Analysis

Course 221
David Simms
School of Mathematics
Trinity College
Dublin

Typesetting
Eoin Curran.

1
2
Contents

1 Lebesgue Measure 7
1.1 Algebra of Subsets . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2 The Interval Algebra . . . . . . . . . . . . . . . . . . . . . . . 9
1.3 The length measure . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 The σ-algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.5 The outer measure . . . . . . . . . . . . . . . . . . . . . . . . 14
1.6 Extension of measure to σ-algebra, using outer measure . . . . 16
1.7 Increasing Unions, Decreasing Intersections . . . . . . . . . . . 20
1.8 Properties of Lebesgue Measure . . . . . . . . . . . . . . . . . 22
1.9 Borel Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2 Integration 29
2.1 Measure Space, Measurable sets . . . . . . . . . . . . . . . . . 29
2.2 Characteristic Function . . . . . . . . . . . . . . . . . . . . . . 30
2.3 The Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.4 Monotone Convergence Theorem . . . . . . . . . . . . . . . . 34
2.5 Existence of Monotone Increasing Simple Functions converg-
ing to f . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.6 ’Almost Everywhere’ . . . . . . . . . . . . . . . . . . . . . . . 39
2.7 Integral Notation . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.8 Fundamental Theorem of Calculus . . . . . . . . . . . . . . . 46
2.9 Fatou’s Lemma . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.10 Dominated Convergence Theorem . . . . . . . . . . . . . . . . 49
2.11 Differentiation under the integral sign . . . . . . . . . . . . . . 50

3 Multiple Integration 53
3.1 Product Measure . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.2 Monotone Class . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.3 Ring of Subsets . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.4 Integration using Product Measure . . . . . . . . . . . . . . . 59
3.5 Tonelli’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 63

3
 3.6 Fubini’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 64

4 Differentiation 71
4.1 Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.2 Normed Space . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.3 Metric Space . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.4 Topological space . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.5 Continuous map of topological spaces . . . . . . . . . . . . . . 75
4.6 Homeomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.7 Operator Norm . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.8 Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.9 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.10 C r Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.11 Chain Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

5 Calculus of Complex Numbers 89

5.1 Complex Differentiation . . . . . . . . . . . . . . . . . . . . . 89
5.2 Path Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.3 Cauchy’s Theorem for a triangle . . . . . . . . . . . . . . . . . 95
5.4 Winding Number . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.5 Cauchy’s Integral Formula . . . . . . . . . . . . . . . . . . . . 101
5.6 Term-by-term differentiation, analytic functions, Taylor series 104

6 Further Calculus 107

6.1 Mean Value Theorem for Vector-valued functions . . . . . . . 107
6.2 Contracting Map . . . . . . . . . . . . . . . . . . . . . . . . . 108
6.3 Inverse Function Theorem . . . . . . . . . . . . . . . . . . . . 109

7 Coordinate systems and Manifolds 115

7.1 Coordinate systems . . . . . . . . . . . . . . . . . . . . . . . . 115
7.2 C r -manifold . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7.3 Tangent vectors and differentials . . . . . . . . . . . . . . . . . 118
7.4 Tensor Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.5 Pull-back, Push-forward . . . . . . . . . . . . . . . . . . . . . 125
7.6 Implicit funciton theorem . . . . . . . . . . . . . . . . . . . . 127
7.7 Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
7.8 Lagrange Multipliers . . . . . . . . . . . . . . . . . . . . . . . 132
7.9 Tangent space and normal space . . . . . . . . . . . . . . . . . 133
7.10 ?? Missing Page . . . . . . . . . . . . . . . . . . . . . . . . . . 134
7.11 Integral of Pull-back . . . . . . . . . . . . . . . . . . . . . . . 135
7.12 integral of differential forms . . . . . . . . . . . . . . . . . . . 136

4
 7.13 orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

8 Complex Analysis 145

8.1 Laurent Expansion . . . . . . . . . . . . . . . . . . . . . . . . 145
8.2 Residue Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 147
8.3 Uniqueness of analytic continuation . . . . . . . . . . . . . . . 151

9 General Change of Variable in a multiple integral 153

9.1 Preliminary result . . . . . . . . . . . . . . . . . . . . . . . . . 153
9.2 General change of variable in a multiple integral . . . . . . . . 156

5
6
Chapter 1

Lebesgue Measure

1.1 Algebra of Subsets

Definition Let X be a set. A collection A of subsets of X is called an
algebra of subsets of X if

1. ∅ ∈ A

2. E ∈ A =⇒ E 0 ∈ A. E 0 = {x ∈ X : x 6∈ E}

3. E1 , . . . , Ek ∈ A =⇒ E1 ∪ E2 ∪ · · · ∪ Ek ∈ A. i.e., A is closed under

complements and under finite unions. Hence A is closed under finite
intersection. (because E1 ∩ · · · ∩ Ek = (E1 0 ∪ · · · ∪ Ek 0 )0 )

Example Let I be the collection of all finite unions of intervals in R of the

form:

1. (a, b] = {x ∈ R : a < x ≤ b}

2. (−∞, b] = {x ∈ R : x ≤ b}

3. (a, ∞) = {x ∈ R : a < x}

4. (−∞, ∞) = R

I is an algebra of subsets of R, called the Interval Algebra.

  
  

7
 We want to assign a ‘length’ to each element of the interval algebra I, so
we want to allow ‘∞’ as a length. We adjoin to the real no’s the 2 symbols
∞ and −∞ to get the Extended Real Line:

Definition The Extended Real Line:

R ∪ {−∞, ∞} = [−∞, ∞]

We extend ordering, addition, multiplication to [−∞, ∞] by:

1. −∞ < x < ∞ ∀x ∈ R

2. Addition

(a) ∞ + ∞ = x + ∞ = ∞ + x = ∞
(b) (−∞) + (−∞) = x + (−∞) = (−∞) + x = −∞

(Don’t define ∞ + (−∞) or (−∞) + ∞.)

3. Multiplication

(a) ∞∞ = ∞ = (−∞)(−∞)
(b) ∞(−∞) = (−∞)∞ = −∞
 
 ∞ x>0 
(c) x∞ = ∞x = 0 x=0
 
−∞ x < 0
 
 −∞ x>0 
(d) x(−∞) = (−∞)x = 0 x=0
 
∞ x<0

Definition If A ⊂ [−∞, ∞] then

1. supA = ∞ if ∞ ∈ A or if A has no upper bound in R.

2. inf A = −∞ if −∞ ∈ A or if A has no lower bound in R.

Definition Let A be an algebra of subsets of X. A function m : A −→ [0, ∞]

is called a measure on A if:

1. m(∅) = 0
S∞
2. if
P∞ E = j=1 Ej is countably disjoint with Ej ∈ A ∀j then m(E) =
j=1 m(Ej ). (m is Countably Additive.)

8
 E

Figure 1.1: E is a disjoint union of sets

Note:

1. Countable means that E1 , E2 , . . . is either a finite sequence or can be

labelled by (1, 2, 3, . . .).

2. Disjoint means that Ei ∩ Ej = ∅ ∀i 6= j

P
3. m(E) = ∞ j=1 m(Ej ) means either:
P∞
(a) j=1 m(Ej ) is a convergent series of finite no’s with the sum of
the series as m(E)
P∞
(b) j=1 m(Ej ) is a divergent series with m(E) = ∞

(c) m(Ej ) = ∞ for some j, and m(E) = ∞

1.2 The Interval Algebra

Definition For each E ∈ I, the interval algebra (See Section 1.1), write

E = E1 ∪ · · · ∪ Ek

(disjoint) where each Ei is of the form (a, b], (−∞, b], (a, ∞), or(−∞, ∞). The
length of E is m(E) = m(E1 ) + m(E2 ) + · · · + m(Ek ) where m(a, b] = b − a.

m(−∞, b] = m(a, ∞) = m(−∞, ∞) = ∞

Definition a set V ⊂ R is called an Open Subset of R if:

for each a ∈ V ∃ > 0 s.t.(a − , a + ) ⊂ V

 


 a− a a+

i.e., every point is an interior point: it has no end points.

9
Theorem 1.2.1. (Heine-Borel-Lebesgue) The closed interval is compact.
Let {Vi }i∈I be a family of open sets in R which cover the closed interval
[a, b]. Then ∃ a finite number of them: Vi1 , . . . , Vik (say) which cover [a, b]
(i.e. [a, b] ⊂ Vi1 ∪ · · · ∪ Vik .)

Proof. Put

K = {x ∈ [a, b] s.t. ∃( finite set {i1 , . . . , ir } ⊂ I s.t. [a, x] ⊂ Vi1 ∪ · · · ∪ Vir )}

a c b

a ∈ K =⇒ K 6= ∅

Let
c = sup K

Then
a ≤ c ≤ b c ∈ Vj (say)

Vj is open, therefore

∃  > 0 s.t. (c − , c + ) ⊂ Vj

and

∃ k ∈ K s.t. c −  < k ≤ c,

[a, k] ⊂ Vi1 ∪ · · · ∪ Vir (say)

Then

[a, min(c + , b)] ⊂ (Vi1 ∪ · · · ∪ Vir ∪ Vj )
2
therefore

min(c + , b) ∈ K
2

But c + 2
6∈ K (as c = sup K) =⇒ b ∈ K as required.

10
1.3 The length measure
Theorem 1.3.1. The length function

m : I −→ [0, ∞]

is a measure on the interval algebra I

Proof. We have to check that m is countably additive, which reduces to
showing that if
∞
[
(a, b] = (aj , bj ]
j=1

is a countable disjoint union then

∞
X
(bj − aj )
j=1

is a convergent series with sum b − a.

Take the first n intervals and relabel them:

(a1 , b1 ], . . . , (an , bn ]

so that:

a a1 b1 a2 b2 a3 b3 an bn b

then
Pn
j=1 (bj − aj ) = b1 − a1 +b2 − a2 +b3 − a3 + · · · +bn − an
≤ a2 − a +a3 − a2 +a4 − a3 + · · · +b − an
=b−a

therefore ∞
X
(bj − aj )
j=1

converges and has sum ≤ b − a

Let  > 0 and Pput 0 = 2 , 1 = 4 , . . . , j = 2j+1

,...
∞
so j > 0 and j=1 j = 
Then
(a − 0 , a + 0 ), (a1 , b1 + 1 ), (a2 , b2 + 2 ), . . .
is a family of open sets which cover [a, b]

11
 By the compactness of [a, b] ∃ a finite number of these open sets which
cover [a, b], and which by renumbering and discarding some intervals if nec-
essary we can take as:

(a − 0 , a + 0 ), (a1 , b1 + 1 ), . . . , (an , bn + n )
so that:
a − 0 a + 0 a2 b2 + 2 an bn + n

a a1 b1 + 1 a3 b3 + 3 an−1 bn−1 + n−1 b

b − a = a1 − a +a2 − a1 + · · · +an − an−1 +b − an

< 0 +(b +
P1 ∞ 1 − a 1 ) + · · · +(b n−1 +  n−1 − a n−1 ) +(b n + n − a n )
< + j=1 (bj − aj )

is true ∀  > 0. Therefore

∞
X
b−a≤ (bj − aj )
j=1

Therefore ∞
X
(bj − aj ) = b − a
j=1

as required.
Let m :−→ [0, ∞] be a measure m on an algebra A of subsets of X.
 
 T henm(E) ≤ m(F ) 
E ⊂ F =⇒ and
 
m(F ∩ E 0 ) = m(F ) − m(E) if m(F ) 6= ∞
Then
F = E ∪ (F ∩ E 0 )

is a disjoint union, and

m(F ) = m(E) + m(F ∩ E 0 )

12
Theorem 1.3.2. (m is subadditive on countable unions)
∞
[ ∞
X
m( )≤ m(Ei )
j=1 i=1

Proof. We proceed as follows:

Let
Fn = E 1 ∪ E 2 ∪ · · · ∪ E n

E1 E2 E3 En

∞
[
Ej = E1 ∪ (E2 ∩ F1 0 ) ∪ (E3 ∩ F2 0 ) ∪ · · ·
j=1
S∞
m( j=1 Ej ) = m(E1 ) +m(E2 ∩ F1 0 ) +m(E3 ∩ F2 0 ) + · · ·
≤ m(E1 ) +m(E2 ) +m(E3 ) +···
hence the result: ∞ ∞
[ X
m( Ej ) ≤ m(Ej )
j=1 j=1

m is subadditive on countable unions.

1.4 The σ-algebra

Our aim is to extend the notion of length to a much wider class of subsets of
R. In particular to sets obtainable from I by a sequence of taking countable
unions and taking complements.
Definition an algebra A of subsets of X is called a σ-algebra if for each
sequence
E1 , E2 , E3 , . . .
of elements of A, their union
E1 ∪ E 2 ∪ E 3 ∪ · · ·
is also an element of A.
So A is closed under countable unions, and hence also under countable
intersections.

13
1.5 The outer measure
Definition We define the outer measure m̂ associated with m to be the
function:
m̂ : {all subsets of X} −→ [0, ∞]
given by:
∞
X
m̂(E) = inf m(Ej )
j=1

where the inf is taken over all sequences Ej of elements of A such that:
∞
[
E⊂ Ej
j=1

Theorem 1.5.1. m̂(E) = m(E) ∀E ∈ A (i.e. m̂ agrees with m on A)

Proof. 1. if
E∈A
and ∞
[
E⊂ Ej with Ej ∈ A
j=1

then: ∞
X
m(E) ≤ m(Ej )
j=1

therefore
m(E) ≤ m̂(E)

2. if
E∈A
then
E, ∅, ∅, . . .
is a sequence of elements of A whose union contains E. Therefore:

m̂(E) ≤ m(E) + 0 + 0 + · · ·

therefore
m̂(E) ≤ m(E)

14
Theorem 1.5.2. E ⊂ F =⇒ m̂(E) ≤ m̂(F )

Proof. if E ⊂ F then each sequence in A which covers F will also cover E.

Therefore:
m̂(E) ≤ m̂(F )

Theorem 1.5.3. m̂ is subadditive on countable unions

∞
[ ∞
X
m̂( Ej ) ≤ m̂(Ej )
j=1 j=1

for all sequences E1 , E2 , . . . of subsets of X.

Proof. Let  > 0. Choose:

0 , 1 , 2 , . . . > 0

such that
∞
X
j = 
j=0

choose Bij ∈ A s.t.

Ei ⊂ Bi1 ∪ Bi2 ∪ · · · ∪ Bij ∪ · · ·

and s.t.
∞
X
m(Bij ) ≤ m̂(Ei ) + i
j=1

Then
∞
[ ∞ [
[ ∞
Ei ⊂ Bij
i=1 i=1 j=1

and
∞ X
X ∞ ∞
X
m(Bij ) ≤ m̂(Ei ) + 
i=1 j=1 i=1

therefore
∞
[ ∞
X
m̂( Ej ) ≤ m̂(Ej )
j=1 j=1

as required.

15
Definition We call a subset E ⊂ X measurable w.r.t m if:
m̂(A) = m̂(A ∩ E) + m̂(A ∩ E 0 )
for all A ⊂ X
(E splits every set A into two pieces whose outer measures add up.)
A
2
1

1.6 Extension of measure to σ-algebra, using

outer measure
We can now prove the central:
Theorem 1.6.1. Let m be a measure on an algebra A of subsets of X, m̂
the associated outer measure, and M the collection of all subsets of X which
are measurable with respect to m.
Then M is a σ-algebra containing A and m̂ is a (countably additive)
measure on M .
Corollary 1.6.2. the measure m on the algebra A can be extended to a
measure (also denoted by m) on the σ-algebra M by defining:
m(E) = m̂(E) ∀E ∈ M
in particular:
Corollary 1.6.3. the length measure m on the interval algebra I can be
extended to a measure (also denoted by m) on the σ-algebra M of measurable
sets w.r.t. m. We call the elements of M the Lebesgue Measureable sets,
and the extended measure the (one-dimensional) Lebesgue Measure.

I ⊂ {Lebesgue Measurable Sets} ⊂ all subsets of R

length measure m Lebesgue measure outer measure m̂

(countably additive) (not countably additive
=⇒ not a measure)
& ↓ .
[0, ∞]

16
Proof. 1. A ⊂ M
Let E ∈ A, let A ⊂ X
Need to show:
m̂(A) = m̂(A ∩ E) + m̂(A ∩ E 0 )
For  > 0. Let
F1 , F2 , . . .
be a sequence in A s.t.
∞
[
A⊂ Fj
j=1

and s.t.
∞
X
m(Fj ) ≤ m̂(A) + 
j=1

A
Fi

Then

m̂(A) ≤ m̂(A ∩ E) +m̂(A ∩ E 0 ) since m̂ is subadditive

S∞ S∞
≤ m̂( j=1 Fj ∩ E) +m̂( j=1 Fj ∩ E 0 )
P∞ P∞
≤ j=1 m̂(Fj ∩ E) + j=1 m̂(Fj ∩ E 0 )
P∞ P∞
= j=1 m(Fj ∩ E) + j=1 m(Fj ∩ E 0 ) since Fj ∩ E, Fj ∩ E 0 ∈ A
and m̂ = m on A
P∞
= j=1 m(Fj ) since m is additive

≤ m̂(A) + 

17
 therefore,
m̂(A) ≤ m̂(A ∩ E) + m̂(A ∩ E 0 ) ≤ m̂(A) + 
∀ > 0, and therefore
m̂(A) = m̂(A ∩ E) + m̂(A ∩ E 0 )
as required.
2. M is an algebra
let E, F ∈ M ; Let A ⊂ X

4
3

1 2

Then
m̂(A) = m̂(1 + 2) +m̂(3 + 4) since E ∈ M

= m̂(1 + 2) +m̂(3) +m̂(4) since F ∈ M

= m̂(1 + 2 + 3) +m̂(4) since E ∈ M

therefore
E∪F ∈M
and M closed under finite unions.
Also, M closed under complements by symmetry in the definition.
3. M is a σ-algebra
S
Let E = ∞ i=1 Ei be a countable disjoint union with Ei ∈ M . We need
to show that E ∈ M . Put
Fn = E 1 ∪ E 2 ∪ · · · ∪ E n

18
 A

E1 E2 En

Fn

Fn ∈ M since M is an algebra. Let A ⊂ X. Then

m̂(A) = m̂(A ∩ Fn ) +m̂(A ∩ Fn 0 ) since Fn ∈ M

Pn
= k=1 m̂(A ∩ Ek ) +m̂(A ∩ Fn 0 ) since E1 , . . . , En ∈ M,
and are disjoint
Pn
≥ k=1 m̂(A ∩ Ek ) +m̂(A ∩ E 0 ) since E 0 ⊂ Fn 0

is true ∀n. Therefore:

P 
m̂(A) ≥ ∞ 0
k=1 m̂(A ∩ Ek ) +m̂(A ∩ E ) 


S∞ (*)
≥ m̂( k=1 A ∩ Ek ) +m̂(A ∩ E 0 ) since m̂ is countably subadditive 


= m̂(A ∩ E) +m̂(A ∩ E 0 )

≥ m̂(A) since m̂ is subadditive

All the above are equalities:

m̂(A) = m̂(A ∩ E) + m̂(A ∩ E 0 )

therefore
E∈M
and M is closed under countable disjoint unions. But any countable
union:
E1 ∪ E 2 ∪ E 3 ∪ · · ·

19
 can be written as a countable disjoint union

E1 ∪ (E2 ∩ F1 0 ) ∪ (E3 ∩ F2 0 ) ∪ · · ·

where
Fn = E 1 ∪ · · · ∪ E n
therefore M is closed under countable unions as required.
4. m̂ is countably additive on M
Put A = E in (*) to get
∞ ∞
!
X [
m̂(Ek ) = m̂ Ek
k=1 k=1

as required.

1.7 Increasing Unions, Decreasing Intersec-

tions
Definition We use the notation Ej ↑ E to denote that

E
E2
E1

E1 ⊂ E 2 ⊂ · · · ⊂ E j ⊂ · · ·
is an increasing sequence of sets such that
∞
[
Ej = E
j=1

Definition We use the notation Ej ↓ E to denote that

E1
E2
E

20
 E1 ⊃ E 2 ⊃ · · · ⊃ E j ⊃ · · ·
is a decreasing sequence of sets such that
∞
\
Ej = E
j=1

Theorem 1.7.1. Ej ↑ E =⇒ limj→∞ m(Ej ) = m(E)

Proof. 1. if m(Ej ) = ∞ for some j then the result holds

2. if m(Ej ) is finite ∀j then

E = E1 ∪ (E2 ∩ E1 0 ) ∪ (E3 ∩ E2 0 ) ∪ · · ·

is a countable disjoint union and

m(E) = m(E1 ) + [m(E2 ) − m(E1 )] + [m(E3 ) − m(E2 )] + · · ·

 
= limn→∞ m(E1 ) + [m(E2 ) − m(E1 )] + · · · + [m(En ) − m(En−1 )]

= limn→∞ m(En )

as required.

Theorem 1.7.2.

Ej ↓ E
=⇒ lim m(Ej ) = m(E)
m(E1 ) 6= ∞ j→∞

Proof.
(E1 ∩ Ej 0 ) ↑ (E1 ∩ E 0 )
therefore:
lim m(E1 ∩ Ej 0 ) = m(E1 ∩ E 0 )
j→∞

and
lim [m(E1 ) − m(Ej )] = m(E1 ) − m(E)
j→∞
so
lim m(Ej ) = m(E)
j→∞

as required.

21
Example
1
(a − , b] ↓ [a, b]
n

a−
1 a b
n

for Lebesgue measure m:

m[a, b] = limn→∞ m(a − n1 , b]

= limn→∞ (b − a + n1 )

=b−a

as expected.

1.8 Properties of Lebesgue Measure

We now show that the Lebesgue measure is the only way of extending the
‘length’ measure on the interval algebra I to a measure on the Lebesgue
measurable sets.
We need:

Definition a measure m on an algebra A of subsets of X is called σ-finite

if ∃ a sequence Xi in A such that

Xi ↑ X

and m(Xi ) is finite for all i.

Example the Lebesgue measure m is σ-finite because:

(−n, n] ↑ R

and m(−n, n] = 2n is finite

−n
 O n

22
Theorem 1.8.1. (Uniqueness of Extension) Let m be a σ-finite measure on
an algebra A of subsets of X.
Let M be the collection of measurable sets w.r.t. m.
Let l be any measure on M which agrees with m on A.
Then l(E) = m̂(E) ∀E ∈ M

Proof. Let E ∈ M . Then for each sequence {Ai }, Ai ∈ A covering E:

∞
[
E⊂ Ai
j=1

we have
P∞ P∞
l(E) ≤ i=1 l(Ai ) = i=1 m(Ai ) since l = m on A

Therefore,

(∗) l(E) ≤ m̂(E) by definition of the outer measure m̂

Now let Xi ↑ X and m(Xi ) finite, Xi ∈ A. Consider

Xi

l(Xi ∩ E) + l(Xi ∩ E 0 ) = l(Xi ) = m(Xi ) = m̂(Xi ∩ E) + m̂(Xi ∩ E)

By (*) it follows that
l(Xi ∩ E) = m̂(Xi ∩ E)
Take limi→∞ to get l(E) = m̂ as required.

As a consequence we have:

Lemma 1.8.2. Let m be the Lebesgue measure on R. Then for each mea-
surable set E and each c ∈ R we have:

1. m(E + c) = m(E). m is translation invariant

2. m(cE) = |c|m(E)

Proof. Let M be the collection of measurable sets

23
 1. define a measure mc on M by

mc (E) = m(E + c)

then

mc (a, b] = m(a + c, b + c] = (b + c) − (a + c) = b − a = m(a, b]

therefore mc agrees with m on the interval algebra I. Therefore mc

agrees with m on M .
2. define a measure mc on M by

mc (E) = m(cE)

Then
mc (a, b] = m(c(a, b])
 
 m(ca, cb] c > 0 
= m{0} c=0
 
m[cb, ca) c < 0
 
 cb − ca 
= 0
 
ca − cb

= |c|(b − a)

= |c|m(a, b]
Therefore mc agrees with |c|m on the interval algebra. Therefore mc
agrees with |c|m on M .

1.9 Borel Sets

Definition if V is any collection of subsets of X, we denote by G(V) the
intersection of all the σ-algebras of subsets of X which contain V. We have:
1. G(V) is a σ-algebra containing V
2. if W is any σ-algebra which contains V then

V ⊂ G(V) ⊂ W

24
Thus G(V) is the smallest σ-algebra of subsets of X which contains V. G(V)
is called the σ-algebra generated by V.

Definition The σ-algebra generated by the open sets of R is called the

algebra of Borel Sets of R.

Theorem 1.9.1. The σ-algebra generated by the interval algebra I is the

algebra of Borel sets of R.

Proof. Let V be the collection of open sets of R

T
1. (a, b] = ∞ 1
n=1 (a, b + n ) is the intersection of a countable family of open
sets. Therefore (a, b] ∈ G(V).
Therefore
I ⊂ G(V)
and
G(I) ⊂ G(V)

2. let V be an open set in R. For each a ∈ V choose an interval Ia ∈ I

with rational endpoints s.t. a ∈ Ia ⊂ V .

Ia
 
 

a V

[
V = Ia
a∈V

so V is the union of a countable family of elements of I. Therefore

V ∈ G(I)

implies
V ⊂ G(I) =⇒ G(V) ⊂ G(I)

So, combining these two results,

G(I) = G(V) = algebra of Borel sets

25
 We have:

I ⊂ Borel Sets in R ⊂ Lebesgue measurable Sets in R

The following theorem shows that any Lebesgue measurable set in R can
be obtained from a Borel set by removing a set of measure zero.
Theorem 1.9.2. Let E be a Lebesgue measurable subset of R. Then there
is a Borel set B containing E such that

B − E = B ∩ E0

has measure zero, and hence

m(B) = m(E)

B
B-E
E

Proof. 1. Suppose m(E) is finite. Let k be an integer > 0. Choose a

sequence
I1 , I2 , I3 , . . .
in I such that ∞
[
E⊂ Ii
i=1

and s.t. ∞
X 1
m(E) ≤ m(Ii ) ≤ m(E) +
i=1
k
S∞
Put Bk = i=1 Ii then E ⊂ Bk , Bk is Borel, and
1
m(E) ≤ m(Bk ) ≤ m(E) +
k
T∞
Put B = k=1 Bk . Then E ⊂ B, B is Borel, and
1
m(E) ≤ m(B) ≤ m(E) +
k
for all k. So
m(E) = m(B) =⇒ m(B − E) = 0
since m(E) is finite.

26
2. Suppose m(E) = ∞.

−(k + 1) −k k k+1

Put
Ek = {x ∈ E : k ≤ |x| < k + 1}
Then ∞
[
E= Ek
k=0

is a countable disjoint union and m(Ek ) is finite.

For each integer k choose a Borel set Bk s.t.

Ek ⊂ B k

and
m(Bk − Ek ) = 0
S∞
Put B = k=1 Bk . Then E ⊂ B, B is Borel, and
∞
[
B−E ⊂ (Bk − Ek )
k=1

therefore ∞ ∞
X X
m(B − E) ≤ m(Bk − Ek ) = 0=0
k=1 k=1

as required.
B2
B1

E1 E2

27
28
Chapter 2

Integration

2.1 Measure Space, Measurable sets

Definition We fix a set X, a σ-algebra M of subsets of X, and a measure
m on M . The triple (X, M, m) is then called a measure space, the elements
of M are called the measurable sets of the measure space.

Definition We call a function

f : X −→ [−∞, ∞]

measurable if
f −1 (α, ∞] = {x ∈ X : f (x) > α}
is a measurable ∀α ∈ R.
R

α f

>α >α

Example The collection

{E ⊂ R : f −1 (E) is measurable}

29
is a σ-algebra containing the sets
{(α, β] : α, β ∈ R}
therefore contains the Borel sets. Therefore f −1 is measurable for each Borel
set B ⊂ R.

2.2 Characteristic Function

Definition If E ⊂ X, we denote by χE the function on X:

1 x∈E
χE (x) =
0 x 6∈ E
χE is called the characteristic function of E.
R

E E

Definition A real valued function

φ : X −→ R
is called simple if it takes only a finite number of distinct values

a1 , a2 , . . . , an
(say). Each simple function φ can be written in a unique way as:
φ = a 1 χE 1 + · · · + a n χE n
where a1 , . . . , an are distinct and

X = E1 ∪ · · · ∪ En
is a disjoint union.

30
2.3 The Integral
Definition If φ is a non-negative measurable simple function with φ =
a1 χE1 + · · · + an χEn ; X = E1 ∪ · · · ∪ En disjoint union we define the integral
of φ w.r.t. the measure m to be
Z
φ dm = a1 m(E1 ) + · · · + an m(En ) ∈ [0, ∞]

a1

a2

E1 E2

Recall that 0.∞ = 0, a.∞ = ∞ if a > 0.

Definition if E is a measurable subset of X and φ is a non-negative mea-

surable simple function on X, we define the integral of φ over E w.r.t. the
measure m to be: Z Z
φ dm = φχE dm
E
We note that:
R R R
1. (φ + ψ) dm = φ dm + ψ dm
R R
2. cφ dm = c φ dm ∀c ≥ 0
To see 1. we put: X
φ= a i χE i
X
ψ= b j χFj
Let {ck } be the set of distinct values of {ai + bj }. Then
X
φ+ψ = c k χ Gk
S
where Gk = Ei ∩ Fj , the union taken over {i, j : ai + bj = ck }.

31
 R P
(φ + ψ) = ck m(Gk )
P P
= k ck {i,j:ai +bj =ck } m(Ei ∩ Fj )
P
= i,j (ai + bj )m(Ei ∩ Fj )
P P
= ai m(Ei ∩ Fj ) + bj m(Ei ∩ Fj )
P P
= i ai m(Ei ) + j bj m(Fj )

R R
= φ dm + ψ dm
which proves 1.

A very useful property of the integral is:

Theorem 2.3.1. Fix a simple non-negative measurable function φ on X.

For each E ∈ M put Z
λ(E) = φ dm
E

Then λ is a measure on M .

Proof. Let
φ = a 1 χE 1 + · · · a n χE n
for ai ≥ 0. Then
R
λE = E
φ dm
R
= φχE dm
R
= (a1 χE1 ∩E + · · · + an χEn ∩E ) dm

= a1 m(E1 ∩ E) + · · · +an m(En ∩ E)

= a1 mE1 (E) + · · · +an mEn (E)

therefore
λ = a 1 mE 1 + · · · + a n mE n
is a linear combination of measures with non-negative coefficients. Therefore
λ is a measure.

32
 S∞
Corollary 2.3.2. 1. If E = n=1 En is a countable disjoint union then
Z ∞ Z
X
φ dm = φ dm
E n=1 En

S∞
2. If E = n=1 En is a countable increasing union then
Z Z
φ dm = lim φ dm
E n→∞ En

T∞ R
3. If E = n=1 En is a countable decreasing intersection and E1
φ dm <
∞ then Z Z
φ dm = lim φ dm
E n→∞ En

We can now define the integral of any non-negative measurable function.

Definition Let f : X −→ [0, ∞] be a measurable function. Then we define

the integral of f w.r.t. the measure m to be:
Z Z
f dm = sup φ dm
φ

where the sup is taken over all simple measurable functions φ such that:

0≤φ≤f

φ f

Definition If E is a measurable subset of X then we define the integral of

f over E w.r.t. m to be:
Z Z
f dm = f χE dm
E

33
 f

PSfrag replacements f χE
Ε

We then have:
R R
1. f ≤ g =⇒ f dm ≤ g dm
R R
2. E ⊂ F =⇒ E f dm ≤ F f dm

2.4 Monotone Convergence Theorem

We can now prove our first important theorem on integration.

Theorem 2.4.1. (Monotone Convergence Theorem, MCT) Let {fn } be a

monotone increasing sequence of non-negative measurable functions. Then
Z Z
lim fn dm = lim fn dm

Proof. We write fn ↑ f to denote that {fn } is an increasing sequence of

functions with lim fn = f . We need to prove that:
Z Z
lim fn dm = f dm

1.
fn ≤ fn+1 ≤ f
for all n. Therefore:
Z Z Z
fn dm ≤ fn+1 dm ≤ f dm

for all n. Therefore:

Z Z
lim fn dm ≤ f dm

34
2. Let φ be a simple measurable function s.t.:
0≤φ≤f

φ f

fn

Let 0 < α < 1. For each integer n > 0 put

En = {x ∈ X : fn (x) ≥ αφ(x)}
so En ↑ X. Now:

αφ
f

fn

En
Z Z Z
αφ dm ≤ fn dm ≤ fn dm
En En
Let n → ∞: Z Z
α φ dm ≤ lim fn dm

is true ∀0 < α < 1. Therefore:

Z Z
φ dm ≤ lim fn dm

so: Z Z
f dm ≤ lim fn dm

hence Z Z
f dm = lim fn dm

35
Definition In probability theory we have a measure space:
 
X , M , P
 sample events 
space

with P (X) = 1. X is the sure event.

A measurable function:
f : X −→ R
is called a random variable.

P {x : f (x) ∈ B}
is the probability that the random variable f takes value in the Borel set
B ⊂ R. If
f = a 1 χE 1 + · · · + a n χE n
with a1 , . . . , an ; E1 , . . . , En disjoint, and E1 ∪ · · · ∪ En = X, then the proba-
bility that f takes value ai is

P {x ∈ X : f (x) = ai } = P (Ei )

and Z
f dP = a1 P (E1 ) + · · · + an P (En )

is the average value or the expectation of f .

We define the expectation of any random variable f to be:
Z
E(f ) = f dP

2.5 Existence of Monotone Increasing Simple

Functions converging to f
In order to apply the MCT effectively we need:

Theorem 2.5.1. Let f be a non-negative measurable function f : X −→

[0, ∞]. Then there exists a monotone increasing sequence φn of simple mea-
surable functions converging to f

36
 Proof. Put each integer n > 0:
 k k k+1
2n 2n
≤ f (x) < 2n
k = 0, 1, 2, . . . , n2n − 1
φn (x) =
n f (x) ≥ n

k+1
2n

k
2n

PSfrag replacements

Then
1. 0 ≤ φn (x) ≤ φn+1 (x)

2. each φn is simple and measurable

3. limn→∞ φn (x) = f (x)

Theorem 2.5.2. Let f, g be non-negative measurable functions mapping X

to [0, ∞] and c ≥ 0. Then
R R
1. cf dm = c f dm
R R R
2. (f + g) dm = f dm + g dm
Proof. Let φn , ψn be monotonic increasing sequences of non-negative simple
functions with f = lim φn , g = lim ψn . Then
R M CT R R M CT R
1. cf dm = lim cφn dm = c lim φn dm = = c f dm
R M CT R R R M CT
2. R (f + g) dm
R = lim (φ n + ψ n ) = lim φ n dm + lim ψn dm =
f dm + g dm

This enables us to deal with series:

37
Theorem 2.5.3. Let fn be a sequence of non-negative measurable functions
X −→ [0, ∞]. Then:
Z X ∞ ∞ Z
X
( fn ) dm = fn dm
n=1 n=1

Proof. put sn = f1 + · · · + fn sum to n terms. sn is monotone increasing:

Z Z Z Xn X n Z
M CT
( lim sn ) dm = lim sn dm = lim fr dm = lim fr dm
n→∞ n→∞ n→∞ n→∞
r=1 r=1

Therefore: Z X ∞ ∞ Z
X
( fr ) dm = fr dm
r=1 r=1

Theorem 2.5.4. Let f : X −→ [0, ∞] be a non-negative measurable and put

Z
λ(E) = f dm
E
for each E ∈ M . Then λ is a measure in M .
S
Proof. Let E = ∞ k=1 Ek be a countable disjoint union. Then
R
λ(E) = E f dm
R
= f χE dm
R P
= ( ∞k=1 f χEk ) dm

M CT P∞ R
= k=1 f χEk dm
P∞ R
= k=1 Ek
f dm
P∞
= k=1 λ(Ek )
therefore λ is countably additive, as required.
S
Corollary 2.5.5. 1. if E = ∞ k=1 Ek countable disjoint union then
Z X∞ Z
f dm = f dm
E k=1 Ek
R R
2. if Ek ↑ E then limk→∞ Ek f dm = E f dm
R R R
3. If Ek ↓ E and E1 f dm < ∞, then limk→∞ Ek f dm = E f dm

38
 2.6 ’Almost Everywhere’
Definition Let f, g : X −→ [0, ∞]. Then we say that f = g almost every-
where (a.e.) or f (x) = g(x) almost all x ∈ X (a.a.x) if {x ∈ X : f (x) 6= g(x)}
has measure zero.

RTheorem 2.6.1. Let f : X −→ [0, ∞] be non-negative measurable. Then

f dm = 0 ⇔ f = 0 a.e.
Proof. Put E = {x ∈ X : f (x) 6= 0}

1. Put
1
En = {x ∈ X : f (x) > }
n
for each integer n > 0, so
1
f> χE
n n

1
n

PSfrag replacements
En

R
Suppose that f dm = 0. Then
Z
1
0 = f dm ≥ m(En )
n
so
m(En ) = 0
for all n. But En ↑ E, so

m(E) = lim m(En ) = lim 0 = 0

therefore
f = 0 a.e.

2. Suppose f = 0 a.e., so m(E) = 0. Now,

0 ≤ f ≤ lim nχE

39
 nχE

PSfrag replacements
E

therefore R
0 ≤ f dm
R
≤ lim nχE dm

M CT R
= lim nχE dm

= lim nm(E)

= lim 0 = 0
so Z
f dm = 0

Corollary 2.6.2. Let f : X −→ [0, ∞] be non-negative and measurable and

let E have measure zero. Then
Z
f dm = 0
E

Proof.
f χE = 0 a.e.
therefore Z Z
f dm = f χE dm = 0
E

R f, g : XR −→ [0, ∞] be non-negative and measurable

Corollary 2.6.3. Let
and f = g a.e. then f dm = g dm

40
Proof. Let f = g on E and m(E 0 ) = 0 then
Z Z Z Z Z Z
f dm = f dm + f dm = g dm + g dm = g dm
E E0 E E0

R
Thus changing f on a set of measure zero makes no difference to f dm.
Also

Theorem 2.6.4. If fn are non-negative and fn ↑ f a.e. then

Z Z
lim fn dm = f dm

Proof. suppose fn ≥ 0 and fn ↑ f on E with m(E 0 ) = 0. Then

R R R
f dm = E f dm + E 0 f dm

M CT R
= lim E
fn dm + 0
R R 
= lim f dm +
E n E0
fn dm
R
= lim fn dm

So far we have dealt with functions

X −→ [0, ∞]

which are non-negative, but have allowed the value ∞.

Now we look at functions
X −→ R
which may be negative and we do not allow ∞ as a value.

Definition Let f : X −→ R be measurable. Put

 
+ f (x) f (x) ≥ 0
f (x) =
0 f (x) ≤ 0
 
− −f (x) f (x) ≤ 0
f (x) =
0 f (x) ≥ 0

41
 f+

PSfrag replacements
f

f−
f

PSfrag replacements

Thus f = f + − f − and both f + , f − are non-negative. We say that f is

integrable w.r.t. m if
Z Z
+
f dm < ∞ and f − dm < ∞

and we write Z Z Z
+
f dm = f dm − f − dm

and call it the integral of f (w.r.t. measure m).

If E is measurable we write
Z Z Z Z
+
f dm = f χE dm = f dm − f − dm
E E E

Theorem 2.6.5. Let f = f1 − f2 where f1 , f2 are non-negative and measur-

able and Z Z
f1 dm < ∞ and f2 dm < ∞

Then f is integrable and

Z Z Z
f dm = f1 dm − f2 dm

42
Proof. 1. f = f1 − f2 , therefore f + ≤ f1 ; f − ≤ f2 . So
Z Z Z Z
+ −
f dm ≤ f1 dm < ∞; f dm ≤ f2 dm < ∞

therefore f is integrable.
2. f = f1 − f2 = f + − f − . So

f1 + f − = f + + f 2

therefore
Z Z Z Z
− +
f1 dm + f dm = f dm + f2 dm

so Z Z Z Z Z
+ −
f1 dm − f2 dm = f dm − f dm = f dm

as required.

Theorem 2.6.6. Let f, g be integrable and f = g a.e. Then

Z Z
f dm = g dm

Proof. R R
f + = g + a.e. =⇒ f + dm = g + dm
R R
f − = g − a.e. =⇒ f − dm = g − dm
so
Z Z Z Z Z Z
+ − + −
f dm = f dm − f dm = g dm − g dm = g dm

So when integrating we can ignore sets of measure zero.

Theorem 2.6.7. Let f : X −→ R be measurable. Then
1. f is integrable ⇔ |f | is integrable

2. if f is integrable then
Z  Z
 
 f dm ≤ |f | dm
 

43
Proof. 1.
R R
f integrable ⇔ f + dm < ∞ and f − dm < ∞
R
⇔ (f + + f − ) dm < ∞
R
⇔ |f | dm < ∞

⇔ |f | integrable

2. R  R R 
 f dm =  f + dm − f − dm
R R
≤ f + dm + f − dm
R
= |f | dm

Definition A complex valued function f : X −→ C with

f (x) = f1 (x) + if2 (x)

(say) (f1 , f2 real) is called integrable if f1 and f2 are integrable and we define:
Z Z Z
f dm = f1 dm + i f2 dm

Thus Z Z
< f dm = (<f ) dm
Z Z
= f dm = (=f ) dm

We have |f1 | ≤ |f |, |f2 | ≤ |f |, and therefore

|f | integrable ⇔ |f1 | and |f2 | integrable ⇔ f1 and f2 integrable ⇔ f
integrable.
We also have:
f, g integrable and c ∈ C =⇒ f + g and cf integrable, and
Z Z Z
(f + g) dm = f dm + g dm

and Z Z
(cf ) dm = c f dm

44
 Thus, the set L(X, R, m) of all integrable real valued functions on X is
a real vector space, and the set L(X, C, m) of all integrable complex valued
functions on X is a complex vector space. And on each space:
Z
f −→ f dm

is a linear form.

If f : X −→ C is integrable then
Z Z 
 
f dm =  f dm eiθ (say) θ real

therefore
R (real)  R
 f dm = e−iθ f dm

R (real)
−iθ
= e f dm
R
= <[e−iθ f ] dm
R  −iθ 
≤ e f  dm
R
= |f | dm
therefore Z  Z
 
 f dm ≤ |f | dm
 

2.7 Integral Notation

Definition When dealing with 1-dimensional Lebesgue measure we write
Z Z b Z a
f dm = f (x) dx = − f (x) dx
[a,b] a b

if a ≤ b. It follows that
Z b Z c Z c
f (x) dx + f (x) dx = f (x) dx
a b a

and Z b
dx = b − a
a

45
∀a, b, c.
Notice that x is a dummy symbol and that
Z b Z b Z b
f (x) dx = f (y) dy = f (t) dt = · · ·
a a a

just as
n
X n
X
a i bi = a j bj = · · ·
i=1 j=1

2.8 Fundamental Theorem of Calculus

Theorem 2.8.1. Fundamental Theorem of Calculus Let f : R −→ R be
continuous and put Z t
F (t) = f (x) dx
a
then
F 0 (t) = f (t)
Proof. Let t ∈ R, let  > 0. Then ∃δ > 0 s.t.

|f (t + h) − f (t)| ≤  ∀|h| ≤ δ

by continuity of f . Therefore
 
 F (t+h)−F (t) 
 h
− f (t) 
 R R t+h 
 t+h
=  h1 t f (x) dx − 1 
h t
f (t) dx
R 
1  t+h 
= |h|  t [f (x) − f (t)] dx

1
≤ |h|
|h| =

for all |h| ≤ δ; h 6= 0, and hence the result.

Corollary 2.8.2. If G : R −→ R is C 1 (i.e. has a continuous derivative)
Rb
then a G0 (x) dx = G(b) − G(a).
Rt
Proof. put F (t) = a G0 (x) dx. Then

F 0 (t) = G0 (t)

46
for all t. Therefore
F (t) = G(t) + c
for all t, where c is constant. Therefore
Z b
G(b) − G(a) = F (b) − F (a) = G0 (x) dx
a

as required.

Theorem 2.8.3. (Change of Variable)

Let
g f
[t3 , t4 ] −→ [t1 , t2 ] −→ R
with f continuous and g ∈ C 1 ; g(t3 ) = t1 , g(t4 ) = t2 . Then
Z t2 Z t4
f (x) dx = f (g(y))g 0(y) dy
t1 t3

Proof. Put Z t
G(t) = f (x) dx
t1

then G0 (t) = f (t), and therefore:

R t4 R t4
t3
f (g(y))g 0(y) dy = t3
G0 (g(y)g 0(y) dy
R t4 d
= t3
[ dy G(g(y))] dy

= G(g(t4 )) − G(g(t3 ))

= G(t2 ) − G(t1 )
R t2
= t1
f (x) dx

as required.

To deal with sequences which are not monotone we need the concepts of
lim inf and lim sup.

Definition Let
{an } = a1 , a2 , a3 , . . .

47
be a sequence in [−∞, ∞]. Put

bn = inf{an , an+1 , an+2 , . . .}

cn = sup{an , an+1 , an+2 , . . .}

Then

b1 ≤ b2 · · · ≤ bn ≤ bn+1 ≤ · · · ≤ cn+1 ≤ cn ≤ · · · ≤ c2 ≤ c1

Define:
lim inf an = lim bn
lim sup an = lim cn
Then
lim inf an ≤ lim sup an
and an converges iff lim inf an = lim sup an (= lim an ).

2.9 Fatou’s Lemma

Theorem 2.9.1. (Fatou’s Lemma) Let fn be a sequence of non-negative
measurable functions:
fn : X −→ [0, ∞]
Then Z Z
lim inf fn dm ≤ lim inf fn dm

Proof. Put
gr = inf{fr , fr+1 , . . .}
so that
· · · ≤ gr ≤ gr+1 ≤ · · ·
Now:
fn ≥ g r ∀n ≥ r
R R
=⇒ fn ≥ gr ∀n ≥ r
R R
=⇒ lim inf fn ≥ gr ∀r
R R M CT R R
=⇒ lim inf fn ≥ lim gr = lim gr = lim inf fn

as required.

48
2.10 Dominated Convergence Theorem
Theorem 2.10.1. Lebesgue’s Dominated Convergence Theorem, DCT Let
fn be integrable and fn → f . Let
|fn | ≤ g
R R
for all n where g is integrable. Then f is integrable and f = lim fn
Proof. |f | = lim |fn | ≤ g. Therefore |f | is integrable, and therefore f is
integrable. Now
g ± fn ≥ 0
therefore Z Z
lim inf[g ± fn ] ≤F AT OU lim inf [g ± fn ]
so Z Z Z
[g ± f ] ≤ g + lim inf ± fn
so Z Z
± f ≤ lim inf ± fn
Which gives: R R
: f ≤ lim inf fn
R R
: − f ≤ − lim sup fn
therefore Z Z Z Z
f ≤ lim inf fn ≤ lim sup fn ≤ f
which is equivalent to: Z Z
lim fn = f
as required.
For series this leads to:
Theorem 2.10.2. Dominated convergence theorem for series Let fn be a
sequence of integrable functions such that
X∞ Z
|fn | dm < ∞
n=1
P∞
Then n=1 fn is (equal a.e. to) an integrable function and
Z X ∞ X∞ Z
( fn ) dm = fn dm
n=1 n=1

49
Proof. Put
sn = f 1 + f 2 + · · · + f n
sum to n terms. Then
∞
X
|sn | ≤ |f1 | + · · · + |fn | ≤ |fr | = g
r=1

(say). Then
R R P∞
g dm = r=1 |fr | dm

M CT P∞ R
= r=1 |fr | dm < ∞
Therefore g is integrable (a.e. equal to an integrable fn). Therefore lim sn is
integrable and Z Z
DCT
lim sn dm = lim sn dm

therefore
∞
Z X ∞ Z
X
fr dm = fr dm
r=1 r=1

as required.

2.11 Differentiation under the integral sign

Another useful application.

Theorem 2.11.1. Differentiation under the integral sign Let f (x, t) be an

integrable function of x ∈ X for each a ≤ t ≤ b and differentiable w.r.t t.
Suppose  
 ∂f 
 (x, t) ≤ g(x)
 ∂t 

for all a ≤ t ≤ b, where g is integrable. Then

Z Z
d ∂f
f (x, t) dx = (x, t) dx
dt ∂t

Proof. Put Z
F (t) = f (x, t) dx

50
Let a ≤ t ≤ b. Choose a sequence tn in [a, b] s.t. lim tn = t and tn 6= t. Then,
by the Mean Value Theorem:
 
|f (x, tn ) − f (x, t)| =|tn − t|  ∂f
∂t
(x, c(n, x, t)) 

≤ |tn − t|g(x)

Therefore  
 f (x, tn ) − f (x, t) 
  ≤ g(x)
 tn − t 
so R
lim F (tntn)−F
−t
(t)
= lim f (x,tn )−f (x,t)
tn −t
dx

DCT R
= lim f (x,tntn)−f
−t
(x,t)
dx
R ∂f
= ∂t
(x, t) dx
and therefore Z
dF ∂f
= (x, t) dx
dt ∂t
as required.

51
52
 Chapter 3

Multiple Integration

3.1 Product Measure

We have established the Lebesgue measure and Lebesgue integral on R. To
consider integration on
R2 = R × R
we use the concept of a product measure.
We proceed as follows:
Definition Let l be a measure on a σ-algebra L of subsets of X. Let m be
a measure on a σ-algebra M of subsets of Y .
Call
{A × B : A ∈ L, B ∈ M}
the set of rectangles in X × Y

B A×B

PSfrag replacements
A

53
 S
Lemma 3.1.1. Let A×B = ∞ i=1 Ai ×Bi be a rectangle written as a countable
disjoint union of rectangles. Then
∞
X
l(A)m(B) = l(Ai )m(Bi )
i=1

Proof. We have
∞
X
χA×B = χAi ×Bi
i=1
∞
X
=⇒ χA (x)χB (y) = χAi (x)χBi (y)
i=1

Fix x and integrate w.r.t. m term by term using the Monotone Convergence
Theorem: ∞
X
χA (x)m(B) = χAi (x)m(Bi )
i=1

Now integrate w.r.t. x using MCT to get:

∞
X
l(A)m(B) = l(Ai )m(Bi )
i=1

as required.

Definition Let A be the collection of all finite unions of rectangles in X ×Y .

Each element of A is a finite disjoint union of rectangles. For each E ∈ A
such that ∞
[
E= Ai × B i
i=1

is a countable disjoint union of rectangles we define:

∞
X
π(E) = l(Ai )m(Bi )
i=1

Theorem 3.1.2. π is well-defined and is a measure on A.

Proof. 1. well-defined
Suppose
∞
[ ∞
[
E= Ai × B i = Cj × D j
i=1 j=1

54
 then ∞
[
Ai × B i = (Ai ∩ Cj ) × (Bi ∩ Dj )
j=1

Therefore, by Lemma 3.1.1

∞
X
l(Ai )m(Bi ) = l(Ai ∩ Cj )m(Bi ∩ Dj )
j=1

and hence,
P∞ P∞ P ∞
i=1 l(Ai )m(Bi ) = i=1 j=1 l(Ai ∩ Cj )m(Bi ∩ Dj )
P∞
= j=1 l(Cj )m(Dj )

Dj

PSfrag replacements
Bi

Di
Bj Ai Cj
S∞
2. countably additive Let E = i=1 Ei be a countable disjoint union with
E, Ei ∈ A.
ni
[
Ei = Aij × Bij
j=1

is a finite disjoint union (say). Then

∞ [
[ ni
E= Aij × Bij
i=1 j=1

Therefore ∞ X
ni ∞
X X
π(E) = l(Aij )m(Bij ) = π(Ei )
i=1 j=1 i=1

Therefore π is countably additive, as required.

55
Definition The measure π on A extends to a measure (also denoted by π
and called the product of the measures l and m) on the σ-algebra (denoted
L × M) of all subsets of X × Y which are measurable w.r.t. π.
Similarly, if
(X1 , M1 , m1 ), . . . , (Xn , Mn , mn )
is a sequence of measure spaces then we have a product measure on a σ-
algebra of subsets of
X1 × · · · × X n
s.t.
π(E1 × · · · × En ) = m1 (E1 )m2 (E2 ) . . . mn (En )
In particular, starting with 1-dimensional Lebesgue measure on R we get
n-dimensional Lebesgue Measure on

R × · · · × R = Rn

Example If we have two successive, independent events with independent

probability measures, P1 , P2 , then the probability of the first event E and
the second event F is:
P (E × F ) = P [(E × Y ) ∩ (X × F )]

= P (E × Y )P (X × F ) by independence

= P1 (E)P2 (F )

(i.e. product measure)

3.2 Monotone Class

Definition A non-empty collection M of subsets of X is called a monotone
class if

1. En ↑ E, En ∈ M =⇒ E ∈ M

2. En ↓ E, En ∈ M =⇒ E ∈ M

i.e. M is closed under countable increasing unions and under countable

decreasing intersections.

We have: M is a σ-algebra =⇒ M is a monotone class. Therefore, there

are more monotone classes than σ-algebras.

56
Definition If V is a non-empty collection of subsets of X and if M is the
intersection of all the monotone classes of subsets of X which contain V, then
M is a monotone class, called the monotone class generated by V.
M is the smallest monotone class containing V and
V ⊂ monotone class ⊂ σ-algebra
generated by V generated by V

3.3 Ring of Subsets

Definition A non-empty collection R of subsets of X is called a ring of
subsets of X if: 
 E∪F ∈R
E, F ∈ R =⇒

E ∩ F0 ∈ R
i.e. R is closed under finite unions and under relative complements.
Example The collection of all finite unions of rectangles contained in the
interior X of a fixed circle in R2 is a ring of subsets of X.
Note:
1. if R is a ring then R is closed under finite intersections since E ∩ F =
E ∩ (E ∩ F 0 )0 .
2. if R is a ring then ∅ ∈ R since E ∩ E 0 = ∅
3. if R is a ring of subsets of X and if X ∈ R then R is an algebra of
subsets of X since E 0 = X ∩ E 0 , so R is closed under complements.
Theorem 3.3.1. (Monotone class lemma)
Let R be a ring of subsets of X and let M be the monotone class generated
by R.
Then M is a ring.
Proof. We have to show M is closed under finite unions and relative comple-
ments. i.e. that:
E ∩ F 0 , E ∪ F, E 0 ∩ F ∈ M
for all E, f ∈ M . So for each E ∈ M put
ME = {F ∈ M : E ∩ F 0 , E ∪ F, E 0 ∩ F ∈ M }
and we must show
ME = M
for all E ∈ M . Now

57
 1. ME ⊂ M by definition

2. ME is a monotone class, because:

Fn ↑ F, Fn ∈ ME

=⇒ (E ∩ Fn 0 ) ↓ (E ∩ F 0 ) (E ∪ Fn ) ↑ (E ∪ F ) (E 0 ∩ Fn ) ↑ (E 0 ∩ F )

with E ∩ Fn 0 E ∪ Fn E 0 ∩ Fn all ∈ M

=⇒ E ∩ F 0 E∪F E0 ∩ F all ∈ M
(M i monotone

=⇒ F ∈ ME

therefore ME is closed under countable increasing unions, and similarly

ME is closed under countable decreasing intersections.

3. E ∈ R
=⇒ R ⊂ ME since R is a ring. Therefore ME = M by (i), (ii) since
M is smallest monotone class containing R.

4. F ∈ ME ∀E ∈ R, F ∈ M by (iii). Therefore E ∈ MF ∀E ∈ R, F ∈
M.
Therefore
R ⊂ MF ∀F ∈ M
and therefore
MF = M ∀F ∈ M
as required.

Corollary 3.3.2. Let M be the monotone class generated by a ring R, and

let X ∈ M .
Then M = G(R) the σ-algebra generated by R.

Proof. M is a ring and X ∈ M

E ∈ M =⇒ E 0 = E 0 ∩ X ∈ M

therefore M closed under compliments, and therefore M is an algebra.

58
 S∞
Also, if En is a sequence in M and E = 1 En put

Fn = E 1 ∪ · · · ∪ E n ∈ M

then Fn ↑ E, and therefore E ∈ M .

Therefore, M is a σ-algebra, hence the result.

Corollary 3.3.3. Let M be the monotone class generated by an algebra A.

Then M = G(A) the σ-algebra generated by A.

Proof. X ∈ A =⇒ X ∈ M , and hence the result by previous corollary.

3.4 Integration using Product Measure

Let l be a measure on a σ-algebra L of subsets of X.
Let m be a measure on a σ-algebra M of subsets of Y .
Let A be the algebra of finite unions of rectangles A × B; A ∈ L, B ∈ M.
Let π be the product measure on the σ-algebra G(A).
If
f : X −→ [0, ∞]

g : Y −→ [0, ∞]

F : X × Y −→ [0, ∞]

are non-negative measurable, write:

R R
f dl = X
f (x) dx
R R
f dm = Y
g(y) dy
R R
F dπ = X×Y
F (x, y) dx dy

If E ⊂ X × Y write

Ex = {y ∈ Y : (x, y) ∈ E}

E y = {x ∈ X : (x, y) ∈ E}

59
 y

Ex

x E
y

Theorem 3.4.1. if E ∈ G(A) and l, m are σ-finite then:

Z Z
π(E) = m(Ex ) dx = l(E y ) dy
X Y

Proof. To show
Z
π(E) = m(Ex ) dx (3.1)
X

1. Suppose l, m are finite measures: l(X) < ∞, m(Y ) < ∞. Let N be

the collection of all E ∈ G(A) s.t. Equation (3.1) holds. We will show
that N is a monotone class containing A:

A ⊂ N ⊂ G(A)

and hence N = G(A) since by the corollary to the monotone class

lemma, G(A) is the monotone class generated by A.

(a) A ∈ N .
If E = A × B is a rectangle then

m(Ex ) = χA (x)m(B)

60
 E

then Z
m(Ex ) dx = l(A)m(B) = π(E)
X

therefore E ∈ N . Now each element of A can be written as a

finite disjoint union of rectangles, therefore

A⊂N

(b) Let En ↑ E, En ∈ N . So,

(En )x ↑ Ex

for each x ∈ X.

En

61
 Then R R
X
m(Ex ) dx = X
lim m((En )x ) dx

M CT R
= lim X
m((En )x ) dx

= lim π(En )

= π(E)

therefore E ∈ N , and N is closed under converging unions.

(c) Let E ∈ N . Then
R R
X
m((E 0 )x ) dx = X
[m(Y ) − m(Ex )] dx

= l(X)m(Y ) − π(E)x

= π(X × Y ) − π(E)

= π(E 0 ).

Therefore E 0 ∈ N , and N is closed under complements. Therefore

N is a monotone class, and therefore N = G(A).

2. Suppose l, m are σ-finite,

An ↑ X, Bn ↑ Y

(say) with
l(An ) < ∞, m(Bn ) < ∞

Then
Zn ↑ (X × Y )

where Zn = An × Bn and π(Zn ) < ∞.

Let E ∈ G(A). Then (E ∩ Zn ) ↑ E, and

(E ∩ Zn )x ↑ Ex

for all x ∈ X.

62
 Zn E

So: R R
X
m(Ex ) dx = X
lim m((E ∩ Zn )x ) dx

M CT R
= lim X
m((E ∩ Zn )x ) dx

(since Zn has finite measure)

= lim π(E ∩ Zn )

= π(E)
as required.

3.5 Tonelli’s Theorem

Theorem 3.5.1. (Repeated integral of a non-negative funtcion)
Let F : X × Y −→ [0, ∞] be non-negative measurable. Then
Z Z  Z Z Z 
F (x, y) dy dx = F (x, y) dx dy = F (x, y) dx dy
X Y X×Y Y X

(notation as before.)

63
Proof. To show
Z Z  Z
F (x, y) dy dx = F (x, y) dx dy (3.2)
X Y X×Y

1. Equation 3.2 holds for F = χE , since

Z Z  Z Z
χE (x, y) dy dx = m(Ex ) dx = π(E) = χE (x, y) dx dy
X Y X X×Y

therefore Equation 3.2 also holds for any simple function.

2. (General Case)
∃ monotone increasing sequence of non-negative measurable functions
with F = lim Fn
R R  R R 
X Y
F (x, y) dy dx = X Y
lim F n (x, y) dy dx

M CT R  R 
= X
lim Y
Fn (x, y) dy dx

M CT R R 
= lim X Y
Fn (x, y) dy dx

(1.) R
= lim X×Y
Fn (x, y) dx dy

M CT R
= X×Y
lim Fn (x, y) dx dy
R
= X×Y
F (x, y) dx dy

3.6 Fubini’s Theorem

Theorem 3.6.1. (Repeated Integral of an integrable function)
Let F be an integrable function on X × Y . Then:
Z Z  Z Z Z 
F (x, y) dy dx = F (x, y) dx dy = F (x, y) dx dy
X Y X×Y Y X
R R
Where Y F (x, y) dy is equal to an integrable function of x a.e., and X
F (x, y) dx
is equal to an integrable function of y a.e.

64
Proof.
R R R
X×Y
F (x, y) dx dy = X×Y
F + (x, y) dx dy − X×Y
F − (x, y) dx dy

T onelli R R  R R 
= X Y
F + (x, y) dy dx − X Y F − (x, y) dy dx
R R
Therefore Y F + (x, y) dy, and Y F − (x, y) dy are each finite a.a.x., and each
has a finite integral w.r.t. x. (*)
So:
F (x, y) = F + (x, y) − F − (x, y)
is integrable w.r.t y a.e., and:
Z Z Z
+
F (x, y) dy = F (x, y) dy − F − (x, y) dy
Y Y Y

for a.a.x, and is an integrable function of x (by (*)), with:

R R  R R +
 R R −

X Y
F (x, y) dy dx = X Y
F (x, y) dy dx − X Y
F (x, y) dy dx
R
= X×Y
F (x, y) dx dy

Theorem 3.6.2. Let f be an integrable function R −→ R. Then

R R
1. f (x + c) dx = f (x) dx
R 1
R
2. f (cx) dx = |c| f (x) dx (c 6= 0)

Proof. These are true for f = χE , because:

1.
Z Z Z
χE (x + c) dx = χE−c (x) dx = m(E − c) = m(E) = χE (x) dx

2.
Z Z Z
1 1 1
χE (cx) dx = χ 1 E (x) dx = m( E) = m(E) = χE (x) dx
c c |c| |c|

Therefore, these are true for f simple, and therefore true for f non-negative
measurable (by MCT), since ∃fn simple, s.t. fn ↑ f . Therefore, these are
true for f = f + − f − integrable.

65
 We Rnow see how toR deal with integration on change of variable on Rn .
1
Recall f (cx) dx = |c| f (x) dx.

Theorem 3.6.3. (Linear change of variable)

A
Let Rn −→ Rn be an invertable matrix, then
Z Z
1
f (Ax) dx = f (x) dx
| det A|

Proof. we can reduce A to the unit matrix I by a sequence of elementary

row operations.

1. To replace row i by row i + c row j, multiply A by:

 
..
1 .
 .. 
 .. 
 . . 
 
 ··· ··· 1 ··· c ··· ··· 
N = .. 

 . 


 1 

..
.

with c in the i,j position.

2. To interchange row i and row j, multiply A by:

 .. .. 
1 . .
 .. .. 

 1 . . 

 ··· ··· 0 ··· ··· 1 ··· 
 
 .. .. 

 . 1 . 

 .. 
P = . 
 .. . 

 . 1 .. 

 ··· ··· 1 ··· ··· 0 ··· 
 
 
 1 
 .. 
 . 
1

66
 3. To replace row i by c row j, multiply A by:
 
1
 .. 
 . 
 
 1 
 
D= c 
 
 1 
 .. 
 . 
1

Therefore, there exists matrices B1 , . . . , Bk , each of type N, P or D s.t.

B1 B2 · · · Bk A = I

and therefore,
A = Bk−1 · · · B2−1 B1−1
is a product of matrices of type N, P or D.
Now, if the theorem holds for matrices A, B then it also holds for AB
since:
Z Z Z Z
1 1 1
f (ABx) dx = f (Ax) dx = f (x) dx = f (x) dx
| det B| | det B|| det A| | det AB|

therefore, it is sufficient to prove it for matrices of type N , P , D.

1. Let  
1 c 0 ··· 0

 0 1 0 ··· 0 


 0 0 1 0 ··· 

N = .. 
 . 
 .. 
 . 
1
(say). det N = 1. Then
R R
f (N x) dx = R f (x1 + cx2 , x2 , . . . , xn ) dx1 dx2 · · · dxn
= f (x1 , x2 , . . . , xn ) dx1 dx2 · · · dxn
(by Fubini, and translation invariance)

1
R
= | det N |
f (x) dx

67
 2. Let  
0 1 0 ··· 0

 1 0 0 ··· 0 


 0 0 1 0 ··· 

P = .. 
 . 
 .. 
 . 
1
(say). det P = −1. Then
R R
f (P x) dx = R f (x2 , x1 , . . . , xn ) dx1 dx2 · · · dxn
= f (x1 , x2 , . . . , xn ) dx1 dx2 · · · dxn
(by Fubini)

1
R
= | det P |
f (x) dx

3. Let  
c 0 0 ··· 0

 0 1 0 ··· 0 


 0 0 1 0 ··· 

D= .. 
 . 
 .. 
 . 
1
(say). det N = c. Then
R R
f (Dx) dx = fR(cx1 , x2 , . . . , xn ) dx1 dx2 · · · dxn
1
= |c| f (x1 , x2 , . . . , xn ) dx1 dx2 · · · dxn
(by Fubini)

1
R
= | det D|
f (x) dx

A
Corollary 3.6.4. if E ⊂ Rn is measurable and Rn −→ Rn is a linear homo-
morphism then
m(A(E)) = | det A|m(E)
Proof. R
m(E) = R χE (x) dx
= χA(E) R (AX) dx
= | det1 A| χA(E) (x) dx
= | det1 A| m(A(E))

68
as required.

69
70
Chapter 4

Differentiation

4.1 Differentiation
f
If R −→ R is a real valued function od a real variable then the derivative of
f at a is defined to be:
f (a + h) − f (a)
f 0 (a) = lim (4.1)
h−→0 h
We want to define the derivative f 0 (a) when f is a vector-valued function of
a vector variable:
f : M −→ N
where m, N are real (or complex) vector spaces.
We cannot use Equation (4.1) directly since we don’t know how to divide
f (a + h) − f (a), which is a vector in N , by h, which is a vector in M
So, we rewrite Equation (4.1) as:
f (a + h) = f (a) + f 0 (a)h + φ(h)
|{z} | {z } |{z}
(constant) (linear in h) (remainder)

where
φ(h) f (a + h) − f (a)
lim = − f 0 (a)
h−→0 h h
This suggests that we take M, N to be normed spaces and define f 0 (a) to
be a linear operator such that
f (a + h) = f (a) + f 0 (a)h + φ(h)
where
kφ(h)k
lim =0
khk−→0 khk

71
 a+h a+h

f φ(h)

h f’(a) h

f(a+h)
a f(a)

This f 0 (a)h is the linear approx to the change in f when variable changes by
h from a to a + h.

4.2 Normed Space

Definition let M be a real or complex vector space. Then M is called a
normed space if a function k.k exists,:
M −→ R
x −→ kxk
is given on M (called the norm on M ), such that:
1. kx ≥ 0
2. kxk = 0 ⇔ x = 0
3. kαxk = |α| kxk ∀ scalar α
4. kx + yk ≤ kxk + kyk (triangle inequality)
p
Example 1. Rn with k(α1 , . . . , αn )k = α12 + · · · + αn2 is called the Eu-
clidean Norm on Rn .
p
2. Cn with k(α1 , . . . , αn )k = |α1 |2 + · · · + |αn |2 is called the Hilbert
Norm on Cn .
3. Cn with k(α1 , . . . , αn )k = max{|α1 |, . . . , |αn |} is called the sup norm
on Cn .
4. if (X, M, m) is a measure space the the set of integrable functions
L0 (X, R, m) with Z
kf k = |f | dm

is called the L0 -norm (functions are to be regarded as equal if they are

equal a.e.)

72
4.3 Metric Space
Definition a set X is called a metric space if a function

D : X × X −→ R

is given (called a metric on X) such that

1. d(x, y) ≥ 0

2. d(x, y) = 0 ⇔ x = y

3. d(x, z) ≤ d(x, y) + d(y, z). This is known as the triangle inequality

Definition If a ∈ X and r > 0 the we write

BX (a, r) = {x ∈ X : d(a, x) < r}

and call it the ball in X centre a, radius r.

Example if M is a normed space then M is also a metric space with

d(x, y) = kx − yk

4.4 Topological space

Definition a set X is called a topological space id a collection V of subsets
of X is given (called the topology on X) such that:

1. ∅ and X belong to V
S
2. if {Vi }i∈I is any family of elements of V then i∈I vi belongs to V. i.e.
V is closed under unions

3. if U and V belong to V then U ∩ V belongs to V. i.e. V is closed under

finite intersections.

We call the elements of V the open sets of the topological space X, or open
in X.

Example Let X be a metric space. Then X is a topological space where we

define a set V to be open in V if V ⊂ X and each a ∈ V ∃ r > 0 s.t.

BX (a, r) ⊂ V

73
 V

r
a

Theorem 4.4.1. if X is a metric space then each c ∈ X and s > 0 the ball
BX (c, s) is open in X.

Proof. Let a ∈ BX (c, s). Put r = s − d(a, c) > 0.

Then

x ∈ BX (a, r)

=⇒ d(x, a) < r = s − d(a, c)

=⇒ d(x, a) + d(a, c) < s

=⇒ d(x, c) < s

=⇒ x ∈ BX (c, s)

74
 BX (c, s)

BX (a, r)

x
s r
c a

PSfrag replacements

and, therefore BX(a, r) ⊂ BX (c, s), as required.

4.5 Continuous map of topological spaces

Definition a map f : X −→ Y of topological spaces is called continuous if:

V open in Y =⇒ f −1 V open in X

Theorem 4.5.1. let X, Y be metric spaces, then:

f : X −→ Y

is continuous if and only if,

for each a ∈ X,  > 0

∃ δ > 0 s.t. (4.2)
d(x, a) < δ =⇒ d(f (x), f (a)) < 

i.e.
f BX (a, δ) ⊂ BY (f (a), )

Proof. 1. Let f be continuous. Let a ∈ X,  > 0. Then BY (f (a), ) is

open in Y . Therefore, f −1 BY (f (a), ) is open in X.

75
 f −1 BY (f (a), ) BY (f (a), )

ε
f(a)
PSfrag replacements δ
a

∃δ > 0 s.t.
BX (a, δ) ⊂ f −1 BY (f (a), )

and
f BX (a, δ) ⊂ BY (f (a), )

therefore Equation (4.2) holds.

2. Let Equation (4.2) hold. Let V be open in Y , a ∈ f −1 V , and so

f (a) ∈ V .
Now, ∃ > 0 s.t.
BY (f (a), ) ⊂ V

a
f −1 V
f(a)
PSfrag replacements

and ∃δ > 0 s.t.

f BX (a, δ) ⊂ BY (f (a), ) ⊂ V

therefore,
BX (a, δ) ⊂ f −1 V

so, f −1 V is open in X, and f is continuous.

76
4.6 Homeomorphisms
Definition a map f : X −→ Y of topolgical spaces is called homeomorphism
if:

1. f is bijective

2. f and f −1 are continuous

Definition X is homeomorphic to Y (topologically equivalent) if ∃ a home-

omorphism X −→ Y . i.e. ∃ a beijective map under which the open sets of
X correspond to the open sets of Y .

Definition a property P of a topological space is called a topological prop-

erty if X has property P and X homeomorphic to Y =⇒ Y has property
P.

Example ’compactness’ is a topolgical property

Note: we have a category with topological spaces as objects, and con-

tinuous maps as morphisms.

4.7 Operator Norm

Definition Let M, N be finite dimensional normed spaces. Then we can
make the vector space
L(M, N )
of all linear operators T : M −→ N , into a normed space by defining:

kT xk
kT k = sup
x∈M kxk
x6=0

kT k is called the operator norm of T .

We have:
x
1. If x 6= 0 and y = kxk
then

1
kyk = kxk = 1
kxk

77
 x

x
kxk
PSfrag replacements

i.e. y has unit norm. Therefore

kT xk x = kT yk
= T

kxk kxk

and
kT k = sup kT yk
y∈M
kyk=1

kT k

y
T y

PSfrag replacements

2. if α is a scalar then

kαT k = sup kαT yk = |α| sup kT yk = |α|kT k

kyk=1 kyk=1

3. if S, T : M −→ N then

kS+T k = sup k(S+T )yk ≤ sup (kSyk+kT yk) ≤ sup kSyk+sup kT yk = kSk+kT k
kyk=1 kyk=1

78
 4.
kT xk
kT k = 0 ⇐⇒ supx6=0 kxk
=0

⇐⇒ kT xk = 0 ∀x

⇐⇒ T x = 0 ∀x

⇐⇒ T = 0
(by 2,3,4 the operator norm is a norm).
Theorem 4.7.1. 1. if T : M −→ N and x ∈ M then
kT xk ≤ kT k kxk
T S
2. If L −→ M −→ N then
kST k ≤ kSk kT k
Proof. 1.
kT xk
≤ kT k
kxk
∀x 6= 0, be definition. therefore,
kT xk ≤ kT k kxk
for all x
2.
kST k = supkyk=1 kST yk

≤ sup kSk kT yk (by 1.)

≤ supkyk=1 kSk kT k kyk (by 1.)

= kSk kT k

Note:
1. if M is finite dimensional then every choice of norm on M defines the
same topology on M
2. a sequence xr of points in a topological space X is said to converge to
a ∈ M if, for each open set V containing a ∃N s.t. xr ∈ V ∀ r ≥ N .
For a normed space M this is the same as: for each  > 0 ∃N s.t.
kxr − ak <  ∀ r ≥ N .

79
4.8 Differentiation
Definition let
f
M ⊃ V −→ N
where M, N are normed spaces and V is open in M . Let a ∈ V . THen f is
differentiable at a if ∃ a linear operator
f 0 (a)
M −→ N

such that
f (a + h) = f (a0) + f 0 (a)h + φ(h)
where
kφ(h)k
−→ 0 as khk −→ 0
khk

Theorem 4.8.1. if f is differentiable at a the the operator f 0 (a) is uniquely

determined by:
f (a+th)−f (a)
f 0 (a)h = limt−→0 t

= directional derivative of f at a along h

d

= dt
f (a + th)t=0

a+h

a+th
a

f 0 (a) is called the derivative of f at a

Proof.
f (a + th) = f (a) + f 0 (a)th + φ(th)
therefore,

f (a + th) − f (a) 0
φ(th) kφ(th)k
− f (a)h =
= khk
t t kthk

tends to 0 as t −→ 0.

80
 f’(a) h
f(a+h)
h a+h
f(a+th)
a+th f
a
f(a)

Example 1. f : R −→ R, f (x) = x3 . Then

f (a + h) = (a + h)3 = |{z}
a3 + 3a 2
|{z} h + 3ah 2 3
| {z+ h}
f (a) f 0 (a)h φ(h)

so f 0 (a) = 3a2 .

2. f : Rn×n −→ Rn×n , f (X) = X 3 .

f (A + H) = (A + H)3

A3 + A
= |{z} 2
| H + AHA
2
{z + HA}
f (A) f 0 (A)H

2
+ AH
| + HAH{z+ H 2 A + H}3
φ(H)

taking (say) Euclidean norm on Rn and operator norm on Rn×n =

L(Rn , Rn ).
kφ(H)k kAH 2 +HAH+H 2 A+H 3 k
kHk
= kHk

3kAk kHk2
≤ kHk

= 3kAk kHk −→ 0
as kHk −→ 0
therefore, f is differentiable at A, and

f 0 (A) : Rn×n −→ Rn×n

is given by
f 0 (A)H = A2 H + AHA + HA2

81
 f
Theorem 4.8.2. Let Rn ⊃ V −→ Rm be a differentiable function with

f = (f 1 , . . . , f m ) = (f i )

Then  
0 ∂f i
f =
∂xj
Proof. For each a ∈ V we have

f 0 (a) : Rn −→ Rm

is a linear operator. Therefore, f 0 (a) is an m × n matrix whose j th column is

f (a+tej )−f (a)
f 0 (a)ej = limt−→0 t

∂f
= ∂xj
(a)
 
∂f 1 m
= ∂xj
(a), . . . , ∂f
∂xj
(a)

therefore,  
0 ∂f i
f (a) = (a)
∂xj
for all a ∈ V . Therefore,  
0 ∂f i
f =
∂xj
which is the Jacobian matrix.
f
Theorem 4.8.3. Let M ⊃ V −→ N1 × · · · × Nk where


f (x) = f 1 (x), . . . , f k (x)

Then f 1 , 
. . . , f k differentiable at a ∈ V =⇒ f is differentiable at a, and
0 0
f 0 (a)h = f 1 (a)h, . . . , f k (a)h .

Proof. (k = 2)
f
M ⊃ V −→ N1 × N2


f (x) = f 1 (x), f 2 (x)
Take any norms on M, N1 , N2 and define a norm on N! × N2 by

k(y1 , y2 )k = ky1 k + ky2 k

82
Then
f (a + h) = (f 1 (a + h), f 2 (a + h))
0 0

= f 1 (a) + f 1 (a)h + φ1 (h), f 2 (a) + f 2 (a)h + φ2 (h)
 

0 0
= (f 1 (a), f 2 (a)) + f 1 (a)h, f 2 (a)h + φ1 (h), φ2 (h)
| {z } | {z }
linear in h remainder

Now:
k (φ1 (h), φ2 (h)) k kφ1 (h)k kφ2 (h)k
= +
khk khk khk
tends to 0 as h −→ 0. Therefore, f is differentiable at am and
 
0 0
f 0 (a)h = f 1 (a)h, f 2 (a)h

as required.

4.9 Notation
1. Given a function of n variables
f
Rn ⊃ V −→ R

V open, we shall denote by

x1 , x2 , . . . , x n

the usual co-ordinate functions on Rn and shall often denote the partial
derivative of f at a w.r.t. j th variable by:
∂f f (a1 ,a2 ,...,aj +t,...,an )−f (a1 ,a2 ,...,aj ,...,an )
∂xj
(a) = limt−→0 t

d

= dt
f (a + tej )t=0

Notice that the symbol xj does not appear in the definition: it is a

’dummy symbol’ indicating deriv w.r.t. j th ’slot’, sometimes written as
f,j (a).

2. given a function
f
R2 ⊃ V −→ R

83
 V open, we often denote by x, y the usual co-ordinate functions instead
of x1 , x2 and write
∂f ∂f
for
∂x ∂x1
∂f ∂f
for
∂y ∂x2
etc.

3. given
f
M ⊃ V −→ N
V open, if f is differentiable at a, ∀a ∈ V , we say that f is differentiable
on V and call the function on V :

f 0 : a −→ f 0 (a)

the derivative of f. We write:

f (2) = (f 0 )0

f (3) = ((f 0 )0 )0

and call f C r if f (r) exists and is continuous, f C ∞ if f (r) exists ∀r.

4.10 C r Functions
Theorem 4.10.1. Let
f
Rn ⊃ V −→ R
∂f
V open. Then f is C 1 ⇐⇒ ∂xi
exists and is continuous for i = 1, . . . , n.

Proof. (case n = 2)
f
R2 ⊃ V −→ R, f (x, y)
∂f ∂f
1. if f is C 1 , then ,
∂x ∂y
exist and
 
0 ∂f ∂f
f = ,
∂x ∂y
∂f ∂f
therefore, f 0 is continuous, and therefore, ∂x
and ∂y
are continuous

84
 ∂f ∂f
2. Let ,
∂x ∂y
exists and be continuous on V . Then

∂f ∂f
f (a + h) = f (a) + (a)h1 + (a)h2 + φ(h)
∂x ∂y

(say) where h = (h1 , h2 ). Now

φ(h) = f (a + h) − f (a + h2 e2 ) − ∂f
∂x
(a)h1
∂f
2
+f (a + h e2 ) − f (a) − ∂y (a)h2

∂f
= ∂x
(m1 )h1 − ∂f
∂x
(a)h1
+ ∂f
∂y
(m2 )h2 − ∂f
∂x
(a)h2

(say) by mean value theorem.

a + h 2 e2
m2 a+h

m1

PSfrag replacements a

Therefore,
   
kφ(h)k  ∂f ∂f  |h1 |  ∂f ∂f  |h2 |
≤  (m1 ) − (a) + (m2 ) − (a) −→ 0
khk ∂x ∂x  khk  ∂y ∂y  khk
∂f ∂f
as h −→ 0 since ,
∂x ∂y
are continuous at a.
 
∂f
Therefore, f is differentiable at a and f 0 (a) = ∂x
(a), ∂f
∂y
(a) . Also,
 
0 ∂f ∂f
f = ∂x , ∂y is continuous.

85
 ∂r f
Corollary 4.10.2. f is C r ⇐⇒ ∂xi1 ···∂xir
exists and is continuous for each
i1 , . . . , ir .

Theorem 4.10.3. if

f
Rn ⊃ V −→ R V open

∂2f
is C 2 then ∂xi ∂xj
is a symmetric matrix.

Proof. Let
f
Rn ⊃ V −→ R

be C 2 . Need to show:
∂2f ∂2f
=
∂x∂y ∂y∂x

b+h - +

b + -

a x a+h

Let (a, b) ∈ V . Let h 6= 0, k 6= 0 be such that the closed rectangle

(a, b), (a + h, b), (a + h, b + k), (a, b + k)

is contained in V . Put

g(x) = f (x, b + k) − f (x, b)

86
Then

f (a + h, b + k) − f (a + h, b)
−f (a, b + k) + f (a, b) = g(a + h) − g(a)

= hg 0 (c) some a ≤ c ≤ a + h by MVT

 ∂f ∂f

= h ∂x
(c, b + k) − ∂x
(c, b)

∂ f 2
= hk ∂y∂x (c, d) some b ≤ d ≤ b + k by MVT

and, similarly:

f (a + h, b + k) − f (a + h, b)
∂2f
−f (a, b + k) + f (a, b) = kh ∂x∂y (c0 , di0 )

for some a ≤ c0 ≤ a + h and b ≤ d0 ≤ b + k. Therefore

∂2f ∂2f 0 0
(c, d) = (c , d )
∂y∂x ∂x∂y

Now, let (h, k) −→ (0, 0). Then (c, d) −→ (a, b), and (c0 , d0 ) −→ (a, b).
Therefore,
∂2f ∂2f
(a, b) = (a, b)
∂y∂x ∂x∂y
∂2f ∂2f
by continuity of ∂y∂x
and ∂x∂y

4.11 Chain Rule

Theorem 4.11.1. (Chain Rule for functions on finite dimensional real or
complex vector spaces)
Let L, M, N be finite dimensional real or complex vector spaces. Let U be
open in L, V open in M and let
g f
U −→ V −→ N

Let g be differentiable at a ∈ U , f be differentiable at g(a). Then the compo-

sition f · g is differentiable at a and

(f · g)0 (a) = f 0 (g(a)) g 0(a) operator product

87
Proof. we have
f [g(ah )] = f [g(a) + T h + φ(h)] where T = g 0 (a)
and kφ(h)k
khk
−→ 0
as khk −→ 0

= f [g(a) + y] where y = T h + φ(h)

so kyk ≤ kT k khk + kφ(h)k −→ 0
as khk −→ 0

= f (g(a)) + Sy + kykψ(y) where S = f 0 (g(a))

and kψ(h)k −→ 0
as kyk −→ 0

= f (g(a) + S(T h + φ(h) + kykψ(y)

= f (g(a)) + ST h + Sφ(h) + kykψ(y)

Now
 
kSφ(h) + kykψ(y)k kφ(h)k kφ(h)k
≤ kSk + kT k + kψ(y)k
khk khk khk
tends to 0 as khk −→ 0. Therefore, f · g is differentiable at a, and
(f · g)0 (a) = ST = f 0 (g(a)) g 0 (a)

Example
d
f (g 1 (t), . . . , g n (t)) = (f · g)0 (t) = f 0 (g(t)) g 0 (t)
dt | {z } |{z}
1×n n×1

where
g f
R ⊃ U −→ V −→ R
V open in Rn , and f, g differentiable. Then
 d 1

dt
g (t)
f 0 (g(t))g 0(t) = ∂f ∂f
(g(t)), . . . ∂x

 .. 
∂x1 n (g(t))  . 
d n
dt
g (t)

∂f ∂f
= ∂x1
(g(t)) dtd g 1 (t) +···+ ∂xn
(g(t)) dtd g n (t)
is the usual chain rule.

88
Chapter 5

Calculus of Complex Numbers

5.1 Complex Differentiation

Definition let
f
C ⊃ V −→ C
be a complex valued function of a complex variable defined on open V . C is
a 1-dimensional complex normed space.
Let a ∈ V . Then f is differentiable at a as a function of a complex
variable if ∃ a linear map of complex spaces:
f 0 (a)
C −→ C

s.t.
f (a + h) = f (a) + f 0 (a)h + φ(h)
where |φ(h)|
|h|
−→ 0 as h −→ 0. f 0 (a) is a 1 × 1 complex matrix, i.e. a complex
number, and  
 f (a + h) − f (a) 0
 |φ(h)|
 − f (a) =
 h |h|
which tends to 0 as h −→ 0. Therefore
f (a + h) − f (a)
f 0 (a) = lim
h−→0 h
We call f 0 (a) the derivative of f at a. f 0 is called holomorphic on V if
0 df
f (a) exists ∀ a ∈ V . We write f = dy .

Example
f : C −→ C, f (z) = z n

89
then,
f (z+h)−f (z) (z+h)n −z n
limh−→0 h
= limh−→0 h

n(n−1) n−2
= limh−→0 [nz n−1 + 2
z h + higher powers h]

= nz n−1

therefore, f is holomorphic in C and f 0 (z) = nz n−1 .

Example
f (z) = z
then, 
f (z + h) − f (z) h 1 h real
= =
h h −1 h pure imaginary
which does not converge as h −→ 0.

The usual rules for differentiation apply:

d df dg
1. dy
(f + g) = dy
+ dy

d dg df
2. dy
fg = f dy + g dy
df
d f g dy −f dg
3. dy g
= g2
dy
if g 6= 0
d
4. dz
f (g(z)) = f 0 (g(z)) g 0 (z) chain rule

By definition C = R2

z = x + iy = (x, y)
√ i i
and the operation
  by i = −1
of mult C −→ C is the operator R2 −→ R2
0 −1
with matrix , because
1 0

ie1 = i = (0, 1) = 0e1 + e2

ie2 = ii = −1 = −e1 + 0e2 = (−1, 0)

Let f is (real) differentiable on V then each a ∈ v the operator
f 0 (a) f 0 (a)
C −→ C i.e R2 −→ R2

90
preserves addition and commutes with mult by real scalars, for each a ∈ V ,
and f is holomorphic in V iff f 0 (a) also commutes with multiplication by
complex scalars. ⇐⇒ f 0 (a) also commutes with multiplication by i.
 ∂u ∂u       ∂u ∂u 
∂x ∂y 0 −1 0 −1 ∂x ∂y
⇐⇒ ∂v ∂v = ∂v ∂v
∂x ∂y
1 0 1 0 ∂x ∂y
 ∂u   ∂v ∂v 
∂y
− ∂u
∂x
− ∂x − ∂y
⇐⇒ ∂v ∂v = ∂u ∂u
∂y
− ∂x ∂x ∂y
∂u ∂v 
∂x
= ∂y 
⇐⇒ Cauchy-Riemann Equations
∂v ∂u 
∂x
= − ∂y

so, therefore: f = u + iv is holomorphic in V ⇐⇒ f is real-differentiable

on V and satisfies the Cauchy-Riemann Equations.

Example f (z) = z 2 = (x + iy)2 = x2 − y 2 + 2ixy. then

∂u ∂v
u = x2 − y 2 ∂x
= 2x = ∂y

∂v
v = 2xy ∂x
= 2y = − ∂u
∂x

Note, if f = u + iv is holomorphic in V then

∂f f (a+h)−f (a)
∂y
(a) = limh−→0 h

f (a+t)−f (a) f (a+it)−f (a)

= limt−→0 t
= limt−→0 it

∂ 1 ∂
= ∂x
[u + iv] = i ∂y
[u + iv]

therefore,
∂f ∂u ∂v ∂v ∂u
= +i = −i
∂z ∂x ∂x ∂y ∂y
this also gives the Cauchy-Riemann Equations.

5.2 Path Integrals

Definition C 1 map
α
[t1 , t2 ] −→ V ⊂ R2 V open

t −→ α(t)

91
is called a (parametrical) path in R fom α(t1 ) to α(t2 ) and if f, g are complex-
valued continuous on V we write
Z Z t2
d d
(f dx + g dy) = [f (α(t)) x(α(t)) + g(α(t)) y(α(t))] dt
α t1 dt dt

and call it the integral of the differential form f dx + g dy over the path α. If
σ
[s1 , s2 ] −→ [t1 , t2 ]

is C 1 with σ(s1 ) = t1 , σ(s2 ) = t2 and β(s) = α(σ(s)) then β is called a

reparameterisation of α.

We have
R R s2 d d
β
(f dx + g dy) = s1
[f (β(s)) ds x(β(s)) + g(β(s)) ds y(β(s))] ds
R s2 d d
= s1
[f (α(σ(s))) ds x(α(σ(s))) + g(α(σ(s))) ds y(α(σ(s)))] ds
R s2
= s1
[f (α(σ(s)))(x · α)0 (σ(s)) + g(α(σ(s)))(y · α)0 (σ(s))]σ 0 (s) ds
R t2
= t1
[f (α(t)) dtd x(α(t)) + g(α(t)) dt
d
y(α(t))] dt
R
= α
(f dx + g dy)

therefore, the integral over α is independent of parmaterisation.

If we take
σ
[t1 , t2 ] −→ [t1 , t2 ]
with σ(s) = t1 + t2 − s, and β(s) = α(σ(s)) we have σ(t2 ) = t1 , σ(t1 ) = t2
then β is same path but traversed in the opposite direction and
R R t1
β
(f dx + g dy) = t2
[f (α(t)) dtd x(α(t)) + g(α(t)) dtd y(α(t))] ds
R
= − α
(f dx + g dy)

Definition If f is C 1 on V we write

∂f ∂f
df = dx + dy
∂x ∂y

A differential form of this type is called exact, and is called the differential
of f .

92
 If α is a path from a to b, α(t1 ) = a, α(t2 ) = b, then
R R t2
α
df = t1
[f (α(t)) dtd x(α(t)) + g(α(t)) dtd y(α(t))] dt
R t2
= [ d f (α(t))] dt
t1 dt

= f (α(t2 )) − f (α(t1 ))

= f (b) − f (a) so we have path independence

= change of value of f along α

If α is a closed path, (i.e. a = b) then
Z
df = 0
α

If f (z) = u(x, y) + iv(x, y), for z = x + iy with u, v real, put dz = dx + idy

and
f (z) dz = [u + iv][dx + idy]
= (u dx − v dy) + i(v dx + u dy)
then
R R R
α
f (z) dz = α (u dx − v dy) + i α (v dx + u dy)
R t2 
= t1
[u(α(t)) + iv(α(t))] dtd [x(α(t)) + iy(α(t))] dt
R t2
= t1
f (α(t)) α0 (t) dt
If f is holomorphic then
∂u ∂u ∂v
df = du + i dv = ∂x
dx + ∂y
dy + i ∂x dx + i ∂v
∂y
dy

∂u ∂v ∂v
= ∂x
dx − ∂x
dx + i ∂x dx + i ∂u
∂x
dy

∂u ∂v


= ∂x
+ i ∂x ( dx + i dy)

= f 0 (z) dz
therefore, if f is holomorphic on V open then
1. for any path α in V from a to b:
Z
f 0 (z) dz = f (b) − f (a)
α

path independence

93
 2. for any closed path α in V :
Z
f 0 (z) dz = 0
α

Example 1. if α path from a to b

Z Z
n d z n+1 bn+1 − an+1
z dz = [ ] dz = n 6= 1
α α dz n + 1 n+1

in particular,

2. if α is a closed path,
Z
z n dz = 0 n 6= 1
α

But:

3. for the closed path α(t) = eit , 0 ≤ t ≤ 2π

Z Z 2π Z 2π
dz 1 it
= ie dt = i dt = 2πi 6= 0
α z 0 eit 0

1
therefore, z
cannot be the derivative of a holomorphic function on C.

α
Definition if [t1 , t2 ] −→ C is a path in C then we write
Z t2
L(α) = |α0 (t)| dt
t1

and call it the length of α.

If β = α · σ is a reparameterisation of α with α0 (s) > 0
R s2
L(β) = s1
|β 0 (s)| ds
R s2
= s1
|α0 (σ(s))σ 0 (s)| ds
R t2
= t1
|α0 (t)| dt

= L(α)

94
Theorem 5.2.1. (Estimating a complex integral)
Let f be bounded on α:

|f (α(t))| ≤ M

(say), t1 ≤ t ≤ t2 . then
Z 
 
 f (z) dz  ≤ M L(α)
 
α

Proof. R 
R 
 f (z) dz  =  t2 f (α(t))α0 (t) dt
α t1

R t2
≤ t1
|f (α(t))| |α0(t)| dt
R t2
≤ M t1
|α0 (t)| dt

= M L(α)

5.3 Cauchy’s Theorem for a triangle

If f (z) = u(x, y) + i v(x, y) is holomorphic then
Z Z Z
f (z) dz = (u dx − v dy) + i (v dx + u dy)
α α

∂ ∂ ∂ ∂
and ∂y u = ∂x (−v), ∂y v = ∂x u (by Cauchy-Riemann).
Therefore, the neccessary conditions for path-independence are satisfied,
but not always sufficient, e.g.f (z) = z1 . However, we have

Theorem 5.3.1. (Cauchy’s Theorem for a triangle)

f
Let C ⊃ V −→ C be holomorphic on open V , and let T be a triangle
(interior plus boundary δT ) ⊂ V . Then
Z
f (z) dz = 0
δT

Proof. write Z
i(T ) = f (z) dz
δt

95
and join the mid-points of the sides to get 4 triangles

S1 , S2 , S3 , S4

then
4
X
i(T ) = i(Sj )
j=1

and therefore
4
X
|i(T )| ≤ |i(Sj )| ≤ 4|i(T1 )|
j=1

(say) where T1 is one of S1 , . . . , S4 .

T1

Repeat the process to get a sequence of triangles

T1 , T2 , . . . , T n , . . .

with
|i(T )| ≤ 4n |i(Tn )|
T∞
Let j=1 Tj = {c}. Then
R
i(Tn ) = δTn
f (z) dz
R
= δTn
[f (c) + f 0 (c)(z − c) + |z − c|φ(z − c)] dz
R
= δTn
|z − c| φ(z − c) dz where |φ(z − c)| −→ 0 as |z − c| −→ 0

96
 Tn
c
z

Let  > 0. Choose δ > 0 s.t. |φ(z − c)| <  ∀ |z − c| < δ. Let L =length of T .
Choose n s.t. length of Tn = ”ln < δ.
Then
l l l2
|i(Tn )| ≤ n  n = n 
2 2 4
therefore |i(T )| ≤ l2 , and therefore i(T ) = 0.

Definition A set V ⊂ C is called star-shaped if ∃a ∈ V s.t.

[a, z] ⊂ V ∀z ∈ V

f
Theorem 5.3.2. Let C ⊃ V −→ C be a holomorphic function on an open
star-shaped set. Then f = F 0 for some complex-differentable function F on
V and hence: Z
f (z) dz = 0
α

for each closed curve α in V .

Rw
Proof. Choose a ∈ V s.t. [a, z] ⊂ V ∀ z ∈ V . Put F (w) = a
f (z) dz. Then

97
 w+h

R w+h
F (w + h) = a
f (z) dz
Rw R w+h
= a
f (z) dz + w
f (z) dz by Cauchy
Z w+h
= F (w) + f (w)h + [f (z) − f (w)] dy
w
| {z }
φ(h)

Let  > 0. Choose δ > 0 w.t. |f (z) − f (w)| <  ∀|z − w| < δ. Therefore

|φ(h)| ≤ |h| 

for all |h| < δ, and therefore

|φ(h)|
lim =0
h−→0 |h|

Therefore, F is differentiable at w and F 0 (w) = f (w) as required.

5.4 Winding Number

We have seen that Z
dz
= 2πi
circle about o z
More generally:

98
 Theorem 5.4.1. Let a ∈ C ad α : [t1 , t2 ] −→ C − {a} be a closed path. The
Z
1 dz
2πi α z − a

is an integer, called the winding number of α about a

Proof. put Z t
α0 (s)
β(t) = ds
t1 α(s) − a
α0 (t)
so β 0 (t) = α(t)−a
. Then

d
 
dt
[α(t) − a]e−β(t) = [α0 (t) − β 0 (t){α(t) − a}] e−β(t)

= 0

therefore, [α(t)−a]e−β(t) is a constant function of t and is equal to [α(t1 )−a],

since β(t1 ) = 0. Therefore

α(t) − a
eβ(t) =
α(t1 ) − a

therefore,
α(t2 ) − a
eβ(t2 ) = =1
α(t1 ) − a
and therefore, β(t2 ) = 2nπi, with n an integer.

2nπi
a

eβ(t)
β(t)

frag replacements

99
 Therefore,
Z Z t2
dz α0 (s)
= ds = β(t2 ) = 2nπi
α z−a t1 α(s) − a

as required.
Theorem 5.4.2. Let C be a circle and a ∈ C. Then
1. if a is inside C, then C has winding number 1 about a:
Z
dz
= 2πi
C z −a

2. if a is outside C then winding number about a is 0.

Z
dz
=0
C z −a

Proof. 1. Let a be inside C, let C1 be a circle, centre a inside C. Let

α, β, γ be the closed paths shown, each is contained in an open star
1
shaped set on which z−a is holomorphic.

100
 C

C1
α
a

R dz
R dz
R dz
R dz
R dz
α z−a
+ β z−a
+ γ z−a
= C z−a
− C1 z−a

R R
0 + 0 = 0 = C
− C1

therefore,
Z Z Z 2π
dz dz ireiθ
= = dθ = 2πi
C z−a C1 z−a 0 reiθ

(put z = a + reiθ ).

2. Let a be outside C, then C is contained in an open star-shaped set on

1
which z−a is holomorphic. Therefore,
Z
dz
=0
C z −a

5.5 Cauchy’s Integral Formula

Theorem 5.5.1. (Cauchy’s Integral Formula)
Let f be holomorphic on open V in C. Let w ∈ V . Then, for any circle
C around w, such that C and it’s interior is contained in V , we have:
Z
1 f (z)
f (w) = dz
2πi C z − w
(Thus the values of f on any circle uniquely determine the values of f inside
the circle)

101
 V

C
w

Proof. Let  > 0. Choose a circle C1 centre w, radius r (say) inside C such
that |f (z) − f (w)| ≤  ∀z ∈ C1 by continuity of f .

C1
α
w

f (z)
Let α, β, γ be as indicated. Then, by integrating z−w :
Z Z Z Z Z
− = + + =0+0+0=0
C C1 α β γ

f (z)
since each of α, β, γ are contained in a star-shaped set on which z−w
is a
holomorphic function of z. Therefore,
R f (z) R f (z)
C z−w
dz = C1 z−w dz
R f (w) R f (z)−f (w)
= C1 z−w
dz + C1 z−w
dz

= 2πi f (w) + 0

because: Z 
 f (z) − f (w)  
 dz  ≤ 2πr = 2π

C1 z−w r
for all  > 0. Hence result.

102
 f
Let C ⊃ V −→ C be holomorphic on V open. Let w ∈ V . Pick a circle
C around w such that C and it’s interior ⊃ V . We have:
Z
1 f (z)
f (w) = dz
2πi C z − w
differentiating w.r.t. w under the integral sign:
Z
0 1 f (z)
f (w) = dz
2πi C (z − w)2

r z
w1

PSfrag replacements C

Justified since ∃ r > 0 s.t.

 
 f (z)  |f (z)|
 (z − w1 )2  ≤ r 2
 

for all w1 on an open set containing w1 , which is integrable w.r.t. z.

Repeating n times gives:
Z
(n) n! f (z)
f (w) = dz (5.1)
2πi C (z − w)n+1
Thus, f holomorphic on open V =⇒ f has derivatives of all orders and
(n)
f (w) is given by Equation(5.1) for any suitable circle C (or any suitable
closed curve) around w.

103
5.6 Term-by-term differentiation, analytic func-
tions, Taylor series
Theorem 5.6.1. Let {fn (z)} be a sequence of functions which is uniformly
convergent to the function f (z) on a path α. Then
Z Z
lim fn (z) dz = f (z) dz
n−→∞ α α

Proof. Let  > 0. Choose N s.t.

|fn (z) − f (z)| ≤  ∀n ≥ N ∀ z = α(t) t1 ≤ t ≤ t2
(definition of uniform convergence). Then
R R  R
 fn (z) dz − f (z) dz  ≤ |f (z) − f (z) dz|
α α α n

≤  l(α) ∀n≥N
hence the result.
TheoremP 5.6.2. (term by term differentiation)
Let ∞ i=1 fn (z) be a series of holomorphic functions on an open set V ,
which converges uniformly on a circle C which, together with it’s interior, is
contained in V .
Then the series converges inside C to a holomorphic function F (say)
and ∞
X
(k)
f = fn(k)
n=1
inside C
Proof. Let w be inside C, then λ = inf z∈C |z − w| > 0.
1 1
=⇒ ≤ k+1 ∀z ∈ C
|z − w| k+1 λ
P∞ fn (z) f (z)
Therefore, n=1 (z−w)k+1 converges uniformly to (z−w)k+1
z ∈ C. Therefore,
∞ Z Z
X fn (z) f (z)
dz = dz
n=1 C (z − w)k+1 C (z − w)k+1
and therefore,
∞
X
fn(k) (w) = f (n) (w)
n=1
as required.

104
 f
Definition Let C ⊃ V −→ C with V opn. Then f is called analytic on V
if, for each a ∈ V ∃ r > 0 and constants c0 , c1 , c2 , . . . ∈ C, such that:
f (z) = c0 + c1 (z − a) = c2 (z − a)2 + · · · ∀ z s.t. |z − a| < r
The R.H.S. is called the Taylor series of f about a.

Lemma 5.6.3. If 0 < ρ < r then the series converges uniformly on

{z : |z − a| ≤ ρ}

R
r
a

PSfrag replacements

P
Proof. LetP0 < ρ < R < r. cn (z − a)n converges for z − a = R, and
therefore, cn Rn converges. Therefore, ∃K s.t. |cn Rn | ≤ K ∀n. Therefore
K  ρ n
|cn (z − a)n | ≤ |cn | |z − a|n ≤ n ρn = K
R R
for all |z − a| ≤ ρ.
Therefore, we can differentiate term by term n times:
f (n) (z) = n!cn + (n + 1)!cn+1 |z − a| + · · · higher powers of z − a
1 (n)
Therefore, f (n) (a) = n!cn , and therefore, cn = n! f (a).
The Taylor coefficents are uniquely determined, f is C ∞ , and
∞
X 1 (n)
f (z) = f (z − a)n on |z − a| < r
n=1
n!

Thus, f analytic on V =⇒ f holomorphic on V .

Conversely,

105
 f
Theorem 5.6.4. Let C ⊃ V −→ C be holomorphic on open V . Let C be
any circle, centre a s.t. C and it’s interior are contained in V . Then f has
a Taylor series about a convergent inside C. Hence f is analytic on V .
Proof. Let w be inside C. Then
1
R f (z)
f (w) = 2πi C z−w
dz

1
R f (z)
= 2πi C (z−a)−(w−a)
dz

1
R f (z)
= 2πi C (z−a)[1− w−a ]
dz
z−a

1
R f (z) P∞

w−a n
= 2πi C z−a n=1 z−a
dz
Z
P∞ 1 n f (z)
= n=1 (w − a) dz
2πi C (z − a)n=1
| {z }
cn

P∞ 1 (n)
= n=0 n! f (c)(w − a)n

as required.

106
Chapter 6

Further Calculus

6.1 Mean Value Theorem for Vector-valued

functions
Theorem 6.1.1. (Mean value theorem for vector valued functions)
Let M, N be finite dimensional normed spaces and let
f
M ⊃ V −→ N

be a C 1 function, where V is open in M . Let x, y ∈ V and [x, y] ⊂ V . Then

kf (x − f (y)k ≤ k kx − yk

where k = supz∈[x,y] kf 0 (z)k

Proof. R1 d
f (y) − f (x) = 0 dt
f [ty + (1 − t)x] dt
R1
= 0
f 0 [ty + (1 − t)x] (y − x) dt
| {z } | {z }
operator vector

Therefore,
R1
kf (y) − f (x)k ≤ 0
kf 0 [ty + (1 − t)x]k ky − xk dt
R1
≤ 0
k ky − xk dt

= k ky − xk

as required.

107
 6.2 Contracting Map
Theorem 6.2.1. Let M be a finite dimensional real vector space with norm.
Let
f
M ⊃ V −→ M
be a C 1 function, V open in M , f (0) = 0, f 0 (0) = 1.
Let 0 <  < 1, and let B be a closed ball centre O s.t.
k1 − f 0 (x)k ≤  ∀x ∈ B
Then
1.
kf (x) − f (y)k ≥ (1 − ) kx − yk ∀ x, y ∈ B (6.1)
Thus fB is injective.
2. (1 − )B ⊂ f (B) ⊂ (1 + )B

f (B)

r f (1 − )r

PSfrag replacements B

(1 + )r

Proof. Let r =radius of B

1.
kf 0 (x)k = k1 + (f 0 (x) − 1)k ≤ 1 +  ∀x ∈ B
therefore,
MV T
kf (x)k = kf (x) − f (0)k ≤ (1 + )kxk ≤ (1 + )r
for all x ∈ B. Therefore, f (B) ⊂ (1 + )B

108
 2.
MV T
k(1 − f )(x) − (1 − f )(y)k ≤ kx − yk ∀ x, y ∈ B
therefore, kx − y| − kf (x) − f (y)k ≤  kx − yk, and so

kf (x) − f (y)k ≥ (1 − )kx − yk

and hence Eqn(6.1).

3. To show (1 − )B ⊂ f (B). Let a ∈ (1 − )B, define g(x) = x − f (x) + a.

Then
kg 0 (x)k = k1 − f 0 (x)k ≤  ∀x ∈ B
therefore,
MV T
kg(x) − g(y)k ≤ kx − yk
therefore, g is a contracting map (shortens distances).
Also,
kg(x)k = kg(x) − g(0) + ak since a = g(0)

≤ kg(x) − g(0)k + kak

≤ kxk + kak

≤ r + (1 − )r ∀x ∈ B

= r
therefore, g(x) ∈ B ∀ x ∈ B. So g maps B into B and is contracting.
Therefore, by the contraction mapping theorem ∃x ∈ B s.t. g(x) = x.
i.e. x − f (x) + a = x
i.e. f (x) = a. Therefore, a ∈ f (B), and (1 − )B ⊂ f (B) as required.

6.3 Inverse Function Theorem

f
Definition Let M ⊃ V −→ W ⊂ N where M, N are finite dimensional
normed spaces. Then f is called a C r diffeomorphism if
1. V open in M , W open in N
f
2. V −→ W is bijective

109
 3. f and f −1 are C r

V is C r diffeomorphic to W if ∃ a C r -diffeomorphism V −→ W .

Example
f
R −→ (0, ∞)

f (x) = ex is C ∞ , and f −1 (x) = ln x is C ∞ . Therefore, f is a C ∞ diffeomor-

phism.

y = ex

PSfrag replacements
x = ln y

Theorem 6.3.1. (Inverse Function Theorem)

f
Let M ⊃ V −→ N be C r where M, N are finite dimensional normed
spaces and V is open in M . Let a ∈ V be a point at which

f 0 (a) : M −→ N

is invertible. Then ∃ open neighbourhood W of a such that

fW : W −→ f (W )

is a C r diffeomorphism.

110
 N

f(W)

W
a
M

Proof. Let T be the inverse of f 0 (a) and let F be defined by

F (x) = T f (x + a) − T f (a)

We have:
F (0) = T f (a) − T f (a) = 0

U+a f(U+a)

a f

f(a)
T

+a

T f(a)
0
F 0 - T f(a)
U

F(U)

We prove that F maps an open neighbourhood U of 0 onto an open

neighbourhood F (U ) of 0. It the follows that f maps open U + a onto open

111
 f (U + a) by a C r diffeomorphism. Now

F 0 (x) = T · f 0 (x + a)

=⇒ F 0 (0) = T · f 0 (a) = 1M
Choose a closed ball B centre 0 of positive radius s.s.

1
kF 0 (x) − 1M k ≤ ∀x ∈ B
2

and also s.t. det F 0 (x) 6= 0. Then by the previous theorem (with  = 12 ) we
have:
FB is injective
and
1
kF (x) − F (y)k ≥ kx − yk ∀ x, y, ∈ B
2
and
1
B ⊂ F (B)
2

F F(B)

F −1
U 1
2
B
PSfrag replacements

therefore, F −1 : 21 B −→ B is well-defined and continuous. Let B 0 be the

interior of B, an open set. Put U = F −1 ( 21 B 0 ) ∩ B 0 , an open set. F (U ) is
open since F −1 is continuous. So FU : U −→ F (U ) is a homeomorphism of
open U onto open F (U ).
Let G be it’s inverse. To show G is C r .

112
 F

y+l x+h

G
y
x

F(U)

Let x, x + h ∈ F (U ), G(x) = y, G(x + h) = y + l (say), and khk ≥ 12 klk.

Let F 0 (y) = S. Then F (y + h) = F (y) + Sl + φ(l), where kφ(l)k
klk
−→ 0 as
klk −→ 0.
=⇒ x + h = x + Sl + φ(l),
=⇒ l = S −1 h − S −1 φ(l)
=⇒ G(x + h) = y + l = G(x) + S −1 h − S −1 φ(l)
Now,
kS −1 φ(l)k kφ(l)k klk
≤ kS −1 k −→ 0
khk klk khk
klk
as khk −→ 0 since klk ≤ 2khk =⇒ khk ≤ 2.
So, G is differentiable at x ∀x ∈ F (U ) and

G0 (x) = S −1 = [F 0 (y)]−1 = [F 0 (G(x))]−1

It follows that if G is C s for some 0 ≤ s < r then G0 is C s since G0 is a

composition of C s functions

F 0 , G, [·]−1

and therefore G is C s+1 . So

G C s =⇒ G C s+1 ∀0 ≤ s < r

therefore G is C r as required.

113
Example Let f, g be C r on an open set containing (a, b)

t
W
g(a,b)

b
W’
y

x a f(a,b) z

Let  
 ∂f ∂f 
∂(f, g)  ∂x ∂y 
=  6= 0
∂(x, y)  ∂g ∂g


∂x ∂y

at (a, b). Then (f, g) maps an open neighbourhood W of (a, b) onto an open
neighbourhood W 0 of (f (a, b), g(a, b) by a C r diffeomorphism. Therefore, for
each z, t ∈ W 0 ∃ unique (x, y) ∈ W such that

z = f (x, y)

t = g(x, y)
and x = h(z, t), y = k(z, t) where h and k are C r .

114
 Chapter 7

Coordinate systems and

Manifolds

7.1 Coordinate systems

Definition Let X be a topological space. Let V be open in X. A sequence
y = (y 1 , . . . , y n )
of real-valued functions on V is called an n-dimensional coordinate system
on X with domain V if
y
X ⊃ V −→ y(V ) ⊂ Rn
x :−→ y(x) = (y 1 (x), . . . , y n (x))
is a homeomorphism of V onto an open set y(V ) in Rn .

y(V)

y n (x) y(x)
x V

rag replacements
X
y 1 (x) Rn

115
 Example On the set V = {y 6= 0 or x > 0} in R2 the functions r, θ given
by
x = r cos θ

y = r sin θ

−π < θ < π

map V homomorphically onto an open set in R2 .

θ
(x,y)
π
r
(r, θ)
θ
PSfrag replacements v

θ
−π

Therefore, (r, θ) is a 2-dimensional coordinate system on R2 with domain

V.

Definition If y = (y 1 , . . . , y n ) is a coordinate system with domain V and if

f is a real-valued function on V then

f = F (y 1 , . . . , y n )

for a unique function F on y(V ) s.t. F = f · y −1

We call f a C r function of y 1 , . . . , y n if F is C r and we write

∂f ∂F 1
= (y , . . . , y n )
∂y i ∂xi

and call it the partial derivative w.r.t Y i in the coord system y 1 , . . . , y n .

116
 Note:
∂f ∂F
∂y i
(a) = ∂xi
(y 1 (a), . . . , y n (a))

d

= dt
F (y(a) + tei )t=0

d

= dt
f (αi (t))t=0

= rate of change of f along curve αi in V

y(a) + tei

αi (t)

a y(a)

rag replacements

where αi is the curve given by:

y(αi (t)) = y(a) + tei = (y 1 (a), . . . , y i (a) + t, . . . , y n (a))

therefore, along curve αi all coords y 1 , . . . , y n are constant except ith coord
y i and αi (t) is parameterised by change t in ith coord y i .
αi is called the ith coordinate curve at a.

7.2 C r -manifold
Definition Let y 1 , . . . , y n with domain V , and z 1 , . . . , z n with domain W
be two coordinate systems on X. Then these two systems are called C r -
compatible if each z i is a C r function of y 1 , . . . , y n, and each y i is a C r
function of z 1 , . . . , z n on V ∩ W .
We call X an n-dimensional C r -manifold if a collection of n-dimensional
coordinate systems is given whose domains cover X and which are C r -
compatible.

117
Example the 2-sphere S 2 given by:

x1 + y 1 + z 2 = 1 in R3

The function x, y with domain {z > 0} is a 2-dimensional coordinate system

on S 2 .
z

Similarly the functions x, z with domain {y > 0} is a 2-dimensional co-

ordinate system
On the overlap {y > 0, z > 0} we have:

x=x x=x
p p
y= 1 − x2 − y 2 z= 1 − x2 − y 2

these systems are C ∞ -compatible. In this way, we make S 2 into a C ∞ -

manifold.

7.3 Tangent vectors and differentials

Definition Let a ∈ X, X a manifold. Let y 1 , . . . , y n be co-ords on X at a
(i.e. domain an open neighbourhood of a). Then, the linear operators
 
∂  ∂ 
,··· , n (7.1)
∂y 1 a ∂y a

118
 act on the differentiable functions f at a (i.e. functions f which are real-
valued differentiable functions of y 1 , . . . , y n on an open neighbourhood of a)
and are defined by: 
∂  ∂f
j
f = j (a)
∂y a ∂y
The operators (7.1) are linearly independent since
   i
∂  y i = αj ∂y (a) = αj δ i = αi

αj j (7.2)
∂y j a ∂y j

j ∂ 

=⇒ α j
= 0 =⇒ αi = 0 ∀i (7.3)
∂y a
The real vector space with basis (7.1) is denoted Ta X and is called the tangent
space of X at a. If v ∈ Ta X then (7.2) shows that v has components vy i
w.r.t. basis (7.1)

Definition If α(t) is a curve in X then the velocity vector α̇(t) ∈ Tα(t) X is

the tangent vector at α(t) given by taking rate of change along α(t) w.r.t. t:

α̇(t)

α(t)

PSfrag replacements

i.e.
d
α̇(t)f = dt
f (α(t))

d
= dt
F (y 1 (α(t)), . . . , y n (α(t)) if f = F (y 1 , . . . , y n )

∂F
= ∂xj
(y 1 (α(t)), . . . , y n (α(t))) dtd y j (α(t))

∂f
= ∂y j
(α(t)) dtd y j (α(t))
  
d j 
= dt
y (α(t)) ∂y∂ j  f
α(t)

119
 ˙ is the tangent vector with components
therefore, α(t)
d j
y (α(t)) = ẏ j (t)
dt
i.e. 
j ∂ 
α̇(t) = ẏ (t) j 
∂y α(t)
The tangent space Ta X does not depend on the choice of (compatible) coor-
dinates at a since if z 1 , . . . , z n is another coordinate system at a, then
∂ 

∂z j a
= velocity vector of j th coordinate curve of z 1 , . . . , z n at a

∂y i ∂ 
= ∂z j ∂y i 
∈ Ta X
a

therefore,  
∂  ∂ 
,··· , n
∂z 1 a ∂z a
∂y i
is also a basis for Ta X, and ∂z j
(a) is the transition matrix from z 1 , . . . , z n to
y 1, . . . , y n.

Note also that the curve α(t) with coordinates:

y(α(t)) = (y 1 (a) + α1 t, . . . , y n(a) + αn t)

has velocity vector

 
∂ 
1 n ∂ 

α̇(0) = α + · · · α
∂y 1 a ∂y n a

Therefore, every tangent vector (at a) is the velocity vector of some curve,
and vice versa.

Definition Let a ∈ X and f be a differentiable function at a. Then for each

v ∈ Ta X with v = α̇(t) (say), define
d
hdfa , vi = vf = α̇(t)f = f (α(t)) = rate of change of f along v
dt
Thus dfa is a linear form on Ta X, called the differential of f at a dfa measures
the rate of change of f at a. If y 1 , . . . , y n are coordinates on X at a then
  
∂ ∂  i ∂y i
dyai , j =  y = (a) = δji
∂y a ∂y j a ∂y j

120
  
 ∂ 
therefore, dya1 , . . . , dyan is the basis of Ta X ∗ which is dual to the basis ∂y∂ 1  , · · · , ∂y n 
a a
of Ta X.
Thus the linear form dyai gives the ith component of tangent vectors at a:

hdyai , vi = ith component of v = vy i

Also, dfa has components:

 
∂ ∂ ∂f
dfa , j = j f = j (a)
∂y a ∂ a ∂y

therefore,
∂f
dfa = (a) dyaj
∂y j
(which is called the chain rule for differentials.)

7.4 Tensor Fields

From now on assume manifolds, functions are C ∞ .

Definition Let W be an open set in a manifold X. A tensor field S on X

with domain W is a function on W :

x 7−→ Sx

which assigns to each x ∈ W a tensor of fixed type over the tangent space
Ta X at x. e.g.
Sx : Tx X × (Tx X)∗ × Tx X −→ R

We can add tensor fields, contract them, form tensor products and wedge
products by carrying out these operations at each point x ∈ W . e.g.

(R + S)x = Rx + Sx

(R ⊗ S)x = Rx ⊗ Sx

Definition A tensor field with no indices is called a scalar field (i.e. a real-
valued function); a tensor field with one upper index is called a vector field ;
a tensor field with r skew-symmetric lower indices is called a differential r-
form; a tensor field with two lower indices is called a metric tensor if it is
symmetric and non-singular at each point.

121
 If y 1 , . . . , y n is a coordinate system with domain W and S is a tensor field
on W with (say) indices of type down-up-down, then
∂
S = αi j k dy i ⊗ ⊗ dy k
∂y k
where the scalar fields αi j k are the components of S w.r.t. the coordinates
y i.
If f is a scalar field, then df is a differential 1-form. If v is a vector field,
then vf is the scalar field defined by:

(vf )x = vx f = hdfx , vx i = hdf, vix

i.e. vf = hdf, vi = rate of change of f along v.

If (·|·) is a metric tensor on X then for any two vector fields u, v with
common domain W we define the scalar field (u|v) by

(u|v)x = (ux |vx )x

if v is a vector field then we can lower it’s index to get a differential 1-form
ω such that
hω, ui = (v|u)
Conversely, raising the index of a 1-form gives a vector field. Raising the
index of df gives a vector field gradf called the gradient of f :

(gradf |u) = hdf, ui = uf = rate of change of f along u

If (·|·) is positive definite then for kuk = 1 we have:

|(gradf |u)| ≤ kgradf k

Thus the maximum rate of change of f is kgradf k and is attained in the

direction of gradf .
A metric tensor (·|·) defines a field ds2 of quadratic forms called the
associated line element by:

ds2 (v) = (v|v) = kvk2 if metric is positive definite

If y i are coordinates with domain W then, on W :

each vector field u = αi ∂y∂ i with components αi
each differential r-form ω = ωi1 ...ir dy i1 ∧ · · · ∧ dy ir
each differential 1-form ω = ωi dy i

hω, ui = αi ωi

122
 ∂f
if f is a scalar field df = ∂y i
dy i

∂f
hdf, ui = uf = αi
∂y i
if   
∂  ∂
= gij
∂y i  ∂y j
then
(·|·) = gij dy i ⊗ dy j
and
ds2 = gij dy i dy j
gradf has components
∂f
g ij
∂y j
therefore,
∂f ∂
gradf = g ij
∂y j ∂y i
so
∂f ∂f
kgradf k = (gradf |gradf ) = hdf, gradf i = g ij
∂y i ∂y j
Example On R3 with the usual coordinate functions x, y, z the usual metric
tensor is
(·|·) = dx ⊗ dx + dy ⊗ dy + dz ⊗ dz
with components  
1 0 0
 
 
 = g ij
 0 1 0
gij =  
 
0 0 1
∂
, ∂, ∂
∂x ∂y ∂z
are orthonormal vector fields

∂f ∂f ∂f
df = dx + dy + dy
∂x ∂y ∂y
and
∂f ∂ ∂f ∂ ∂f ∂
gradf = + +
∂x ∂x ∂y ∂y ∂z ∂z
The line element is
ds2 = (dx)2 + (dy)2 + (dz)2

123
 If r, θ, φ are spherical polar cordinates on R3 then

x = r sin θ cos φ

y = r sin θ sin φ

z = r cos θ

therefore,

ds2 = (dx)2 + (dy)2 + (dz)2

= (sin θ cos φdr + r cos θ cos φdθ − r sin θ sin φdφ)2

+(sin θ sin φdr + r cos θ sin φdθ + r sin θ cos φdφ)2

+(cos θdr − r sin θdθ)2

= (dr)2 + r 2 (dθ)2 + r 2 sin2 θ(dφ)2

therefore,  
1 0 0
gij =  0 r 2 0 
0 0 r 2 sin2 θ
and  
1 0 0
ij 1
g =  0 r2
0 
1
0 0 r 2 sin2 θ
so
∂f ∂f ∂f
df = dr + dθ + dφ
∂r ∂θ ∂φ
and df has components:  
∂f ∂f ∂f
, ,
∂r ∂θ ∂φ
therefore, gradf has components
 
∂f 1 ∂f 1 ∂f
, ,
∂r r 2 ∂θ r 2 sin2 θ ∂φ
therefore,
∂f ∂ 1 ∂f ∂ 1 ∂f ∂
gradf = + 2 + 2 2
∂r ∂r r ∂θ ∂θ r sin θ ∂φ ∂φ

124
 7.5 Pull-back, Push-forward
Definition Let X and Y be manifolds and
φ
X −→ Y
a continuous map. Then for each scalar field f on Y we have a scalr field
φ∗ f = f · φ
on X (we assume that φ∗ f is C ∞ for each C ∞ f ).
φ∗ f is called the pull-back of f to X under φ.

φ∗ f f
φ φ(a)
a
rag replacements

X Y

(φ∗ f )(x) = f (φ(x))

For each a ∈ X we have a linear operator
φ∗
Ta X −→ Ta Y
If v = α̇(t) ∈ Ta X then we define φ∗ v by:
d
[φ∗ v]f = f (φ(α(t))) = α̇(t)[f · φ] = v[f · φ] = v[φ∗ f ]
dt
for each scalar field f on Y at φ(a).

rag replacements

φ φ∗ α̇(t)
a
α̇(t)
α(t) φ(a)φ(α(t))

125
 Thus the velocity vector of α(t) pushes forward to the velocity vector of
φ(α(t))
φ
Theorem 7.5.1. Let X −→ Y and let y 1 , . . . , y n be coordinates on X at a
and let
φ1 (x), . . . , φn (x)
be the coordinates of φ(x) w.r.t. a coordinate system z 1 , . . . , z n on Y at φ(a).
Then the push-forward
φ∗
Ta X −→ Tφ(a) Y
has matrix
∂φi
(x)
∂y j
Proof. h i
i th
component of φ∗ ∂y∂ j = φ∗ ∂y∂ j zi
a a

∂
= ∂y j a
[φ∗ z i ]

∂φi
= ∂y j
(a)
since φi (x) = z i (φ(x)), so φi = φ∗ z i .
Thus the matrix of φ∗ is the same as the matrix of the derivative in the
case Rn −→ Rn . The push-forward φ∗ is often called the derivative of φ at
a.
The chain rule for vector spaces
(ψ · φ)0 (x) = ψ 0 (φ(x)) φ0 (x)
corresponds, for manifolds, to:
Theorem 7.5.2. (Chain rule for maps of manifolds; functorial property of
the push-forward)
φ ψ
Let X −→ Y −→ Z be maps of manifolds. Then
(ψ · φ)∗ = ψ∗ · φ∗
Proof. Let α̇(t) be a tangent vector on X, then:
(ψ · φ)∗ α̇(t) = velocity vector of ψ(φ(α(t)))

= ψ∗ [velocity vector of φ(α(t))]

= ψ∗ [φ∗ α̇(t)]
as required.

126
 φ
Definition Let X −→ Y be a map of manifolds and ω a tensor field on Y
with r lower indices. Then we define the pull-back of ω to X under φ to be
the tensor field φ∗ ω on X having r lower indices and defined by:

(φ∗ ω)x [v1 , . . . , vr ] = ωφ(x) [φ∗ v1 , . . . , φ∗ vr ]

all v1 , . . . , vr ∈ Tx X, ∀x ∈ X. i.e.

(φ∗ ω)x = φ∗ [ωφ(x) ]

We note the φ∗ on tensor fields preserves all the algebraic operations such
as additions, tensor products, wedge products.
We have:
φ
Theorem 7.5.3. if X −→ Y and f is a scalar field on Y , then

φ∗ df = dφ∗ f

i.e., the pull-back commutes with differentials

Proof. Let v ∈ Tx X. Then
h(φ∗ df )x , vi = h(df )φ(x) , φ∗ vi

= [φ∗ v]f

= v[φ∗ f ]

= h(dφ∗ f )x , vi

therefore, φ∗ df )x = (dφ∗ f )x ∀x ∈ X, and therefore φ∗ df = dφ∗ f .

7.6 Implicit funciton theorem

Theorem 7.6.1 (Implicit Function Theorem). (on the solution spaces
of l equations in n variables)
Let f = (f 1 , . . . , f l ) be a sequence of C r real-valued functions on an open
set V in Rn . Let c = (c1 , . . . , cl ) ∈ Rl and let

X = {x ∈ V : f (x) = c, rankf 0 (x) = l}

Then for each a ∈ X we can select n − l of the usual coordinate functions

x1 , . . . , x n :
xl+1 , . . . , xn

127
 (say) so that on an open neighbourhood of a in X they form a coordinate
system on X. Any two such coordinate systems are C r -compatible. Thus X
is an (n − l)-dimensional C r manifold.
Proof.  
∂f 1 ∂f 1 ∂f 1
∂x1
··· ∂xl
··· ∂xn
f 0 =  ... .. ..
 
. . 
∂f l ∂f l ∂f l
∂x1
··· ∂xl
··· ∂xn l×n
0 0
Let a ∈ X. THen f (a) has rank l, so the matrix f (a) has l linearly inde-
pendent columns: the first l columns (say). Put

F = (f 1 , . . . , f l , xl + 1, . . . , xn )

then  
∂f 1 ∂f 1
∂x1
··· ∂xl
0 ···
 .. .. 

 . . 

∂f l ∂f l

 ∂x1
··· ∂xl
0 ··· 

0
F =




 0 ··· 0 1 0 

 .. .. .. 
 . . . 
0 ··· 0 0 1 l×n
therefore,  
∂f 1 ∂f 1

 ∂x1
(a) ··· 
∂xl
 (a)
det F 0 (a) = 
 .. .. 
. .  6= 0
 ∂f l ∂f l


∂x 1 (a) · · · ∂x l (a) 

PSfrag replacements
n−l
Rl R X Rn−l
Rn−l
W
F F (W )
U a
G

Rl c Rl

128
By the inverse function theorem, F maps an open neighbourhood W of a onto
an open neighbourhood F (W ) by a C r -diffeomorphism with inverse G (say).
Note that F , and hence also G leave the coordinates xl+1 , . . . , xn unchanged.
F maps W ∩ X homeomorphically onto {c} × U where U is open in Rn−l .
Thus (xl+1 , . . . , xn ) maps W ∩ X homeomorphically onto U and is there-
fore an (n − l)-dimensional coordinate system on X with domain W ∩ X.
Also, if
G = (G1 , . . . , Gl , xl+1 , . . . , xn )
on F (W ) then
x1 = G1 (c1 , . . . , cl , xl+1 , . . . , xn )
.. ..
. .
x = Gl (c1 , . . . , cl , xl+1 , . . . , xn )
l

on W ∩ X. Therefore x1 , . . . , xl and C r functions of xl+1 , . . . , xn on W ∩ X.

Hence any two such coordinate systems on X are C r compatible.

Note: the proof shows that if (say) the first l columns of the matrix
∂f i
∂xj
(a)
are linearly independent, then the l equations in n unknowns

f 1 (x1 , . . . , xn ) = c1
.. ..
. .
f l (x1 , . . . , xn ) = cl

determine x1 , . . . , xn as C r functions of xl+1 , . . . , xn on an open neighbour-

hood of a in the solution space.

Example 1.
f
Rn ⊃ V −→ R
a C r function.
   
0 ∂f ∂f ∂f
f = = 1
,..., n = ∇f
∂xi 1×n ∂x ∂x

are the components of the gradient vector field.

If ∇f is non-zero at each point of the space X of solutions of the
equations:
f (x1 , . . . , xn ) = c
(one equation in n unknowns) then X is an (n − 1)-dimensional C r -
manifold.

129
 ∂f
If (say) ∂x1
(a) 6= 0 at a ∈ X then

x2 , x3 , . . . , x n

are coordinates on an open neighbourhood W of a in X and x1 is a C r

function of x2 , . . . , xn on W .
f 1 ,f 2
2. Rn ⊃ V −→ R two C r functions.

  !
∂f 1 ∂f 1
∂f i ∂x1
··· ∂xn
= ∂f 2 ∂f 2
∂xj ∂x1
··· ∂xn

If the two rows ∇f 1 , ∇f 2 are linearly independent at each point of the

space X of solutions of the two equations:

f 1 (x1 , . . . , xn ) = c1

f 2 (x1 , . . . , xn ) = c2
(two equations in n unknowns) then X is an (n − 2)-dimensional C r
manifold. If (say)

∂(f 1 , f 2 )
6= 0 at a ∈ X
∂(x1 , x2 )

then x3 , x4 , . . . , xn are coordinates on an open neighbourhood W of a

in X and x1 and x2 are C r functions of x3 , . . . , xn on W .

7.7 Constraints
If f 1 , . . . , f l are C r functions on an open set V in Rn and c = (c1 , . . . , cn ) we
consider the equations
f 1 = c1 , . . . , f l = cl
as a system of constraints which are satisfied by points on the constraint
manifold :
 
1 1 l l ∂f i
X = x ∈ V : f (x) = c , . . . , f (x) = c , rank j (x) = l
∂x
i
Let X −→ Rn be the inclusion map i(x) = x ∀x ∈ X. Then for each scalar
field f on V :
(i∗ f )(x) = f (i(x)) = f (x) ∀x ∈ X

130
the pull-back i∗ f is the restriction of f to X. We call i∗ f the constrained
function F . If ω is a differential r-form on V the i∗ ω is called the constrained
r-form ω. Similarly if (·|·) is a metric tensor on V and ds2 the line element
the i∗ (·|·) is the constrained matrix and i∗ ds2 is the constrained line element.
For each a ∈ X the push-forward
i
∗
Ta X −→ T a Rn

is injective and enables us to identify Ta X with a vector subspace of Ta Rn .

The constrained metric and constrained line element are just the restrictions
to Ta X of the matrix and line element, for each a ∈ X.

Example Let X be the constraint surface in R3 :

f (x, y, z) = c f Cr

We have:
∂f ∂f ∂f
df = dx + dy + dz unconstrained
∂x ∂y ∂z
Pulling back to X, where f is constant, we get:
∂f ∂f ∂f
0= dx + dy + dz constrained
∂x ∂y ∂z

If (say) ∂f
∂z
6= 0 at a ∈ X then, by the implicit function theorem, the con-
strained functions x, y are coordinates on X in a neighbourhood W of a and
constrained z is a C r function of x, y on W .

z = F (x, y)

say. Now    
∂f . ∂f ∂f . ∂f
dz = − dx − dy
∂x ∂z ∂y ∂z
Therefore,  
∂z ∂F ∂f . ∂f
= =−
∂x
y ∂x ∂x ∂z
 
∂z ∂F ∂f . ∂f
= =−
∂y x ∂y ∂y ∂z
The usual line element on R3 is

ds2 = (dx)2 + (dy)2 + (dz)2 unconstrained

131
pulling back to X gives:
h  .   .  i2
ds2 = (dx)2 + (dy)2 + − ∂f
∂x
∂f
∂z
dx − ∂f
∂y
∂f
∂z
dy
  . 2  ∂f ∂f
  . 2 
∂f ∂f
= 1+ ∂x ∂z
2
(dx) + 2 ∂x ∂y
∂f 2
dx dy + 1 + ∂f
∂y
∂f
∂z
(dy)2
( )
∂y

The coefficients give the components of the constrained metric on X with

respect to the coordinates x and y

Example using spherical polar coordinates on R3 the usual line element is

ds2 = (dr)2 + r 2 (dθ)2 + r 2 sin2 θ(dφ)2

Pulling back to the sphere S 2 by the constraint:

r = const

we have:
ds2 = r 2 (dθ)2 + r 2 sin2 (dφ)2
Thus the constrained line element has components:
 2 
r 0
0 r 2 sin2 θ

7.8 Lagrange Multipliers

A scalar field f on a manifold X has a critical point a ∈ X if dfa = 0.
∂f i
i.e. ∂y i (a) = 0 for a coordinate system y at a. (e.g. a local maximum or

minimum or saddle point)

Problem given a scalar field F on V open in Rn , to find the critical point
of constrained F , where the constraints are:

f 1 = c1 , . . . , f l = cl

and f i are C r functions on V

Method Take f 1 , . . . , f l , xl+1 , . . . , xn (say) as coordinates on R3 as in the

proof of the implicit function theorem, so
l n
X ∂f i X ∂f i
dF = df + dx unconstrained
i=1
∂f i i=l+1
∂xi

132
 Thus n
X ∂F i
dF = 0 + dx constrained
i=l+1
∂xi
∂F
This is zero at a critical point a, so ∂x i (a) = 0, i = l + 1, . . . , n, and

hence
l
X ∂F
dFa = i
(a)df i
i=1
∂f
at a critical point a of constrained F . If we put
∂F
λi = (a) = rate of change of F at a with respect to the ith constraint
∂f i
we have:
dF = λ1 df 1 + · · · + λl df l
at a where the scalars λ1 , . . . , λl are called Lagrange multipliers.
Since dF is a linear combination of df 1 , . . . , df l at a we can take the
wedge product to get the equations:

dF ∧ df 1 ∧ · · · ∧ df l = 0

f 1 = c1 , . . . , f l = cl
which must hold at any critical point of constrained F .

7.9 Tangent space and normal space

If f is constant on the constraint manifold X then

df = 0 constrained

therefore,
hdf, α̇(t)i = 0
for all curves in X. Hence the system of l linear equations

df 1 = 0, . . . , df l = 0

give the tangent space to X at each point. Also if (·|·) is a metric tensor
with domain V then

(gradf |α̇(t)) = hdf, α̇(t)i = 0

133
for all curves in X. Therefore gradf is orthogonal to the tangent space to X
at each point. Hence
gradf 1 , . . . , gradf l
is a basis for the normed space to X at each point.
At a critical point of constrained F we have:
dF = λ1 df 1 + · · · + λl df l
and raising the index gives:
gradF = λ1 gradf 1 + · · · + λl gradf l
thus gradF is normal to the constraint manifold X at each critical point of
constrained F .

7.10 ?? Missing Page

1. =⇒ ω is closed.
However ω is not exact because it’s integral around circle α(t) =
(cos t, sin t) 0 ≤ t ≤ 2π is
Z Z 2π
2 π [cos t cos t − (sin t)(− sin t)]
ω=0 dt = dt = 2 π 6= 0
α cos2 t + sin2 t 0

Note: on R2 − {negative x-axis} we have:

x = r cos θ −π <θ <π
y = r sin θ
therefore,
r cos θ[sin θ dr + r cos θ dθ] − r sin θ[cos θ dr − r sin θ dθ]
ω= = dθ
r 2 cos2 θ + r 2 sin2 θ
R
so.
R α
ω is path-independent provided α does not cross the -ve x-axis.
α
ω = θ(b) − θ(a), and ω is called the angle-form
2. Let ω be a differential n-form with domain V open in Rn
ω = f (x1 , . . . , xn )dx1 ∧ · · · ∧ dxn (say) xi usual coord
then we define:
Z Z
ω= f (x1 , . . . , xn )dx1 dx2 . . . dxn Lebesgue integral xi dummy
V V

134
 7.11 Integral of Pull-back
φ
Theorem 7.11.1. Let V −→ φ(V ) be a C 1 diffeomorphism of open V in Rn
onto open φ(V ) in Rn , with det φ0 > 0. Let ω be an n-form on φ(V ). Then
Z Z
∗
φω= ω
V φ(V )

[so writing hω, V i to denote integral of ω over V we have: hφ∗ ω, V i = hω, φV i

adjoint]
Proof.

φ
ω
φ∗ ω
Sfrag replacements

V φV

Let
ω = f (x1 , . . . , xn )dx1 ∧ · · · ∧ dxn

φ = (φ1 , . . . , φn )
i.e. φi (x) = xi (φ(x)) i.e. φi = xi · φ = φ∗ xi .
Then
φ∗ ω = f (φ1 , . . . , φn )dφ1 ∧ · · · ∧ dφn

∂φ 1 ∂φn
i1
= f (φ(x)) ∂x i1 dx ∧ ···∧ ∂xin
dxin
 
∂φi
= f (φ(x)) det ∂xj
(x) dx1 ∧ · · · ∧ dxn
therefore,
Z Z Z
∗ 0
φω= f (φ(x)) det φ (x)dx1 dx2 . . . dxn = f (x) dx
V V φ(V )

by the general change of variable theorem for multiple integrals (still to be

proved).

135
 Corollary 7.11.2. Let y 1 , . . . , y n be positively oriented coordinates with do-
main V open in Rn . Then
Z Z
1 n 1 n
f (y , . . . , y )dy ∧ · · · ∧ dy = f (y1 , . . . , yn )dy1 dy2 . . . dyn
V y(V )
| {z } | {z }
non-dummy y i with wedge Lebesgue integral over y(V ). y1 , . . . , yn dummy. No wedge

y = (y 1 , . . . , y n )

y ∗ xi xi
PSfrag replacements

V y(V )

Proof.

y ∗ xi = x i · y = y i
therefore,
R
LHS = V
y ∗ [f (x1 , . . . , xn ) dx1 ∧ cdots ∧ dxn ]
R
= y(V )
f (x1 , . . . , xn ) dx1 ∧ · · · ∧ dxn
R
= y(V )
f (x1 , . . . , xn ) dx1 dx2 . . . dxn x1 , . . . , xn dummy

= RHS

7.12 integral of differential forms

R
To define X ω where ω is an n-form and X is an n-dimensional manifold
(e.g. ω is a 2-form, X a surface) we need some topological notions:

1. A topological space X is called Hausdorff if, for each a, b ∈ X, a 6= b, ∃

open disjoint V, W s.t. a ∈ V, b ∈ W .

136
PSfrag replacements W
V

a b

2. A set A ⊂ X is called closed in X if it’s complement in X

A0 = {x ∈ X : x 6∈ Z}

is open in X.

A0
PSfrag replacements A

3. If A ⊂ X then the intersection of all the closed subsets of X which

contain A is denoted Ā and is called the closure of A in X; Ā is the
smallest closed subset of X which contains A.

4. Let S be a tensor field on a manifold X. THe closure in X of the set

{x ∈ X : Sx 6= 0}

is called the support of S, denoted supp S.

supp S
PSfrag replacements S 6= 0

Theorem 7.12.1. Let X be a Hausdorff manifold and let A ⊂ X be compact.

Then ∃ scalar fields
F1 , . . . , Fk
on X s.t.

137
 1. Fi ≥ 0
2. F1 + · · · + Fk = 1 on A ( partition of unity)
3. each supp Fi is contained in the domain Vi of a coordinate system on
X.
Proof. (sketch) Let  
1
e− t t>0
g(t) =
0 t≤0

g(t)

PSfrag replacements

let
g[1 − t2 ]
b(t) =
g(1)

b(t)
PSfrag replacements
1
−1 1

b is C ∞ , supp b = [−1, 1], 0 ≤ b ≤ 1. (b is for bump).

Let a ∈ A. Pick a coordinate system y on X at a with domain V . Pick
a ball centre y(a) radius 2r > 0 contained in y(V ). Put
(   )
b ky(x)−y(a)k if x ∈ V
h(x) = r
0 if x 6∈ V

138
 then h is C ∞ , supp h ⊂ V , 0 ≤ h ≤ 1, h(a) = 1; ’bump’ at a.
Let W = {x ∈ X : h(x) > 0}. then W is an open neighbourhood of a
and h > 0 on W .
Thus, for each a ∈ A we have an open neighbourhood Wa of a and a
scalar field fa s.t. ha > 0 on Wa , ha is C ∞ , supp ha ⊂ a coord domain Va ,
0 ≤ ha ≤ 1.

Va

Wa
2r
a
rag replacements r
y(a)

Rn

ha is a bump at a, for each a ∈ A. Since A is compact we can select a

finite number of points a1 , . . . , ak such that Wa1 , . . . , Wak cover A. THen put

( )
ha i
ha1 +···+hak
6 0
ifhai =
Fi =
0 ifhai = 0

to get the required scalar fields F1 , . . . , Fk .

139
 h a2
h a1

PSfrag replacements F1 + F 2 = 1

F1 F2

7.13 orientation
Definition Let y 1 , . . . , y n with domain V , z 1 , . . . , z n with domain W be two
coordinate systems on a manifold X. Then they have the same orientation
if
∂(y 1 , . . . , y n) ∂y i
= det >0
∂(z 1 , . . . , z n ) ∂z j
on V ∩ W .
Since
∂ ∂y i ∂
=
∂z j ∂z j ∂y i
this means that ∂z∂ 1 a , · · · , ∂z∂n a has the same orientation in Ta X as ∂y∂ 1 , · · · , ∂y∂n
a a
each a ∈ V ∩ W .
We call X oriented if a family of mutually compatible coordinate systems
is given on X, whose domains cover X and any two of which have the same
orientation.

∂(x,y)
Example x = r cos θ, y = r sin θ. Then ∂(r,θ)
= r > 0. Therefore x, y and
r, θ have the same orientation.

140
 Definition Let ω be a differential n-form with compact support on an ori-
ented n-dimensional Hausdorff manifold X. We define
Z
ω
X

the integral of ω over X as follows:

1. Suppose supp ω ⊂ V where V is the domain of positively oriented

coordinates y 1 , . . . , y n and ω = f (y 1 , . . . , y n ) dy 1 ∧ · · ·∧ dy n on V . Then
define:
Z Z
ω= f (y1 , . . . , yn ) dy1 . . . dyn dummy yi
x y(V )

W
V

PSfrag replacements X

y z

y(V ) z(W )
φ

ω2
ω1

This is independent of the choice of coordinates since if supp ω ⊂ W

where W is the domain of positively oriented coordinates z 1 , . . . , z n
then

ω = f (y 1 , . . . , y n) dy 1 ∧ · · · ∧ dy n = g(z 1 , . . . , z n ) dz 1 ∧ · · · ∧ dz n (say) on V ∩ W

= y ∗ [f (x1 , . . . , xn ) dx1 ∧ · · · ∧ dxn = z ∗ [g(x1 , . . . , xn ) dx1 ∧ · · · ∧ dxn

= y ∗ ω1 = z ∗ ω2

141
 (say). So ω1 = φ∗ ω2 where φ = z · y −1 : y(V ∩ W ) −→ z(V ∩ W ).
Therefore,
R R R R
y(V )
f (x 1 , . . . , x n ) dx 1 . . . dx n = y(V )
ω 1 = y(V ∩W )
ω 1 = ω
z(V ∩W ) 2

R R
= ω =
z(W ) 2 z(W )
g(x1 , . . . , xn ) dx1 . . . dxn

as required.

2. Choose a partition of unity

F1 + · · · + F k = 1

on supp ω. Put ωi = Fi ω. Then

ω1 + · · · + ω k = ω

and supp ωi ⊂ supp Fi ⊂ Vi where Vi is the domain of a coordinates

system. Define Z Z Z
ω= ω1 + · · · + ωk
X X X

using (1.). This is independent of the choice of partition of unity since

if
G1 + · · · + G l = 1 on supp ω
then Pl R Pl R Pk
j=1 X
Gj ω = j=1 X i=1 Fi G j ω
Pl Pk R
= j=1 i=1 X
Fi G j ω

similarly Pk R
= i=1 X
Fi ω

Definition If A ⊂ X is a Borel set we define

Z Z
ω= χA ω
A X

Example to find the area of the surface X:

x2 + y 2 + z = 2, z>0

We have:
2x dx + 2y dy + dz = 0

142
constrained to X. Therefore,

ds2 = (dx)2 + (dy)2 + (2x dx + 2y dy)2 constrained to X

= (1 + 4x2 )(dx)2 + 8xy dx dy + (1 + 4y 2 )(dy)2

 
1 + 4x2 4xy
gij =
4xy 1 + 4y 2
√ p p p
therefore, g = det gij = 1 + 4x2 + 4y 2 . So, area element is 1 + 4x2 + 4y 2 dx∧
dy. Therefore: Rp
area = 1 + 4x2 + 4y 2 dx ∧ dy
R√
= 1 + 4r 2 r dr ∧ dθ
R 2π hR √2 √ i
= r 1 + 4r 2 dr dθ
0 0

h 3
i√2
2 1 2
= 2π 3 8
(1 + 4r ) 2
0

π
= 6
[27 − 1]

13
= 3
π

143
144
 Chapter 8

Complex Analysis

8.1 Laurent Expansion

Theorem 8.1.1. Let f be holomorphic on V − {a} where V is open and
a ∈ V . Let C be a circle, centre a, radius r s.t. C and it’s interior is
contained in V . Then ∃ {cn } n = 0, ±1, ±2, . . . ∈ C s.t.
∞
X
f (z) = cn (z − a)n in 0 < |z − a| < r
n=−∞

The RHS is called the Laurent series of f about a. The coefficient cn are
uniquely determined by:
Z
1 f (z)
cn = dz
2πi C (z − a)n+1

Proof. Let w be inside C. Choose circles C1 , C2 as shown with centres a, w:

w
PSfrag replacements C2

a
C1 C

145
 Then
1
R f (z)
f (w) = 2πi C2 z−w
dz

1
R f (z) 1
R f (z)
= 2πi C z−a
d− 2πi C1 z−w
dz

1
R f (z) 1
R f (z)
= 2πi C (z−a)−(w−a)
dz + 2πi C1 (w−a)−(z−a)
dz

1
R f (z) 1
R f (z)
= 2πi C (z−a)[1− w−a ]
dz + 2πi C1 (w−a)[1− z−a ]
dz
z−a w−a

1
R f (z) P∞

w−a n 1
R f (z) P∞

z−a n
= 2πi C (z−a) n=0 z−a
dz + 2πi C1 (w−a) n=0 w−a
dz
Z Z
P∞ 1 f (z) P −n−1 1 f (z)
= n=0 (w − a)
n
dz + ∞n=0 (w − a) dz
2πi C (z − a) n+1 2πi C! (z − a)−n
| {z } | {z }
+ve powers (w − a) -ve powers (w − a)

as required.

Definition If f is holomorphic in V − {a} anf

∞
X c−2 c−1
f (z) = cn (z − a)n = · · · + 2
+ +c0 + c1 (z − a) + · · ·
n=−∞
(z − a) z − a
| {z }
p(z)

P
is the Laurent series of f at a. p(z) = −1 n
n=−∞ cn (z−a) is called the principal
part of f at a. p(z) is holomorphic on C − {a}.
f (z) − p(z) is holomorphic in V (defining it’s value at 0 to be c0 ).
For any closed curve α in C − {a}:
R R h c−2 c−1
i
α
p(z) dz = α
···+ (z−a)2
+ z−a
dz
R dz
= · · · + 0 + c1 α z−a

= 2πi Res(f, a) W (α, a)

1
R dz
where W (α, a) = 2πi α z−a
is the winding number of α about a. and
Res(f, a) = c−1 is the residue of f at a

146
8.2 Residue Theorem
Theorem 8.2.1 (Residue Theorem). Let f be holomorphic on V −{a1 , . . . , an }
where a1 , . . . , an are distinct points in a star-shaped (or contractible) open set
V . Then for any closed curve α in V − {a1 , . . . , an } we have:
Z n
X
f (z) dz = 2πi Res(f, aj ) W (α, aj )
α j=1

Proof. Let p1 , . . . , pn be the principal parts of f at aR1 , . . . , an respectively.

Then f −(p1 +· · ·+pn ) is holomorphic on V . Therefore, α [f − (p1 + · · · + pn )] dz =
0. So,
Z n Z
X n
X
f (z) dz = pj (z) dz = 2πi Res(f, aj )W (α, aj )
j=1 α j=1

as required.

Definition If f is holomorphic on a neighbourhood of a, excluding possibly

a itself, and if the Laurent expansion has at most a finite number of negative
powers:
P∞ r
f (z) = r=n cr (z − a)

= cn (z − a)n + cn+1 (z − a)n+1 + · · · cn 6= 0

= (z − a)n [cn + cn+1 (z − a) + · · · ]

= = (z − a)n f1 (z) f1 holomorphic on a nbd of a and f1 (a) 6= 0

then we say that f has order n at a.

e.g. order 2:

f (z) = c2 (z − a)2 + c3 (z − a)3 + · · · c2 6= 0

order -3:
c−2 c−2 c1
f (z) = 3
+ 2
+ + c0 + c1 (z − a) + · · · c−3 6= 0
(z − a) (z − a) (z − a)

If n > 0 we call a a zero of f .

If n < 0 we call a a pole of f .
n = 1: simple zero; n = 2: double zero.

147
 n = −1: simple pole; n = −2: double pole.
If f has order n and g has order m:
f (z) = (z − a)n f1 (z) f1 (a) 6= 0

g(z) = (z − a)n g1 (z) g1 (a) 6= 0

then:
f (z) g(z) = (z − a)n+m f1 (z) g1 (z) f g has order n + m

f (z) (z)
g(z)
= (z − a)n−m gf11 (z) f
g
has order n − m
The residue theorem is very useful for evaluating integrals:
R∞
Example to evaluate 0 xx2sin
+a2
x
, a > 0 we put
z eiz
f (z) = z 2 +a2
[eiz is easier to handle than sin z]

z eiz
= (z−ia)(z+ia)

Simple poles at ia, −ia.

ia

−R R
PSfrag replacements

−ia

Choose a closed contour that goes along x-axis −R to R (say) then loops
around a pole, upper semi-circle α (say).
Z R Z
x eix z eiz
2 2
dx + 2 2
dz = 2πi Res(f, ia)
−R x + a α z +a

148
 c
1. if f (z) = z−ia + c0 + c1 (z − ia) + · · · on neighbourhood of ia then
(z − ia)f (z) = c + c0 (z − ia) + c1 (z − ia)2 + · · · on neighbourhood of
ia. Then limz−→ia (z − ia)f (z) = c = Res(f, ia). Therefore
f (z)
Res(f, ia) = c = limz−→ia z−ia

z eiz ia −a
= limz−→ia z+ia
= 2ia
e

1 −a
= 2
e

2. α(t) = Reit 0 ≤ t ≤ π = R(cos t + i sin t). Therefore

R  R 
 z eiz   π Reit eiR(cos t+i sin t) iReit 
 α z 2 +a2 dz  =  0 R2 e2it +a2
dt

R2
Rπ
≤ R2 −a2 0
e−R sin t dt

which −→ 0 as R −→ ∞ since
Z π Z π
−R sin t DCT
lim e dt = lim e−R sin t dt = 0
R→∞ 0 0 R→∞

Therefore Z ∞
x(cos x + i sin x) e−a
dx + 0 = 2πi
−∞ x2 + a 2 2
so Z ∞
x sin x
dx = πe−a
−∞ x2 + a 2
R∞ sin x eiz
Example to calculate −∞ x
dx put f (z) = z
holomorphic except for
simple pole at z = 0.

α
PSfrag replacements

−R −r r R

149
 Choose a closed contour along x-axis from −R to −r, loops around 0 by
β(t) = reit then r to R then back along α(t) = Reit .
Z −r Z Z R Z
eix eiz eix eiz
dx + dz + dx + dz = 0
−R x β z r x α z

1. Res(f, 0) = limz→0 (z − 0)f (z) = limz→0 eiz = 1

2. R  R 
 eiz   π eiR(cos t+i sin t) iReit 
 α z
dz  =  0 Reit
dt
Rπ
≤ 0
e−R sin t dt
which −→ 0 as R −→ ∞.
eiz 1
3. z
= z
+ g(z), g holomorphic in neighbourhood of 0. Therefore
Z Z Z
eiz dz
dz = + g(z) dz
β z β z β

Let sup |g(z)| = M (say on a closed ball centre 0 redius δ > 0 (say).
Z 
 
 g(z) dz  ≤ M πr −→ 0 as r −→ 0
 
β

if r ≤ δ Z Z Z
π π
dz ireit
=− dt = − i dt = −iπ ∀r
β z 0 reit 0

therefore Z
eiz
lim dz = −iπ
r−→0 β z
therefore Z Z
0 ∞
eix eix
dx − iπ + dx + 0 = 0
−∞ x 0 x
so Z ∞
cos x + i sin x
dx = iπ
−∞ x
and Z ∞
sin x
dx = π
−∞ x

150
8.3 Uniqueness of analytic continuation
Theorem 8.3.1 (Uniqueness of analytic continuation). Let f, g be holo-
morphic on a connected open set V . Let a ∈ V and let {zk } be a sequence in
V (6= a) converging to a s.t.

f (zk ) = g(zk ) ∀k

Then f = g on V .
Proof. Put F = f − g, so F (zk ) = 0. To show F = 0 on V .

1. F is holomorphic at a. Therefore F (z) = b0 +b1 (z −a)+b2 (z −a)2 +· · ·

on |z − a| < R (say). F (a) = lim F (zk ) = 0. Therefore b0 = 0.
Suppose we know that b0 , b1 , . . . , bm−1 are all zero.

F (z) = (z − a)m [bm + bm+1 (z − a) + bm+2 (z − a)2 + · · · ]

= (z − a)m Fm (z)

(say). F (zk ) = 0, so Fm (zk ) = 0 and Fm (a) = 0. Therefore bm = 0.

br = 0∀r and F = 0 on an open ball centre a.

2. Put V = W ∪W 0 where W = {z ∈ V : F = 0 on an open ball centre z}

and W 0 = V − W . Then W is open, and a ∈ W by 1. Therefore W
is non-empty. Suppose c ∈ W 0 and c is a non-interior point of W 0 .
Then each integer r > 0∃ wr ∈ W s.t. |wr − c| < 1r , F (wr ) = 0 and
lim wr = c. Therefore F = 0 on an open ball centre c by 1., and c ∈ W .
Therefore W 0 is empty.

151
152
Chapter 9

General Change of Variable in

a multiple integral

9.1 Preliminary result

φ
Theorem 9.1.1. Let Rn ⊃ V −→ Rn be C 1 , V open, a ∈ V , det φ0 (a) 6= 0.
Then
m(φB)
lim = | det φ0 (a)|
m(B)
where the limit is taken over cubes B containing a with radius B −→ 0.
Proof. Let k · k be the sup norm on Rn .

k(α1 , . . . , αn )k = max(|α1 |, . . . , |αn |)

so a ball, radius r is a cube, with side 2r.

Let 0 <  < 1. Put T (x) = [φ0 (x)]−1 . Fix a closed cube J containing
a and put k = supx∈J kT (x)k. Choose δ > 0 s.t. kφ0 (x) − φ0 (y)k ≤ k all
kx − yk ≤ 2δ; x, y ∈ J.
If B is a cube ⊂ J containing a of radius ≤ δ. and centre c (say).
Consider:
T (c)φ has derivative T (c)φ0 (x) which equals 1 at x = c and

kT (c)φ0 (x) − 1k = kT (c)φ0 (x) − T (c)φ0 (c)k

≤ kT (ck kφ0 (x) − φ0 (c)k


≤ k k

= 

153
therefore (1 − )B1 ⊂ T (c)φB ⊂ (1 + )B1 , where B1 is a translate of B to
new centre T (c)φ(x). Therefore

(1 − )n m(B) ≤ | det T (c)|m(φB) ≤ (1 + )n m(B)

m(φB)
=⇒ (1 − )n ≤ | det T (c)| ≤ (1 + )n
m(B)
m(φB)
=⇒ lim | det T (c)| =1
m(B)
m(φB)
=⇒ | det T (a)| lim =1
m(B)
Therefore:
m(φB) 1
lim = = | det φ0 (a)|
m(B) | det T (a)|
as required

Theorem 9.1.2. Let f be a continuous real valued function on an open

neighbourhood of a in Rn . THen
R
f (x) dx
lim B = f (a)
m(B)

where the limit is taken over cubes B containing a with radius B −→ 0.

Proof. Let  > 0. Choose δ > 0 s.t. |f (x) − f (a)| <  ∀ kx − ak < δ. Then
each cube B containing a of radius ≤ 2δ we have:
R  R 
 B f (x) dx  B
[f (x) − f (a)] dx  m(B)

 m(B) − f (a) = ≤ =
 m(B) m(B)

hence result.

Recall that the σ-algebra generated by the topology of Rn is called the

collection of Borel Sets in Rn .

Theorem 9.1.3. Let A be a Borel set in Rr , B be a Borel set in Rs . Then

A × B is a Borel set in Rr+s .

Proof. For fixed V open in Rr , the sets:

{V × W : W open in Rs } (9.1)

154
are all open in Rr+s . Therefore the σ-algebra generated by Eqn(9.1):

{V × B : B Borel in Rs }

consists of Borel sets in Rr+s . Hence for fixed B Borel in Rs , the set:

{V × B : V open in Rr (9.2)

are all Borel sets in Rr+s . Therefore the σ-algebra generated by Eqn(9.2):

{A × B : A Borel in Rr }

consists of Borel sets in Rr+s . Therefore A × B is Borel for each A Borel in

Rr , B Borel in Rs .
Theorem 9.1.4. Let E be a lebesgue measurable subset of Rn . THen there
is a Borel set B containing E such that

B − E = B ∩ E0

has measure zero, and hence m(B) = m(E).

Proof. We already know this is true for n = 1. So we use induction on n.
Assume true for n − 1. Let E ⊂ Rn , E measurable.
1. Let m(E) < ∞. Let k be an integer > 0.

E ⊂ Rn−1 × R

choose a sequence of rectangles

A1 × B − 1, A2 × B2 , . . .

covering E with Ai ⊂ Rn−1 measurable and Bi ⊂ R measurable, and

such that ∞
X 1
m(E) ≤ m(Ai )m(Bi ) ≤ m(E) +
i=1
k
By the induction hypothesis choose Borel sets Ci , Di s.t.

Ai ⊂ Ci , Bi ⊂ Di , m(Ai ) = m(Ci ), m(Bi ) = m(Di )

S
Put Bk =P ∞ i=1 Ci × Di . Then E ⊂ Bk , Bk is Borel and m(E) ≤
∞
m(Bk ) ≤ i=1 m(Ci )m(di ) ≤ m(E) + k1 .
T
Put B = ∞ k=1 Bk . Then E ⊂ B, B is Borel and m(E) = m(B).
Therefore m(B ∩ E 0 ) = m(B) − m(E) = 0 as required.

155
 S
2. Let m(E) = ∞. Put Ek = {x ∈ E : k ≤ |x| < k + 1}. E = ∞ k=0 Ek is
a countable disjoint union, and m(Ek ) < ∞. For each integer k choose
by 1. Borel Bk s.t. Ek ⊂ Bk and m(Bk ∩ Ek0 ) = 0
S
Put B = Bk . Then E ⊂ B and m(B ∩ E 0 ) = m(B) − m(E) as
required. ???????????

9.2 General change of variable in a multiple

integral
Theorem 9.2.1 (General change of variable in a multiple integral).
φ
Let Rn ⊃ V −→ W ⊂ Rn be a C 1 diffeomorphism of open V onto open W .
Let f be integrable on E. Then
Z Z
f (x) dx = f (φ(x))| det φ0 (x)| dx (Lebesgue Integrals) (9.3)
W V

Proof. 1. we have f = f + − f − with f + , f − ≥ 0. Therefore, it is sufficient

to prove Eqn(9.3) for f ≥ 0.
2. if f ≥ 0 then ∃ a monotone increasing sequence of non-negative simple
functions {fn } such that f = lim fn so, using the monotone convergence
theorem, it is sufficient to prove Eqn(9.3) for f simple.
Pk
3. if f is simple then f = i=1 ai χEi with {Ei } Lebesgue measurable,
so it is sufficient to prove Eqn(9.3) with f = χE with E Lebesgue
measurable.
4. if E Lebesgue measurable ∃ Borel B s.t. E ⊂ B and Z = B − E has
measure zero. Therefore, χE = χB − χZ , and it is sufficient to prove
Eqn(9.3) for f = χB , B Borel, and for f = χZ , Z measure zero.
5. If f = χE with E Borel then E = φF with F Borel and Eqn(9.3)
reduces to Z Z
χφF (x) dx = χF (x)| det φ0 (x)| dx
W V
i.e. Z
m(φF ) = | det φ0 (x)| dx (9.4)
F

If Eqn(9.4) holds for each rectangle F ⊂ V then Eqn(9.4) holds for F ∈

ring R of finite disjoint unions of rectangles ⊂ V .

156
 Therefore Eqn(9.4) holds for F ∈ monotone class generated by R.
Therefore Eqn(9.4) holds for F ∈ σ-algebra generated by R (monotone
class lemma). So Eqn(9.4) holds for any Borel set F ⊂ V

6. to show Eqn(9.4) holds for any rectangle F ⊂ V : for each rectangle

F ⊂ V put Z
λ(F ) = m(φF ) − | det φ0 (x)| dx
F
Then, λ is additive: [ X
λ( Bj ) = λ(Bj )
for a disjoint union.
Must prove λ(F ) = 0 for each rectangle F . Suppose B is a rectangle
for which λ(B) 6= 0.
Suppose B is a cube. |λ(B)| > 0. Therefore, ∃  > 0 s.t. |λ(B)| ≥
 m(B).
Divide B into disjoint subcubes, each of 21 the edge of B. For one such,
B1 say,
|almbda(B1 )| ≥  m(B1 )
Sivide again
|λ(B2 )| ≥  m(B2 )
continuing we get a decreasing sequence of cubes {Bk } converging to a
say, with
|λ(Bk )|
lim ≥>0
h−→∞ m(Bk )

But R
λ(Bk ) m(φBk )− Bk | det φ0 (x)| dx
limh→∞ m(B k)
= limh→∞ m(Bk )

= | det φ0 (a)| − | det φ0 (a)|

= 0
a contradiction.
R Therefore λ(B) = 0 for all cubes B. Therefore,
m(φB) = B | det φ0 (x)| dx for all cubes B as required.

7. Now suppose Z has measure zero, and choose a Borel set B s.t.

Z⊂B⊂V

and s.t. B has measure zero.

157
 Then Z
m(φZ) ≤ m(φB) = | det φ0 (x)| dx = 0
B
since B has measure zero. Therefore φZ has measure zero.
Therefore under a C 1 diffeomorphism we have:

Z of measure zero =⇒ φZ of measure zero

therefore f = χZ with Z of measure zero =⇒ f · · · φ = χφ−1 Z with

φ−1 Z of measure zero.
Therefore, Eqn(9.3) holds for f = χZ since then both sides are zero.
This completes the proof.

158