) Find the volume of the set

S3 := { (x1 , x2 , x3 ) : 1 ≥ x1 ≥ x2 ≥ x3 ≥ 0 }.

Extend the n-dimensions.

Problem 86. Let n be a positive integer. Find

Z π
cos2 (nx)dx.
0

Problem 87. Let n ≥ 0. The Legendre polynomial of degree n is defined

as
1 dn 2
Pn (x) = (x − 1)
2n n! dxn
136 Problems and Solutions

with P0 (x) = 1. Let n ≥ 1. Show that

Z 1
xk Pn (x)dx = 0
0

for k = 0, 1, . . . , n − 1

Problem 88. Let k = 0, 1, . . .. Calculate

2 π
Z
ck = arcsin(cos(θ)) cos(kθ)dθ.
π 0

Problem 89. Find α > 0 such that

Z ∞
3
2αxe−2αx /3 dx = 1.
0

Problem 90. The complete elliptic integral of first kind K(m) can be
defined by
Z 1
K(m) = ((1 − t2 )(1 − mt2 ))−1/2 dt, |m| < 1.
0

The beta function can be defined by

Z 1
B(a, b) = ta−1 (1 − t)b−1 dt, <(a) > 0, <(b) > 0.
0
√
Show that 2 2K(1/2) = B(1/4, 1/2).

√
Problem 91. Consider the function f : [0, 1] → [0, 1], f (x) = x with
1 1
√
Z
2 3/2  2
I= xdx = x  = .
0 3 0 3

Find a approximation (upper bound) of the integral in the sense of Lebesgue

by partioning the ordinate interval into the intervals [0, 1/4], (1/4, 1/2],
(1/2, 3/4], (3/4, 1] with a0 = 0, a1 = 1/4, a2 = 1/2, a3 = 3/4, a4 = 1.

Problem 92. Let a > 0. Calculate

Z
dx
.
(a2 − x2 )3/2
 Integration 137

Set x = a sin(θ). Then dx = a cos(θ)dθ.

Double Integrals

Problem 93. Let f : R2 → R be a continuous function and f (x1 , x2 ) =

f (x2 , x1 ) for all x1 , x2 ∈ R. Let b, a ∈ R and b > a. Calculate
Z b Z b
f (x1 , x2 ) sin(x1 − x2 )dx1 dx2 .
a a

Problem 94. Let f be a continuous function.

(i) Show that the double integral
Z x Z ξ
dξ f (t)dt
0 0

can be expressed by a single integral.

(ii) Show that (n ≥ 2)
Z x Z ξ1 Z ξn−1
dξ1 dξ2 · · · f (ξn )dξn
0 0 0

can be expressed by a single integral.

Problem 95. Calculate the integral

Z π/2
I= sin2n (θ) cos2n+1 (θ)dθ
0

by considering the double integral

Z Z
2
(r sin(θ))2n (r cos(θ)2n+1 e−r rdrdθ
D

where D is the first quadrant.

Problem 96. Find the integral

!
Z 2π Z | cos(θ)|(`/2)
I(`) = dp dθ.
0 0
138 Problems and Solutions

Problem 97. Evaluate the quadruple integral

Z 1 Z 1 Z 1 Z 1
0
I= dx dy dx dy 0
0 0 0 0
 
1 1
− .
((x − x0 )2 + (y − y 0 )2 )1/2 ((x − x0 )2 + (y − y 0 )2 + 1)1/2

Problem 98. Let z ∈ (0, 1] and x ∈ (0, 1]. Find the double integral
Z 1 Z 1
f (z) = 1 − dx dy.
z z/x

p
Problem 99. (i) Let r = x21 + · · · + x2n . Show that the integral

dx1 · · · dxn
Z Z Z
···
r≥1 rα

is finite for αp> n and infinite for α ≤ n.

(ii) Let r = x21 + · · · + x2n . Show that the integral

dx1 · · · dxn
Z Z Z
···
r≤1 rα

is finite for α < n and infinite for α ≥ n.

Problem 100. Consider a three-dimensional probability distribution

f (x1 , x2 , x3 ) such that for all j
!
1 −1 + 2x2j 2
fx (xj ) = √ 2
1 + 2xj − √ e−xj
2 π 2

where fx (x) is the probability density associated with an individual vari-

able. This means
Z Z
fx (x1 ) = f (x1 , x2 , x3 )dx2 dx3
R R

etc. Is it possible that the probabilty density associated with the sum of
these variables s = x1 + x2 + x3 is given by

−1 + 2s2
 
1 2
fs (s) = √ 1 + 2s2 + √ e−s
2 π 2
 Integration 139

provided that f (x1 , x2 , x3 ) is non-negative?

Problem 101. Consider a one-dimensional chain of length N with open

end boundary conditions. The counting is from left to right starting at
0. The canonical partition function Z(β) (β > 0) is given by the multiple
integral
Z 1 Z 1 Z 1
ZN (β) = ds0 ds1 · · · dsN −1 eβ|s0 −s1 | eβ|s1 −s2 | · · · eβ|sN −2 −sN −1 | .
−1 −1 −1

Show that there is a coordinate transformation which decouples the sites.

Find Z2 (β) and Z3 (β).

Problem 102. Let f : Rn → Rn be an analytic function. Consider the

map
xj = f (xj−1 ) = · · · = f (x0 ).
To study the evolution of phase-space distributions, we can introduce the
evolution operator U (x0 , x, j) such that any initial phase-space distribution
ρ(x, 0) evolves into
Z
ρ(x0 ; j) = U (x0 , x; j)ρ(x, 0)dx
Ω

where Ω is the phase space area. Find U (x0 , x; j).

Problem 103. Let

S := { (x, y) ∈ R2 : x, y ≥ 0, 0 ≤ x21 + x22 ≤ 1 }.
Let m, n be nonnegative integers. Find the integral
Z
x2m+1 y 2n+1 dxdy.
S

Problem 104. Calculate Z Z

dxdy
A
where
A = {(x, y) : x, y ≥ 0, x + y ≤ 1, x ≥ 1/3 }.

Problem 105. Calculate the integral

Z ∞ Z ∞ √ 2 p
x − 1 y2 − 1
I= dx dy
1 1 (x + y)6
140 Problems and Solutions

using the substitution x = cosh(α), y = cosh(β).

Problem 106. Find the integral

Z ∞ Z ∞ √ 2 p
x − 1 y2 − 1
I= dx dy .
1 1 (x + y)6
Hint. Use the substitutions x = cos(α), y = cosh(β) and show that the
integral can be written as
Z ∞ Z ∞
sinh2 (α) sinh2 (β)
I= dα dβ .
0 0 (cosh(α) + cosh(β))6

Problem 107. Calculate the double integral

Z πZ π
1
dαdα0 ln(2 − cos(α) − cos(α0 ))
2π 2 0 0
utilizing the identity
α + α0 α − α0
   
0
cos(α) + cos(α ) ≡ 2 cos cos
2 2
the transformation x = (α + α0 )/2, y = (α − α0 )/2 and the integral
Z π
ln(1 + sin(x))dx = −π ln(2) + G
0

where G is the Catalan constant.

Problem 108. Let a > 0 and R ≥ 0. Find

Z RZ R Z RZ R
dxdy dxdy
Ia (R) = 2 + a2
≡ 2 + y 2 − 2xy + a2
.
0 0 (x − y) 0 0 x

Problem 109. Calculate the integral

Z 1Z 1Z 1Z 1Z 1Z 1
dx1 dx2 dx3 dx4 dx5 dx6
.
0 0 0 0 0 0 1 + x 1 + x2 + x3 + x4 + x5 + x6

Problem 110. Let n1 , n2 ∈ Z. Consider a function f : R2 → R with a

period 2π for both x1 and x2
∞
X
f (x1 , x2 ) = cn1 ,n2 ei(n1 x1 +n2 x2 )
n1 ,n2 =−∞
 Integration 141

and Z 2π Z 2π
f (x1 , x2 )dx1 dx2 = 0.
0 0

Then Z 2π Z 2π
1
cn1 ,n2 = f (x1 , x2 )e−i(n1 x1 +n2 x2 ) dx1 dx2
4π 2 0 0

(i) Let
f (x1 , x2 ) = sin(x1 ) + cos(x2 ).
Find cn1 ,n2 .
(ii) Let
f (x1 , x2 ) = sin(x1 ) cos(x2 ).
Find cn1 ,n2 .

Problem 111. Find a non-negative analytic function f : R2 → R such

that Z Z
f (x, y)dxdy = 1
R R

and Z Z Z Z
2 1 1
x dxdy = , y 2 dxdy = .
R R 2 R R 2

Problem 112. Consider the compact set

√
S := { (x, y) : y ≥ x2 , y ≤ x, x, y ≥ 0 }.

Thus S ⊂ [0, 1] × [0, 1]. Find

Z Z
I(S) = dµ = dxdy.
S S

Problem 113. Let a > 0. Consider the compact set (lemniscate)

S := { (x, y) : (x2 + y 2 )2 ≤ 2a2 xy }.

Find Z Z
I(S) = dµ = dxdy.
S S

Introduce polar coordinates x(r, φ) = r cos(φ), y(r, φ) = r sin(φ). Thus we

have
1
x2 + y 2 = r 2 , xy = r2 cos(φ) sin(φ) ≡ sin(2φ).
2
142 Problems and Solutions

Problem 114. Calculate the integrals

Z 1 Z 1−x2  Z 1 Z 1−x3 Z 1−x3 −x2 
dx2 dx1 dx2 , dx3 dx2 dx1 .
0 0 0 0 0

Extend to n-dimensions. Give an interpretation of the result.

Problem 115. (i) Is the subset of R2

S2 = { (x1 , x2 ) : x1 ≥ 0, x2 ≥ 0, x1 + x2 ≤ 1 }

convex?
(ii) Calculate Z
A= dx1 dx2 .
S2

(iii) The area of a triangle in the plane R2 with vertices at (x1 , y1 ), (x2 , y2 ),
(x3 , y3 ) is given by  
x1 y1 1
1
A = ± det  x2 y2 1
2
x3 y3 1
where the sign is chosen so that the area is nonnegative. Let (x1 , y1 ) =
(0, 0), (x2 , y2 ) = (1, 0), (x3 , y3 ) = (0, 1). Find A. Compare to (ii).
(iv) Is the subset of R3

S3 = { (x1 , x2 , x3 ) : x1 ≥ 0, x2 ≥ 0, x3 ≥ 0 }

convex?
(v) Calculate Z
V = dx1 dx2 dx3 .
S3

Problem 116. Consider

Z π Z π Z π
1
I= dα sin(β)dβ F (α, β, γ)dγ.
8π 2 −π 0 −π

Find I for F (α, β, γ) = 1, F (α, γ, β) = α + β + γ, F (α, β, γ) = αβγ.

Problem 117. Let γ > 0. A random variable X is said to be Lorentzian

with parameters (α, γ) if its probability density is given by

1 γ
fX (x) = , x ∈ R.
π (x − α)2 + γ 2
 Integration 143

Let X, Y be two independent Lorentzian random variables with parameters

(α, γ) and (β, δ), respectively. Let  ≥ 0. Show that
(a) X is distributed Lorentzian (α, γ)
(b)  + X is distributed Lorentzian ( + α, γ)
(c) −X is distributed Lorentzian (−α, γ)
(d) X + Y is distributed Lorentzian  (α + β, γ + δ)

γ
(e) X −1 is distributed Lorentzian α
α2 +γ 2 , α2 +γ 2

Problem 118. A random variable X is said to be Lorentzian if its prob-

ability density pX is a the form
1 γ
pX (x) =
π (x − α)2 + γ 2

where γ > 0. We say that X is (α, γ) to indicate that the random variable
is Lorentzian with the probability density given above. Let X, Y be two
independent random variables with (α, γ) and (β, δ). Let  be a nonnegative
real number. Show that
X is (α, γ)
+X is ( + α, γ)
−X is (−α, γ)
X +Y is (α + β, γ + δ)
 
−1 α γ
X is , .
α2 + γ ∗ 2 α2 + γ 2

Problem 119. Consider the set

M = { (x1 , x2 , x3 ) : x21 + x22 ≤ 1 and x21 + x23 ≤ 1 }

which is subset of R3 . Obviously M is compact and measureable

Z Z Z
I(M ) = dµ = χM dµ = χM dx1 dx2 dx3
M R3 R3

where χM is the indicator function. Find I(M ).

Problem 120. Given a smooth Hamilton function

n
X p2j
H(p, q) = + U (q)
j=1
2
144 Problems and Solutions

with n degrees of freedom (p = (p1 , . . . , pn ), q = (q1 , . . . , qn ). Let V (E) be

the classical phase space volume at energy E of a smooth Hamilton function
is given by
Z
V (E) = Θ(E − H(p, q))dn pdn q
R2n

where Θ is the step function. Assume that U (q) = m U (q).

(i) Consider the transformation

p = E 1/2 p0 , q = E 1/n q0

with the inverse transformation

p0 = E −1/2 p, q0 = E −1/n q.

Find dn p0 dn q0 and H(p0 , q0 ).

(ii) Calculate V (E) with the assumption that E > 0. Find the asymptotic
behaviour.

Problem 121. Let a > 0. Find the area of the surface in R3 given by
the intersection of a hyperbolic paraboloid x3 = x1 x2 /a and the cylinder
x21 + x22 = a2 . We have the parameter representation

x1 x2 1
x1 (r, φ) = r cos(φ), x2 (r, φ) = r sin(φ), x3 (r, φ) = = r2 cos(φ) sin(φ).
a a
1
Note that sin(φ) cos(φ) ≡ 2 sin(2φ).

8.2 Programming Problems

Problem 1. Calculate the integral

Z 1
| cos(2πx)|dx
0

using the random numnber generator described in problem 7, chapter 10,

page 250, Problems and Solutions in Scientific Computing. Compare to the
exact result by solving the integral.
 Integration 145

8.3 Supplementary Problems

Problem 1. Let α > 0. Show that

Z ∞
2
sech(αt)dt = .
−∞ α

Problem 2. Consider the function φ : R → R


1 for x ∈ [0, 1]
φ(x) :=
0 otherwise
Find
ψ(x) := φ(2x) − φ(2x − 1).
Draw the function. Calculate
Z +∞
ψ(x)dx.
−∞

Problem 3. Show that

Z ∞ 
2 1 1
e−x /2
= exp − y 2 + ixy dy.
2π −∞ 2

Problem 4. Let ` > 0 and r0 > 0. Find the integral

Z r0
r3
p dr.
0 1 + r2 /`2

Problem 5. Let 0 ≤ r < 1. Consider the Hilbert space L2 [0, 2π] and
f (θ) ∈ L2 [0, 2π]. Show that
Z 2π ∞
1 2π X j
Z
1
f (θ)dθ + r f (θ) cos(j(φ − θ))dθ
2π 0 π 0 j=1
2π
1 − r2
Z
1
= f (θ) dθ.
2π 0 1 − 2r cos(φ − θ) + r2

Problem 6. Show that

Z ∞  2
x + (y − γ)2

dk 1
e−kx sin(ky) sin(kγ) = ln .
0 k 4 x2 + (y + γ)2
146 Problems and Solutions

Note that Z ∞
dk df 1
f (x) = e−kx ⇒ =− .
0 k dx x

Problem 7. Show that

∞
x2 x(x2 − a2 )
Z
1
= 2 2 + 2 arctan(x/a).
0 (x2 2
+a ) 3 8a (x + a2 )3 8a

Problem 8. Show that

3 π/6
Z
ln(2 cos(x))dx = 0.338314... .
π 0

Problem 9. The content (n-dimensional volume) bounded by a hyper-

sphere of radius r is known to be
2rn π n/2
Vn =
nΓ(n/2)
where Γ is the gamma function. Let r = 1. Show that
lim Vn = 0.
n→∞

Problem 10. Let k = 0, 1, 2, . . . and

Z 1
xk
yk := 2
dx.
0 1+x+x

Show that
1
yk+2 + yk+1 + yk = .
k+1
Note that
Z 1 x=1
1 2 2x + 1 
y0 = dx = √ arctan √
0 1 + x + x2 3 3 
x=0
Z 1 1 Z 1 1
x 1 2 
 1 1 1 2 
 1
y1 = 2
dx = ln(1 + x + x ) − 2
dx = ln(1 + x + x ) − y0 .
0 1+x+x 2 0 2 0 1+x+x 2 0 2

Problem 11. Let c ∈ C with |c| < ∞ and <(c) > 0. Consider the integral
Z ∞
I(c) = exp(−cx − x2 )dx.
0
 Integration 147

Show that
∞
1X Γ(1/2 + k/2) k
I(c) = (−1)k c .
2 k!
k=0

Problem 12. Find α > 0 such that

Z ∞
3
2αx2 e−2αx /3 dx = 1.
0

Problem 13. Let x = (x1 x2 · · · xn ), y = (y1 y2 · · · yn ) and x · y =

x1 y1 + · · · + xn yn . Show that
Z
1
exp(ay2 ) = n/2 exp(−x2 + 2a1/2 x · y)dx1 . . . dxn .
π R n

Problem 14. Let b > a. Show that

Z br    r 
y−a γ dy γ a
1− =π 1− .
a b−y y y−x x b

Problem 15. Let c > 0 and k ∈ R. Show that

Z ∞ √
1 π
exp(−cs2 ) cos(2ks)ds = √ exp(−k 2 /c).
0 2 c

This integral plays a role in optics.

Problem 16. Let x > 0. Show that

Z ∞ ∞
eu
Z
−s2
2 e ds = √
0 x2 u

Problem 17. Let x ≥ 0. Consider the function f defined by

Z x
ln(1 + y)
f (x) = dy.
0 y

Show that for small X we can write

x2 x3 x4
f (x) = x − + − + ···
4 9 16
148 Problems and Solutions

Show that
π2
 
1 1
f (x) = + ln2 (x) − f .
6 2 x
Give a value of f (1).

Problem 18. Calculating

Z
cos3 (x) sin3 (x)dx

student Alice tells you the result is

1 1
sin4 (x) − sin6 (x) + C1
4 6
and student Bob tells that
1 1
cos6 (x) − cos4 (x) + C2
6 4
is the result. C1 , C2 are the constants of integration. Discuss.

Problem 19. Let f : Rn+1 → R be an analytic function, gj : R → R

(j = 1, . . . , n) are analytic functions and c1 , . . . , cn are constants. Consider
Z g1 () Z g2 () Z gn ()
I() = dx1 dx2 · · · f (x1 , x2 , . . . , xn , ).
c1 c2 cn

Show that
Z g1 () Z g2 () Z gn ()
dI() ∂f (x1 , x2 , . . . , xn , )
= dx1 dx2 · · ·
d c1 c2 cn ∂
Z g2 () Z Z gn ()
dg1
+ dx2 g3 () · · · f (x1 , x2 , . . . , xn , )
d c2 c3 cn

dg2 g1 ()
Z Z g3 () Z gn ()
+ dx1 dx3 · · · f (x1 , x2 , . . . , xn , )
d c1 c3 cn
+···
dgn g1 ()
Z Z g2 () Z gn−1 ()
+ dx1 dx2 · · · dxn−1 f (x1 , x2 , . . . , xn , ).
d c1 c2 cn−1

Problem 20. Show that

Z ∞ Z ∞
tdt 1 τ dτ 1 π2 1
2πt − 1
= 2 τ −1
= 2 6
= .
0 e 4π 0 e 4π 24
 Integration 149

Problem 21. Let k ∈ N. Show that

Z 2π
(k − 1)!!
cosk (θ)dθ = 2(1 + (−1)k )π .
0 k!!

Problem 22. Let k ∈ N. Show that

Z 2π
(k − 1)!!
cosk (θ)dθ = 2(1 + (−1)k )π .
0 k!!

Problem 23. Show that

Z 1 √ 8
q
1− xdx = .
0 15

Hint. Set x = sin4 (y). Then dx = 4 sin3 (y) cos(y)dy.

Problem 24. Let a > 0. Show that

x2 dx
Z
x
=√ − arcsin(x/a) + C
(a2 − x2 )3/2 a − x2
2

Problem 25. Show that

Z ∞
sin(2px) sin(qx) |2p + q|
2 dx = ln .
0 x |2p − q|

Problem 26. For a λ/2 antenna we obtain the expression

Z λ/4
ωI0 sin θ
E(r, θ) = − cos(k`) sin(ω(t − c−1 (r − ` cos θ)))d`.
4π0 c2 r −λ/4

Calculate E(r, θ).

Problem 27. Let <(a) > 0 and n = 0, 1, . . .. Show that

Z ∞ Z ∞
2 ∂n 2
xn e−ax +px dx = n e−ax +px dx
−∞ ∂p −∞
∂n
r
π p2 /(4a)
= n e .
∂p a
150 Problems and Solutions

Problem 28. Let a 6= 0. Show that

Z ∞
sin(ax)
dx = πsgn(a).
−∞ x

Problem 29. Let c > 1. Show that

Z c Z 2
1 c sin(τ )
sin(x2 )dx = √ dτ.
1 2 1 τ

Problem 30. Show that

Z 1
1
x(1 + x2 )1/2 dx = (23/2 − 1)
0 3

setting u = 1 + x2 and hence du = 2xdx.

Problem 31. Let (Bτ , τ ≥ 0) be the linear Brownian motion starting

from 0. Show that Z 1
Γ+ := ds 1(Bs >0)
0
follows the arcsine distribution, i.e.
1 dτ
P (Γ+ ∈ dτ ) = p .
π τ (1 − τ )

Problem 32. Let a > 0. Show that

sin2 (ax/2)
Z
1
2
dx = πa.
R x 2

Problem 33. Let k = 1, 2, . . .. Study the integrals

Z πZ πZ π
1
Lk (bcc) = 3 (cos(x1 ) cos(x2 ) cos(x3 ))k dx1 dx2 dx3
π 0 0 0
Z πZ πZ π
1 1
Lk (sc) = 3 k (cos(x1 ) + cos(x2 ) + cos(x3 ))k dx1 dx2 dx3
π 3 0 0 0
Z πZ πZ π
1 1
Lk (f cc) = 3 k (cos(x1 ) cos(x2 )+cos(x2 ) cos(x3 )+cos(x3 ) cos(x1 ))dx1 dx2 dx3 .
π 3 0 0 0
 Integration 151

They play a role in solid state physics for the body-centred cubic lattice,
simple cubic lattice, face-center cubic lattice.

Problem 34. Draw the functions

f1 (x) = cos(2πx)
f2 (x) = cos(2π(cos(2πx)))
f3 (x) = cos(2π(cos(2π(cos(2πx))))).

Extend it to fn (x). Find the integral

Z 2π
fn (x)dx.
0

Problem 35. Show that (Fresnel’s integral)

Z ∞ Z ∞ √
2 2 π
cos(x )dx = sin(x )dx = √ .
0 0 2 2

Problem 36. Show that

1 ∞
Z
1
sech(α − β)eiωβ dβ = sech( πω)eiωα .
π −∞ 2

Problem 37. Let σ > 0. Show that

∞  1/4
x2
Z  
1 5/2 32
x exp − 2 dx = Γ(3/4)σ 3/2 .
4σ 2 0 8σ 10

Problem 38. Let α > 0. Show that

Z ∞
2 1 −β 2 /(4α)
xe−αx J0 (βx)dx = e .
0 2α

Problem 39. Show that

Z ∞
dx π
3 = .
0 cosh (x) 4
Chapter 9

Functional Equations

A functional equation is any equation that specifies a function in implicit

form. An example is the Gamma function Γ(z)

Γ(z + 1) = zΓ(z)

with Γ(1) = Γ(2) = 1. The trigonometric function sin(x) and cos(x) satisfy
a system of functional equations. So functional equations are equations
where the unkonws are functions. In most cases it is assumed that the
functions are continouos.

9.1 Solved Problems

Problem 1. To solve a number of nonlinear functional equation it is
helpful to have the solution of the linear functional equation

f (x + y) = f (x) + f (y) (1)

where we assume that the function f is continouos.

Problem 2. Find the solution of the functional equations

f (x + y) = f (x)f (y) (1)

where we assume that f is continuous.

152
 Functional Equations 153

Problem 3. Find the solution of the functional equation

f (xy) = f (x) + f (y) (1)
where we assume that the function f is continuous.

Problem 4. Find the solution of the functional equation

f (xy) = f (x)f (y). (1)
We assume that the function f is continuous.

Problem 5. Find the solution of the Jensen equation

 
x+y f (x) + f (y)
f = (1)
2 2
where we assume that f is continuous.

Problem 6. Show that the functional equation

f (x) + f (y)
f (x + y) =
1 − f (x)f (y)
admits the solutions f (x) = tan(cx), where c is a constant.

Problem 7. Show that the functional equation

f (x) + f (y)
f (x + y) = f (x)f (y)
(1)
1+ C2

admits the solution

f (x) = C tanh(cx). (2)

Problem 8. Show that the functional equation

f (x)f (y)
f (x + y) =
f (x) + f (y)
admits the solution
c
f (x) = .
x

Problem 9. Show that the functional equation

f (x) + f (y) − 1
f (x + y) =
2f (x) + 2f (y) − 2f (x)f (y) − 1
154 Problems and Solutions

admits the solution

1
f (x) = .
1 + tan(cx)

Problem 10. Show that the functional equation

p p
f (x + y) = f (x)f (y) + f (x)2 f (y)2 − 1
admits the solution
f (x) = cosh(cx).

Problem 11. Show that the functional equation

f (x + y + axy) = f (x)f (y)
admits the solution
f (x) = (1 + ax)c .

Problem 12. Show that the functional equation

p
f (x + y) − f (x − y) = 4 f (x)f (y)
admits the solution
f (x) = cx2 .

Problem 13. Solve the functional equation

f (x + y) + f (x − y) = 2f (x)f (y)
assuming that f is a continuous function.

Problem 14. Show that the trigonometric functions f (x) = cos(x) and
g(x) = sin(x) satisfy the system of functional equations
g(x + y) = g(x)f (y) + f (x)g(y)
f (x + y) = f (x)f (y) − g(x)g(y)
g(x − y) = g(x)f (y) − g(y)f (x)
f (x − y) = f (x)f (y) + g(x)g(y).

Problem 15. Show that the Jacobi elliptic functions satisfy the system
of functional equations
f (x)g(y)h(y) ± f (y)g(x)h(x)
f (x ± y) =
1 − k 2 f (x)2 f (y)2
 Functional Equations 155

g(x)g(y) ∓ f (x)f (y)h(x)h(y)

g(x ± y) =
1 − k 2 f (x)2 f (y)2
h(x)h(y) ∓ k 2 f (x)f (y)g(x)g(y)
h(x ± y) = .
1 − k 2 f (x)2 f (y)2

Problem 16. Let a be a positive integer with a ≥ 2. Let 1 ≤ x ≤ a.

Consider the equation

g(x − 1) − 2g(x) + g(x + 1) = −λg(x).

Show that

gj (x) = sin(jπx/(a + 1)), λj (x) = 2(1 − cos(jπ/(a + 1)))

satisfy this equation.

Problem 17. Let a, c,  be positive constants. Solve the functional equa-

tion
g((θ + c)(mod 1)) = ag(θ) +  sin(2πθ).
156 Problems and Solutions

9.2 Supplementary Problems

Problem 1. Show that the functional equation

f (x) + f (y) + 2f (x)f (y)

f (x + y) =
1 − f (x)f (y)

admits the solution

cx
f (x) = .
1 − cx

Problem 2. Consider the analytic function f : R → R

f (x) = arctan(x).

Show that  
x+y
f (x) + f (y) = f .
1 − xy
What happens at xy = 1?

Problem 3. Consider the functional equation

g(x) = αg(g(x/α))

with g(0) = 0 and g 0 (x = 0) = 1. Show that

x
g(x) =
1 − cx
is a solution with c an arbitarary constant.

Problem 4. We know that

tan(α) + tan(β)
tan(α + β) ≡ .
1 − tan(α) tan(β)

Show that
tan(α + β + γ) =
tan(α) + tan(β) + tan(γ) − tan(α) tan(β) tan(γ)
.
1 − tan(α) tan(β) − tan(α) tan(γ) − tan(β) tan(γ)
Chapter 10

Inequalities

10.1 Solved Problems

Problem 1. Let a, b ∈ R. Show that
2ab ≤ a2 + b2 .

Problem 2. Let a, b ∈ R. Show that

|a + b| ≤ |a| + |b|.

Problem 3. Let a, b ∈ R+ . Show that

1 √
(a + b) ≥ ab.
2

Problem 4. Let x ≥ 0 and 0 < p < 1. Show that

1
(1 − xp ) ≥ 1 − x.
p

Problem 5. Let x ∈ (0, 1). Show that

x(1 − x) < x.

157
158 Problems and Solutions

Problem 6. Let n ∈ N.
(i) Show that n < 2n .
(ii) Show that n2 < 4n .
(iii) Show that if n ≥ 4 then 2n < n!.

Problem 7. Let x, y be two nonnegative real numbers. Show that

 2
x+y 1
xy ≤ ≡ (x2 + y 2 + 2xy).
2 4

Problem 8. Let a, b, c, d be nonnegative real numbers. Show that

1
(abcd)1/4 ≤ (a + b + c + d).
4

Problem 9. Let xj (j = 1, . . . , n) be nonegative real numbers. Show

that  n
x1 + x2 + · · · + xn
x1 x2 · · · xn ≤ .
n

Problem 10. Let a, b, c, d be nonnegative real numbers. Show that

a4 + b4 + c4 + d4 ≥ 4abcd.

Problem 11. Let x ∈ R and n ∈ N. Show that (Bernoulli inequality)

(1 + x)n ≥ 1 + nx.

Problem 12. Let x ≥ 0. Show that

x
1 − e−x ≥ .
1+x

Problem 13. Let

1 1 1
en := 1 + + + ··· +
1! 2! n!
where n = 1, 2, . . .. Let m > n. Show that
2
|em − en | = em − en ≤ .
(n + 1)!
 Inequalities 159

Problem 14. Does the inequality

1 + 2 cos(θ) − cos(2θ) ≤ 2

hold for all θ ∈ [0, 2π)?

Problem 15. Let x > 0. Show that

x x2 √ x
1+ − ≤ 1+x≤1+ .
2 8 2

Problem 16. Let x > 0. Show that

1
1− ≤ ln(x) ≤ x − 1
x
with equality iff x = 1.

Problem 17. Show that if a twice differentiable function f : R → R has

a second order derivative which is non-negative (positive) everywhere, then
the function is convex (strictly convex).

Problem 18. Let a1 , a2 , . . . , an be positive numbers and b1 , b2 , . . . , bn be

nonnegative numbers such that
n
X
bj > 0.
j=1

Show that (log-sum inequality)

  P 
n  n n
X aj
 X j=1 aj
aj log ≥ aj  log P 
bj n
j=1 j=1 j=1 bj

with the conventions based on continuity arguments

p
0 · log 0 = 0, 0 · log = ∞, p > 0.
0
Show that equality holds if and only if aj /bj = constant for all j =
1, 2, . . . , n.

Problem 19. Let xj (j = 1, 2, . . . , n) be positive real numbers. Show

that Pn
xk
(x1 x2 · · · xn )1/n ≤ k=1 .
n
160 Problems and Solutions

This is the arithmetic-geometric mean inequality.

Problem 20. Let n be a positive integer and a, b ≥ c/2 > 0. Show that

|a−n − b−n | ≤ 4nc−n−1 |a − b|.

Problem 21. Let X ⊂ R be an interval. A function ψ : X → R is

convex if for all x1 , x2 ∈ X and numbers α1 , α2 ≥ 0 with α1 + α2 = 1,

ψ(α1 x1 + α2 x2 ) ≤ α1 ψ(x1 ) + α2 ψ(x2 ). (1)

This means that every chord of the graph of ψ lies above the graph. Let
ψ:X→ PR be convex, let x1 , x2 , . . . , xn ∈ X, and let α1 , α2 , . . . , αn ≥ 0
n
satisfy j=1 αj = 1. Show that (Jensen’s inequality)
n
! n
X X
ψ αi xi ≤ αi ψ(xi ). (2)
i=1 i=1

Problem 22. Consider the differentiable function f : [0, ∞) → R

x
f (x) = .
1+x
Let a, b ∈ R+ . Show that

f (|a + b|) ≤ f (|a| + |b|). (1)

Problem 23. Let a, b ∈ R. Show that

|a + b| |a| |b|
≤ + .
1 + |a + b| 1 + |a| 1 + |b|

Problem 24. Show that f : R → R, f (x) = |x| is convex.

Problem 25. Let n ∈ N and n ≥ 2. Show that

en ln(n + 1) < en+1 ln(n).

Problem 26. Show that the function f : (0, ∞) → R

f (x) = x ln(x)
 Inequalities 161

is convex.

Problem 27. Let x, y ∈ R. Show that

x+y |x| |y|
≤ + .
1 + |x + y| 1 + |x| 1 + |y|

Problem 28. Let A, B be n × n matrices over C. Show that

kABk ≤ kAk · kBk.

Problem 29. Let A, B be n × n matrices over C. Show that

kA + Bk ≤ kAk + kBk.

Problem 30. Let H be a Hilbert space. Let ψ, φ ∈ H. Assume that

ψ 6= 0, φ 6= 0. Show that
p p
|hφ, ψi| ≤ hφ, φi hψ, ψi (1)
where h , i denotes the scalar product in the Hilbert space.

Problem 31. Let n ≥ 2. Let x1 , x2 , . . . , xn be given positive real number

with
x1 < x2 < · · · < xn .
Pn
Let λ1 , . . . , λn ≥ 0 and j=1 λj = 1. Show that
  
n
X n
X
 λ j xj   λj x−1j
 ≤ A2 G−2
j=1 j=1

where
1
A= (x1 + xn ), G = (x1 xn )1/2 .
2

Problem 32. Let x, y ∈ Rn (column vectors) and  > 0. Show that

1
2xT y ≤ xT x + yT y.


Problem 33. Let a, b ∈ R and  > 0. Show that

2ab ≤ a2 + −1 b2 .
162 Problems and Solutions

Problem 34. Show that the analytic function f : R2 → R

f (α, β) = sin(α) sin(β) cos(α − β)

is bounded between −1/8 and 1.

Problem 35. Let n be an integer and n ≥ 2. Show that

en ln(n + 1) < en+1 ln(n).

Problem 36. Let a, b, c be positive numbers. Assume that

x2 y2 z2
+ + = 1.
a2 b2 c2
Show that p
x+y+z ≤ a2 + b2 + c2 .
Use the Lagrange multiplier method.

Problem 37. Let a, b, c, d be positive numbers and s := a + b + c + d.

Show that
s s s s 16
+ + + ≥ .
s−a s−b s−c s−d 3

Problem 38. Let x ≥ 0 and 0 < p < 1. Show that

1
(1 − xp ) ≥ 1 − x.
p

Problem 39. Let b be a real number such that the elements of the infinite
sequence (ak )∞
k=1 satisfy

ak+m ≤ ak + am + b

for all k, m = 1, 2, . . .. Show that

ak
a := lim
k→∞ k
exists and ak ≥ ka − b for all k.

Problem 40. Let n ∈ N and n ≥ 2. Show that

1 1 1 √
√ + √ + · · · + √ > n.
1 2 n
 Inequalities 163

Problem 41. Let a, b ≥ 0. Show that

a+b a + b + ab a b
≤ ≤ + .
1+a+b 1 + a + b + ab 1+a 1+b

Problem 42. Show that

Z π/2 Z π/2
1 sin2 (x − y) 1
2 2
dxdy < .
π 0 0 (x − y) 4

Problem 43. Let x1 , x2 , x3 ∈ R. Show that

x21 + x22 + x23 ≥ x1 x2 + x2 x3 + x3 x1 .

Why could it be useful to consider the function f : R3 → R

f (x1 , x2 , x3 ) = x21 + x22 + x23 − x1 x2 − x2 x3 − x3 x1 ?

Problem 44. Let x1 , x2 , . . . , xn be positive numbers. Assume that the

numbers satisfy
Xn Xn
xj = 1, x2j = b2 .
j=1 j=1

Show that
1 √ p
max{ xj : 1 ≤ j ≤ n} ≤ (1 + n − 1 nb2 − 1).
n

Problem 45. Consider the function f : R → R, f (x) = x2 . Show that

|f (x2 ) − f (x1 )| ≤ 4|x2 − x1 |

for all x1 and x2 in [−2, 2]. Hint. Apply

x21 − x22 ≡ (x1 + x2 )(x1 − x2 ).

Problem 46. Let n ∈ N. Show that

n 1 1 1
< 1 + + + ··· + n ≤ n.
2 2 3 2 −1
164 Problems and Solutions

Problem 47. Let

0 ≤ a1 ≤ a2 ≤ · · · ≤ an , 0 ≤ b1 ≤ b2 ≤ · · · ≤ bn

and r ≥ 1. Show that (Chebyshev inequality)

 1/r  1/r  1/r
n n n
1 X 1 X 1 X
 ar   br  ≤ (aj bj )r  .
n j=1 j n j=1 j n j=1

Problem 48. Let a, x, y ∈ R. Show that

p p
| a2 + x2 − a2 + y 2 | ≤ |x − y|.

Problem 49. Let b > a > 0. Show that

 
 a b b
1− < ln < − 1.
b a a

Problem 50. Show that

Z 1 
 cos(nx) 
 ≤ ln(2)

0 x+1 

for n ∈ N.

Problem 51. Let n ≥ 2. Show that

 
2n
2n < < 22n .
n

Problem 52. Consider the two manifolds

x21 + x22 = 1, y12 + y22 = 1.

Show that
|x1 y1 + x2 y2 | ≤ 1.
Hint. Set

x1 (t) = cos(t), x2 (t) = sin(t), y1 (t) = cos(τ ), y2 (t) = sin(τ ).

 Inequalities 165

Problem 53. Let n ∈ N and x > 0. Show that

1 1
xn+1 + > xn + n .
xn+1 x

Problem 54. Find all x ∈ R and x > 0 such that

√ √
| x − 2| < 1.

Problem 55. Let j, k be positive intgers. Find all pairs (j, k) such that
the following four inequalities are satisfied
j j
j + k < 10, j + k ≥ 6, > 1, < 2.
k k

Problem 56. Let n be a positive integer. Show that

1 1 1 √
√ + √ + · · · + √ ≥ n.
1 2 n

Problem 57. Let z = x + iy with x, y ∈ R. Show that

1
√ (|x| + |y|) ≤ |z| ≤ |x| + |y|.
2
166 Problems and Solutions

Matrices

Problem 58. Let A, B be n × n matrices over C. Show that

|tr(AB ∗ )|2 ≤ tr(AA∗ )tr(BB ∗ ).

Problem 59. Let A and B be two n × n matrices over R. It can be

shown that
1
treA+B ≤ tr(eA eB ) ≤ tr(e2A + e2B )
2
treA+B ≤ tr(eA eB ) ≤ (trepA )1/p (treqB )1/q

where p > 1, q > 1 with

1 1
+ = 1.
p q
Is
1
(trepA )1/p (treqB )1/q ≤ tr(e2A + e2B ) ?
2
Prove or disprove.

Problem 60. Let A, B be hermitian matrices. Then

tr(eA+B ) ≤ tr(eA eB ).

Assume that    
0 0 0 1
A= , B= .
0 1 1 0
Calculate the left and right-hand side of the inequality. Does equality hold?

Problem 61. Let A, B be positive definite matrices. Show that

p+1 p
tr(AB)2 ≤ tr(A2 B 2 )2 , p a non-negative integer

Problem 62. Let A be an n × n matrix with kAk < 1.

(i) Show that (In + A)−1 exists.
(ii) Show that
1
k(In + A)−1 k ≤ .
1 − kAk
 Inequalities 167

Problem 63. Let A be an n × n matrix over R. Assume that ajj ≥ 1 for

all j and
n
X
a2jk < 1.
j6=k

Show that A is inververtible.

Problem 64. Let A be an n × n matrix over C. Let In be the n × n

identity matrix. Assume that
n
X
|ajk | < 1
k=1

for each j. Show that In − A is invertible.

Problem 65. Let A be an n × n positive definite matrix over R. Let B

be an n × n positive semidefinite matrix over R. Show that
det(A + B) ≥ det(A).

Problem 66. Let A be an n × n matrix over R. Show that there exists

nonnull vectors x1 , x2 in Rn such that
xT1 Ax1 xT Ax xT Ax2
T
≤ T ≤ 2T
x1 x1 x x x2 x2
for every nonnull vector x in Rn .

Problem 67. Let A be an n × n skew-symmetric matrix over R. Show

that
det(In + A) ≥ 1
with equality holding if and only if A = 0.

Problem 68. Let A be an n × n matrix over R. Assume that ajj ≥ 1 for

j = 1, 2, . . . , n and
Xn
a2jk < 1.
j6=k

Show that A is invertible.

Problem 69. Let A be an n × n matrix over C. Assume that

n
X
|ajj | > |ajk |
k6=j
168 Problems and Solutions

for all j = 1, 2, . . . , n. Show that A is invertible.

Problem 70. Let A, B be n × n positive definite matries. Show that

tr(A ln(A)) − tr(A ln(B)) ≥ tr(A − B).

Problem 71. Let v be a normalized (column) vector in Cn and let A be

an n × n hermitian matrix. Is
∗
v ∗ eA v ≥ ev Av

for all normalized v? Prove or disprove.

 Inequalities 169

10.2 Supplementary Problems

Problem 1. Let n ∈ N and n ≥ 2. Show that
1 √ √ 1
√ >2 n−2 n−1> √ .
n−1 n
Use this reult to show that
1 1
1 + √ + √ + ···
2 3
diverges.

Problem 2. Let b > a. Show that

b−a b−a
2
< arctan(b) − arctan(a) < .
1+b 1 + a2
Note that arctan(−x) = − arctan(x).

Problem 3. Let n be is positive integer. Is

p
n+1
√
n
(n + 1)! > n!
for all n?

Problem 4. Let a, b, c ∈ C with |a| = |b| = |c| = 1. Suppose that

=(a) ≥ 0, =(b) ≥ 0, =(c) ≤ 0. Show that
 
 −1 + ab + bc + ca 
  ≤ √1 .
 4  2

Problem 5. Let n be a positive integer with n ≥ 2. Show that

1 1 1 1
+ + + · · · + n−1 < 1.
2 4 8 2

Problem 6. Let n ≥ 2 and x1 , x2 , . . . , xn be positive numbers. Assume

that the numbers satisfy
n
X n
X
xj = 1, x2j = b2 .
j=1 j=1

Show that
1 √ p 
max{ xj : 1 ≤ j ≤ n } ≤ 1 + n − 1 nb2 − 1 .
n
170 Problems and Solutions

Problem 7. Show that

1 1 1 1
1+ + + ··· + 2 ≤ 2 − .
4 9 n n

Problem 8. Let x ≥ 0. Show that 1 + x ≤ 3x .

Problem 9. Let A be an n × n matrix. Is

| det(In + A)| ≤ exp(kAk)?

Does it depend on the chosen norm?

Problem 10. Let zj (j = 1, . . . , p) be fixed complex numbers. Is

p
Y p
Y
|z − zj | > (|zj | − |z|) ?
j=1 j=1

Problem 11. Let x ∈ (0, 1). Show that

x2
x > ln(1 + x) > x − .
2

Problem 12. Let x, y ∈ R. Show that

|x + y| |x| |y|
≤ + .
1 + |x + y| 1 + |x| 1 + |y|

Problem 13. Let x1 > 0, x2 > 0, α1 > 0, α2 > 0 and α1 + α2 = 1. Show

that
xα1 α2
1 x2 ≤ α1 x1 + α2 x2 .

Apply ln(x).
Inequalities 171
Bibliography

Abramowitz M. and Stegun (editors)

Pocketbook of Mathematical Functions
Harry Deutsch (1984)

Aldous J. M. and Wilson R. J.

Graphs and Applications: An Introductory Approach
Springer (2000)

Armstrong M. A.
Groups and Symmetry
Springer (1988)

Barbeau E. J.
Polynomials, Springer-Verlag (1989)

Bender C. M. and Orszag S. A. Mathematical Methods for Scientists and

Engineers, McGraw-Hill, New Yrok 1978

Bredon G. E.
Introduction to Compact Transformation Groups
Academic Press (1972)

Bronson R.
Matrix Operations
Schaum’s Outlines, McGraw-Hill (1989)

Chern S. S., Chen W. H. and Lam K. S.

Lectures on Differential Geometry
World Scientific (1999)

DasGupta Ananda
American J. Phys. 64 1422–1427 (1996)

172
Bibliography 173

de Souza P. N. and Silva J.-N.

Berkeley Problems in Mathematics
Springer (1998)

Englefield M. J.
Group Theory and the Coulomb Problem
Wiley-Interscience, New York (1972)

Erdmann K. and Wildon M.

Introduction to Lie Algebras,
Springer (2006)

Fuhrmann, P. A.
A Polynomial Approach to Linear Algebra
Springer (1996)

Golub G. H. and Van Loan C. F.

Matrix Computations, Third Edition,
Johns Hopkins University Press (1996)

Gradshteyn I. S. and Ryzhik I. M.

Tables of Integrals, Series and Products
Academic Press, New York (1980)

Hall B. C.
Lie Groups, Lie Algebras, and Representations: an elementary introduction
Springer (2003)

Hardy G. H., Littlewood J. E. and Pólya G.

Inequalities, 2nd edition, Cambridge University Press, Cambridge, 1952

Helgason S.
Groups and Geometric Analysis, Integral Geometry, Invariant Differential
Operators and Spherical Functions
Academic Press (1984)

Horn R. A. and Johnson C. R.

Topics in Matrix Analysis
Cambridge University Press (1999)

Kedlaya K. S., Poonen B. and Vakil R.

The William Lowell Putnam Mathematical Competition 1985–2000,
174 Bibliography

The Mathematical Association of America (2002)

Keener J. P.
Principles of Applied Mathematics: Transformation and Approximation,
Perseus Books, second revised edition (2000)

Lang S.
Linear Algebra
Addison-Wesley, Reading (1968)

Lay S. R.
Analysis: with an introduction to proof
Pearson Prentice Hall (2005)

Letac G.
Integration and Probability by Paul Malliavin: Exercises and Solution
Springer (1995)

Malik S. C. and Arora S.

Mathematical Analysis, 2nd edition
New Age International (1992)

Miller W.
Symmetry Groups and Their Applications
Academic Press, New York (1972)

Posamentier A. S. and Salkind C. T.

Challenging Problems in Algebra, Dover Publications, New York 1996

Riley K. F. and Hobson M. P.

Essential Mathematical Methods for the Physical Sciences: Student Solu-
tion Manual
Cambridge Unversity Press, 2011

Rosenlicht M.
Introduction to Analysis
Dover (1968)

Sneddon I.
Elements of Partial Differential Equations, McGraw-Hill (1957)

Spiegel M. R.
Advanced Calculus, Schaum’s Outline Series, McGraw-Hill (1962)
Bibliography 175

Spiegel M. R.
Complex Variables, Schaum’s Outline Series, McGraw-Hill (1964)

Steeb W.-H.
Matrix Calculus and Kronecker Product with Applications and C++ Pro-
grams
World Scientific Publishing, Singapore (1997)

Steeb W.-H.
Continuous Symmetries, Lie Algebras, Differential Equations and Com-
puter Algebra
World Scientific Publishing, Singapore (1996)

Steeb W.-H.
Hilbert Spaces, Wavelets, Generalized Functions and Quantum Mechanics
Kluwer Academic Publishers, Dordrecht (1998)

Steeb W.-H.
Problems and Solutions in Theoretical and Mathematical Physics,
Second Edition, Volume I: Introductory Level
World Scientific Publishing, Singapore (2003)

Steeb W.-H.
Problems and Solutions in Theoretical and Mathematical Physics,
Second Edition, Volume II: Advanced Level
World Scientific Publishing, Singapore (2003)

Steeb W.-H., Hardy Y., Hardy A. and Stoop R.

Problems and Solutions in Scientific Computing with C++ and Java Sim-
ulations
World Scientific Publishing, Singapore (2004)

Steele J. M.
The Cauchy-Schwarz Master Class, Cambridge University Press, 2000

Stewart J.
Calculus
Brooks/Cole (2012)

Watkins D. S.
Fundamentals of matrix computations
Wiley
176 Bibliography

Whittaker E. T.
A treatise on the analytical dynamics of particles and rigid bodies

Whittaker E. T. and Watson G. N. A Course of Modern Analysis 4th edi-

tion, Cambridge University Press (1962)

Wybourne B. G.
Classical Groups for Physicists
John Wiley, New York (1974)
Index

z-transform, 110 Convex set, 56

convolution integral, 119
Airy functions, 49 Coulomb potential, 47
algebraic, 88 Critcal points, 74
Alternating function, 47 Cubic B-spline, 119
Angular distance, 51 Cubic equation, 94
Arc length, 42
Associated Legendre polynomials, 76 Dawson’s integral, 130
Diffeomorphism, 28
Bargmann space, 85 Diffusion equation, 122
Bell polynomials, 86 Dirichlet integral identity, 123
Bernoulli inequality, 158 Dirichlet kernel, 23
Bernoulli numbers, 1, 85 Dominant tidal potential, 75
Bernoulli polynomials, 84
Bernstein basis polynomials, 77 Elementary symmetric functions, 75
Bernstein polynomials, 77 Euler dilogarithm function, 50
Bessel function, 114 Euler summation formula, 1
Beta function, 136
Bethe ansatz, 3 Fejér kernel, 8
Braid-like relation, 76 Fermat number, 21
Brillouin function, 87 Fibonacci numbers, 15, 37
Buchberger algorithm, 89 Fixed points, 109
fixed points, 74
Cantor function, 4 Fresnel’s integral, 151
Cantor series, 15
Cantor series approximation, 5 Gamma function, 22, 53, 68
Cauchy integral theorem, 110 Gauss sum, 12
Cauchy principal value, 132
Cauchy sequence, 103 Haar wavelet, 122
Cesáro sum, 4 Hankel matrix, 82, 83
Chebyshev inequality, 164 Harmonic series, 1
Chebyshev polynomials, 73 Hermite polynomial, 74
Chordal distance, 105 Hermite polynomials, 73
Convex, 160 Hilbert transform, 132–134
Convex function, 56 Hurwitz zeta-function, 68

177
178 Index

Invariant, 77 Symmetric group, 80

Jensen equation, 153 Toroidal coordinates, 47

Trace, 78
Kernel, 44
Kustaanheimo-Stiefel transformation, Vector product, 35
39
Wronksi determinant, 45
Laguerre polynomials, 73
Legendre map, 36
Legendre polynomials, 75
Lemniscate, 141
log-sum inequality, 159
Logarithmic derivative, 42
Lorenz model, 79
Lotka Volterra model, 80

Newton’s method, 27, 29

Padé approximant, 64
Partition, 30
Poisson summation formula, 16
Poisson wavelet, 122
Polar coordinates, 141
Polar form, 110
Principal argument, 109
Principal value, 133, 134
Prolate spheroidal coordinates, 47
Pyramide, 10

Quartic equation, 87
Quintic equation, 84

Radon transform, 133

Recursion relation, 75
recursion relation, 89
Residue theorem, 111
Resultant, 82
Rudin-Shapiro substitution, 36

Schwarzian derivative, 60
Sonine polynomials, 87
Spherical cosine theorem, 75
Star product, 32
Sum rule, 74