0% found this document useful (0 votes)
50 views21 pages

Measure Theory

1) The document discusses measurable functions on a measurable space (X,A). It proves that a function f: A → [-∞,∞] is measurable if and only if the sets {x ∈ A: f(x) ≤ t} are in A for every real number t. 2) It also proves various properties of measurable functions, including that continuous functions are Borel measurable, characteristic functions are measurable if and only if the set is measurable, and operations like maximum, minimum, addition and multiplication of measurable functions are also measurable. 3) It defines simple functions as those that take only finitely many values and proves they are measurable if and only if the preimage of each

Uploaded by

Upasana Devi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
50 views21 pages

Measure Theory

1) The document discusses measurable functions on a measurable space (X,A). It proves that a function f: A → [-∞,∞] is measurable if and only if the sets {x ∈ A: f(x) ≤ t} are in A for every real number t. 2) It also proves various properties of measurable functions, including that continuous functions are Borel measurable, characteristic functions are measurable if and only if the set is measurable, and operations like maximum, minimum, addition and multiplication of measurable functions are also measurable. 3) It defines simple functions as those that take only finitely many values and proves they are measurable if and only if the preimage of each

Uploaded by

Upasana Devi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 21

Functions and Integrals

January 2019

1 Measurable functions
Proposition 1.1. Let(X, A ) be a measurable space and let A be a subset of X that
belongs to A .Let f : A → [−∞, ∞],the following conditions are equivalent:

(a) For each real number t, the set {x ∈ A : f (x) ≤ t} belongs to A .

(b) For each real number t, the set {x ∈ A : f (x) < t} belongs to A .

(c) For each real number t, the set {x ∈ A : f (x) ≥ t} belongs to A .

(d) For each real number t, the set {x ∈ A : f (x) > t} belongs to A .

Proof. (a)⇒ (b):


{x ∈ A : f (x) < t} = ∪n {x ∈ A : f (x) ≤ t − n1 }
Since for each real number t, the set {x ∈ A : f (x) ≤ t} belongs to A
[i.e. the set {x ∈ A : f (x) ≤ any real number} belongs to A ]
,therefore for all value of t − n1 ,{x ∈ A : f (x) ≤ t − n1 } belongs to A
Therefore, ∪n {x ∈ A : f (x) ≤ t − n1 } ∈ A
i.e.{x ∈ A : f (x) < t} ∈ A
(b)⇒ (c):
If for each real number t, the set {x ∈ A : f (x) < t} belongs to A ,
therefore, for each real number t, the set {x ∈ A : f (x) < t}0 belongs to A [as A is a σ−
algebra].Also,A ∈ A ,so,A ∩ {x ∈ A : f (x) < t}0 ∈ A i.e. {x ∈ A : f (x) ≥ t} is measurable.
(c)⇒ (d):
Again, {x ∈ A : f (x) > t} = ∪n {x ∈ A : f (x) ≥ t + n1 } ∈ A [∵ {x ∈ A : f (x) ≥ t} ∈ A ]
(d)⇒ (a):
{x ∈ A : f (x) ≥ t} ∈ A , ∴ A − {x ∈ A : f (x) ≥ t} ∈ A i.e. {x ∈ A : f (x) < t} ∈ A
Hence all the conditions are equivalent.

1.1
Chapter 2 Functions and integrals

Definition 1.1. Let (X, A ) be a measurable space and let A ∈ A .A function f : A → [−∞, ∞]
is measurable with respect to A if it satisfies any of the conditions from the above proposition.

Note:

1. A function that is measurable with respect to A is called A −measurable.If the


σ−algebra is clear from the context,then we shall simply say ”measurable”.

2. A function which is measurable with respect to B(Rd ) is called Borel measurable


function or simply Borel function.

3. A function which is measurable with respect to Mλ∗ is called Lebesgue measur-


able function.

4. Of course every Borel measurable function on Rd is Lebesgue measurable, be-


cause if f is a Borel function,then {x ∈ Rd : f (x) < t} ∈ B(Rd ) ⊆ Mλ∗ .

Problem: Prove that every continuous real valued function on Rd is Borel


measurable and hence Lebesgue measurable.
Soln:
Let f : Rd → R be continuous.We consider an arbitrary real number t and consider
the set {x ∈ Rd : f (x) < t}.Then {x ∈ Rd : f (x) < t} = f −1 (−∞, t),(−∞, t) being
an open interval is an open set and f is continuous.As we know that for a continuous
map,inverse image of every open set is open,so,f −1 (−∞, t),(−∞, t) is open i.e.a Borel
set.Thus {x ∈ Rd : f (x) < t} ∈ B(Rd ) for every t. Hence f is Borel function.Also f
is Lebesgue measurable function as every Borel function is also Lebesgue measurable
function.

1.2
Chapter 2 Functions and integrals

Problem:
Let (X, A ) be a measurable space, and let B be a subset of X,then χB ,the character-
istic function of B, is A −measurable iff B ∈ A
Soln:
χB : X → R such that
χB (x) = 1, if x ∈ B

= 0, if x ∈
/B

We consider the set {x ∈ X : χB (x) < t}, where t is an arbitrary real number.
If t < 0,then {x ∈ X : χB (x) < t} = φ, an A − measurable set.
If 0 < t < 1,then {x ∈ X : χB (x) < t} = X − B,which is an A − measurable set if B
is A − measurable set.
If t > 1,then {x ∈ X : χB (x) < t} = X,which is an A − measurable set. Thus
if B is A − measurable set, then in any case χB ,the characteristic function of B, is
A −measurable
Conversely if χB is A − measurable, then for each t,{x ∈ X : χB (x) < t}
is A −measurable.In particular taking t = 0.5(say){x ∈ X : χB (x) < 0.5} =
X − B which is A −measurable. Thus X − B is A −measurable also, X is
A −measurable.Therefore B = X − (X − B) is A −measurable.

Problem:
A function is called simple if it has only finitely many values.
Let (X, A ) be a measurable space, let f : X → [−∞, ∞] be simple, and let
α1 , α2 , ..., αn be the values of f, then f is measurable iff {x ∈ X : f (x) = αi } ∈ A
holds for i = 1, 2, ..., n

Proposition 1.2. Let (X, A ) be a measurable space,let A be a subset of X that belong


to A , and let f,g be [−∞, ∞]- valued measurable function on A,then the sets

{x ∈ A : f (x) < g(x)},

{x ∈ A : f (x) ≤ g(x)},

{x ∈ A : f (x) = g(x)},

belong to A .

Proof.
[
{x ∈ A : f (x) < g(x)} = [{x ∈ A : f (x) < r} ∩ {x ∈ A : r < g(x)}]
r∈Q

Since f(x),g(x) are measurable , for each value of r,

1.3
Chapter 2 Functions and integrals

{x ∈ A : f (x) < r}, {x ∈ A : r < g(x)} are measurable.


∴ {x ∈ A : f (x) < r} ∩ {x ∈ A : r < g(x)}
[
∴ {x ∈ A : f (x) < g(x)} = [{x ∈ A : f (x) < r} ∩ {x ∈ A : r < g(x)}]
r∈Q

is measurable.[Since A is a σ− algebra]
Similarly, {x ∈ A : g(x) < f (x)} ∈ A Again,

A, {x ∈ A : g(x) < f (x)} ∈ A

∴ A − {x ∈ A : g(x) < f (x)} ∈ A

i.e.
{x ∈ A : f (x) ≤ g(x)} ∈ A

Again

{x ∈ A : f (x) = g(x)} = {x ∈ A : f (x) ≤ g(x)} − {x ∈ A : f (x) < g(x)} ∈ A

Definition 1.2. Let f and g be two [−∞, ∞] valued functions having common domain
A.The maximumand minimum of f and g, written f ∨ g and f ∧ g are the functions
from A to [−∞, ∞] defined by

(f ∨ g)(x) = max{f (x), g(x)}

and
(f ∧ g)(x) = min{f (x), g(x)}

Proposition 1.3. Let (X, A ) be a measurable space.Let A be a subset that belongs to


A .Let f and g be two [−∞, ∞] valued measurable functions on A.Then f ∨ g,f ∧ g are
measurable.

Proof. Let us consider the set {x ∈ A : (f ∨ g)(x) ≤ t}, for each t.


i.e.
{x ∈ A : max{f (x), g(x)} ≤ t}

i.e.
{x ∈ A : f (x) ≤ t} ∩ {x ∈ A : g(x) ≤ t} ∈ A

1.4
Chapter 2 Functions and integrals

[Since f,g are measurable, so, {x ∈ A : f (x) ≤ t} ∈ A and {x ∈ A : g(x) ≤ t} ∈ A ]

Hence f ∨ g is measurable.
Similarly, let us consider the set {x ∈ A : (f ∧ g)(x) ≤ t}, for each t.
i.e.
{x ∈ A : min{f (x), g(x)} ≤ t}

i.e.
{x ∈ A : f (x) ≤ t} ∪ {x ∈ A : g(x) ≤ t} ∈ A

[Since f,g are measurable, so, {x ∈ A : f (x) ≤ t} ∈ A and {x ∈ A : g(x) ≤ t} ∈ A ]

Hence f ∧ g is measurable.

Proposition 1.4. Let (X, A ) be a measurable space.Let A be a subset of X that belongs


to A and let {fn } be a sequence of [−∞, ∞] valued measurable functions on A.Then

(a) the functions supn fn and inf n fn are measurable.

(b) the functions lim supn fn and lim inf n fn are measurable.

(c) the functions limn fn (whose domain is {x ∈ A : lim supn fn = lim inf n fn ) is measur-
able.

Proposition 1.5. Let (X, A ) be a measurable space.Let A be a subset of X that belongs


to A .Let f and g be two [0, ∞] valued measurable functions on A and let α be a non
negative real number.Then αf and f + g is measurable.

Proof. If α = 0, then the set

{x ∈ A : (αf )(x) < t} = {x ∈ A : (0)(x) < t}

= φ, if t ≤ 0

= A, if t > 0

which is measurable in both the cases. For α > 0,

t
{x ∈ A : (αf )(x) < t} = {x ∈ A : f (x) < }
α

1.5
Chapter 2 Functions and integrals

which belongs to A
Let us consider the set {x ∈ A : (f + g)(x) < t}. If f (x) < r, where r is a rational number, then
g(x) < t − r.Thus, {x ∈ A : (f + g)(x) < t}
[
= [{x ∈ A : f (x) < r} ∩ {x ∈ A : g(x) < t − r}]
r∈Q

and so {x ∈ A : (f + g)(x) < t} is the union of countable collection of sets that belong to A ,itself
belongs to A

Proposition 1.6. Let (X, A ) be a measurable space.Let A be a subset of X that belongs


to A ,let f and g be measurable real valued functions on A, and let α be a real
number.Then αf, f + g, f − g, f g, fg (Domain of f /g is {x ∈ A : g(x) 6= 0}) are measurable

Proof. If α = 0, then the set

{x ∈ A : (αf )(x) < t} = {x ∈ A : (0)(x) < t}

= φ, if t ≤ 0

= A, if t > 0

which is measurable in both the cases. For α > 0,

t
{x ∈ A : (αf )(x) < t} = {x ∈ A : f (x) < }
α
which belongs to A
For α < 0,

t
{x ∈ A : (αf )(x) < t} = {x ∈ A : f (x) > }
α
which belongs to A
Let us consider the set {x ∈ A : (f + g)(x) < t}. If f (x) < r, where r is a rational number, then
g(x) < t − r.Thus, {x ∈ A : (f + g)(x) < t}
[
= [{x ∈ A : f (x) < r} ∩ {x ∈ A : g(x) < t − r}]
r∈Q

and so {x ∈ A : (f + g)(x) < t} is the union of countable collection of sets that belong to A ,itself
belongs to A
Since g is measurable, so −g is measurable.[αf ’s measurability.]
So, f + (−g) is measurable, i.e. f − g is measurable.
Let h : A → R be measurable.We shall show that h2 is measurable.

1.6
Chapter 2 Functions and integrals

Let us consider the set {x ∈ A : h2 (x) < t}, which is equal to φ if t ≤ 0 and so, measurable.
For t > 0
p p
{x ∈ A : h2 (x) < t} = {x ∈ A : − (t) < h(x) < (t)}
p \ p
= {x ∈ A : − (t) < h(x)} {x ∈ A : h(x) < (t)} (1.1)

Since, h is measurable, so,{x ∈ A : − (t) < h(x)}, {x ∈ A : h(x) < (t)} ∈ A


p p

hence by (1.1),{x ∈ A : h2 (x) < t} is measurable.


Thus for any real value of t,{x ∈ A : h2 (x) < t} is measurable.
Hence, we have obtained, if h is a measurable function, h2 is also measurable.
Thus f,g being measurable, implies f 2 , g 2 , (f + g)2 are measurable[f,g,measurable,implies f + g
is measurable]
∴ f g = 21 ((f + g)2 − f 2 − g 2 ) is measurable.
Next, we shall show that fg is measurable.[Domain of fg is {x ∈ A : g(x) 6= 0}]
Let
A0 = {x ∈ A : g(x) 6= 0}

= {x ∈ A : g(x) > 0} ∪ {x ∈ A : g(x) < 0}

Since g is measurable, so,{x ∈ A : g(x) > 0}, {x ∈ A : g(x) < 0} ∈ A , therefore A0 , being the
union of these two measurable sets is measurable,i.e. A0 ∈ A
Now,
f
{x ∈ A0 : (x) < t}
g
f (x)
= {x ∈ A0 : < t}
g(x)
[
= {x ∈ A : f (x) < t g(x) if g(x) > 0} {x ∈ A : f (x) > t g(x) if g(x) < 0}
[
= ({x ∈ A : f (x) < t g(x)} ∩ {x ∈ A : g(x) > 0}) ({x ∈ A : f (x) > t g(x)} ∩ {x ∈ A : g(x) < 0})
(1.2)
We know that {x ∈ A : g(x) > 0}, {x ∈ A : g(x) < 0} ∈ A
Since g is measurable, so, tg is measurable,so,{x ∈ A : f (x) < t g(x)}, {x ∈ A : f (x) > t g(x)} ∈
A . [Using the theorem {x ∈ A : f (x) < g(x)} is measurable,when f,g are measurable]
∴(1.2) is measurable, i.e.{x ∈ A0 : fg (x) < t} ∈ A , for every value of t, so, fg is measurable.

Proposition 1.7. Let (X, A ) be a measurable space, let A be a subset of X that belongs to A , and
let f be a [0, ∞]-valued measurable function on A.Then there is a sequence {fn } of [0, ∞)-valued
simple measurable functions on A that satisfy

f1 (x) ≤ f2 (x) ≤ ... (1.3)

1.7
Chapter 2 Functions and integrals

and
f (x) = lim fn (x) (1.4)
n

at each x in A.

Proof. For each positive integer n and for k = 1, 2, ..., n2n let

k−1 k
An,k = {x ∈ A : n
≤ f (x) < n }
2 2

The measurability of f implies that each An,k ∈ A. We define a sequence {fn } of functions from A
to R as
k−1
fn (x) = for each point x in An,k
2n
= n for each point x in A − ∪k An,k

We prove(3)
Let us consider fn , fn+1
k−1 k
An,k = {x ∈ A : ≤ f (x) < }
2n 2n
k−1 k−1 1 k−1 1 k−1 1 1
= {x ∈ A : n
≤ f (x) < n
+ n+1 } ∪ {x ∈ A : n
+ n+1 ≤ f (x) < n
+ n+1 + n+1 }
2 2 2 2 2 2 2 2
2(k − 1) 2k − 1 2k − 1 k−1 1
= {x ∈ A : ≤ f (x) < } ∪ {x ∈ A : ≤ f (x) < + }
2n+1 2n+1 2n+1 2n 2n
= An+1,2k−1 ∪ An+1,2k

Let x ∈ A
then x ∈ An,k for exactly one value of k,k = 1, 2, 3, ..., n2n , then x ∈ An+1,2k−2 or x ∈ An+1,2k−1
∴ fn (x) = k−1
2n
2(k − 1)
∴ fn+1 (x) =
2n+1
, or,
2k − 1
2n+1
∴ fn+1 (x) ≥ fn (x)
If x ∈ A − ∪k An,k , then fn (x) = n
Since x ∈ A − ∪k An,k , therefore,
f (x) ≥ n

, therefore,
n2n+1
f (x) ≥
2n+1
n2n+1
∴ fn+1 (x) ≥ =n
2n+1

1.8
Chapter 2 Functions and integrals

Thus, in any case


fn+1 (x) ≥ fn (x)

Now we shall prove that


lim fn (x) = f (x)
n

Let f (x) = ∞, now x ∈ A − ∪k An,k for any finite value of n. Therefore fn (x) = n, therefore,
limn fn (x) = ∞ . If f (x) < ∞, then for some value of m, we can have x ∈ Am,k . Hence for
n ≥ m, x ∈ An,k for some value of k.Therefore,
x ∈ An,k implies k−1
2n
≤ f (x) < 2kn .Also, fn (x) = k−1
2n
k−1 k k−1 1
Therefore, 0 ≤ f (x) − 2n < 2n − 2n = 2n
i.e.
1
0 ≤ f (x) − fn (x) < n
2
∴ lim fn (x) = f (x)
n

1.9
Chapter 2 Functions and integrals

Note: For n = 1,possible values of k are k = 1, 2,therefore the sets are A1,1 , A1,2 For
n = 2,possible values of k are k = 1, 2, 3, 4, 5, 6, 7, 8,therefore the sets are

A2,1 , A2,2 , A2,3 , A2,4 , A2,4 , A2,5 , A2,6 , A2,7 , A2,8

f(A11 )=[0,1/2),f(A12 )=[1/2,1)


f(A21 )=[0,1/4),f(A22 )=[1/4,1/2),f(A23 )=[1/2,3/4),f(A24 )=[3/4,1),f(A25 )=[1,5/4),
f(A26 )=[5/4,3/2),f(A27 )=[3/2,7/4),f(A28 )=[7/4,1)

1.10
Chapter 2 Functions and integrals

Note: Define f + (x) = max{f (x), 0} = f ∨ 0


f − (x) = − min{f (x), 0} = max{−f (x), 0} = (−f ) ∨ 0
Clearly, f = f + − f − and |f | = f + + f −
f is measurable implies |f | is measurable.
This is because, f is measurable implies f ∨ 0, (−f ) ∨ 0 are measurable. i.e. f + , f − are
measurable.Hence |f | = f + + f − is measurable.
So, if f is [−∞, ∞]valued measurable function,then f + , f − are [0, ∞] valued measurable
function,so, by the above theorem there exist sequence of simple functions{fn }, {gn } each
of them are monotonic decreasing, as shown in the theorem,such that limn fn = f + and
limn gn = f − , therefore we can construct {fn − gn } sequence of simple functions (may not
be monotonic increasing now), such that limn (fn − gn ) = f + − f − = f

Proposition 1.8. Let (X, A ) be a measurable space, and let A be a subset of X that belongs to
A . For a function f : A → R, the following conditions are equivalent.

(a) f is measurable with respect to A

(b) For each open subset U of R, the set f −1 (U ) ∈ A

(c) For each closed subset C of R, the set f −1 (C) ∈ A

(d) For each Borel subset B of R, the set f −1 (B) ∈ A

Proof. Let F = {B ⊆ R : f −1 B ∈ A }. Now f −1 (R) = A ∈ A , so, R ∈ F


f −1 (φ) = φ ∈ A , so, φ ∈ F
If B ∈ F , then f −1 B ∈ A
therefore f −1 B c = A − f −1 B ∈ A , therefore, f −1 B c ∈∈ A
therefore, B c ∈ F .
If {Bn } ∈ F , then f −1 (∪n Bn ) = ∪n f −1 (Bn ) ∈ A
therefore ∪n Bn ∈ F . Therefore F is a σ-algebra on R.
(a) → (d)
Let f be measurable w.r.t A

∴ {x ∈ A : f (x) ≤ t} ∈ A

for every real number t.


∴ (−∞, t] ∈ F .F being a σ−algebra, the σ-algebra generated by (−∞, t]i.e. B(R) ⊆ F . Therefore
(d).
(d) → (a)
∴ f −1 (B) ∈ A for every B ∈ B(R)
In particular f −1 (−∞, t] ∈ A ∀t ∈ R

1.11
Chapter 2 Functions and integrals

{x ∈ A : f (x) ≤ t} ∈ A ∀t ∈ R
therefore, f is A −measurable.
(d) → (b)
Let f −1 (B) ∈ A for every B ∈ B(R)
In particular,f −1 (U ) ∈ A for every U ∈ B(R),where U is open set.
(b) → (d)
Let f −1 (U ) ∈ A for every B ∈ B(R)
In particular f −1 (−∞, t) ∈ A as (−∞, t)is open and hence belongs to B(R),therefore f is A -
measurable
therefore (d)[as (a)implies (d)].
(d) → (c)
Let f −1 (B) ∈ A for every B ∈ B(R)
In particular,f −1 (C) ∈ A for every C ∈ B(R),where C is closed set.
(c) → (d)
Let f −1 (C) ∈ A for every C ∈ B(R),where C is closed set.
Therefore, C ∈ F , F being a σ−algebra, the σ-algebra generated by Ci.e. B(R) ⊆ F .
∴ (d)

2 The Integral
Let (X, A ) be a measurable space. Let J be the collection of all real valued simple A measurable
functions on X, and let J+ be the collection of non negative functions in J .If f ∈ J+ and is
given by f = m
P
i=1 ai χAi ,where a1 , a2 , ..., am are non negative real numbers and A1 , A2 , ..., Am are
disjoint subsets of X that belongs to A ,then
Z m
X
f dµ = ai µ(Ai ).
i=1

Let f be also given by nj=1 bj χBj , where B1 , B2 , ..., Bn are disjoint subsets of X that belong to
P

A . We can assume that ∪m n


i=1 Ai = ∪j=1 Bj (If necessary eliminate the sets Ai for which ai = 0
and the sets Bj for which bj = 0).Clearly ai = bj if Ai ∩ Bj 6= φ implies that m
P
i=1 ai µ(Ai ) =
Pm Pn Pn
ai µ(Ai ∩ Bj ) = j=1 bj µ(Bj )
Ri=1 j=1
∴ f dµ does not depend on the representation of f.

Proposition 2.1. Let (X, A , µ) be a measure space, let f and g belong to J+ and let α be a
non-negative real number.Then
R R
a) αf dµ = α f dµ
R R R
b) (f + g)dµ = f dµ + gdµ

2.1
Chapter 2 Functions and integrals

R R
c) if f (x) ≤ g(x) holds at each x ∈ X, then f dµ ≤ gdµ

Pm
Proof. Suppose that f = i=1 ai χAi , where a1 , a2 , ..., am are non negative real numbers and
A1 , A2 , ..., Am are disjoint subsets of X that belong to A and that g = nj=1 bj χBj ,where b1 , b2 , ..., bn
P

are non negative real numbers and B1 , B2 , ..., Bn are disjoint subsets of X that belong to A ,we can
assume that m
S Sn
i=1 Ai = j=1 Bj .Then
(αf )dµ = i=1 αai µ(Ai ) = α m
R Pm P R
i=1 a i µ(Ai ) = α f dµ
Z m X
X n
(f + g)dµ = (ai + bj )µ(Ai ∩ Bj )
i=1 j=1
Z m X
X n m X
X n
(f + g)dµ = ai µ(Ai ∩ Bj ) + bj µ(Ai ∩ Bj )
i=1 j=1 i=1 j=1

m X
X n m X
X n
= ai µ(Ai ) + bj µ(Bj )
i=1 j=1 i=1 j=1
Z Z
= f dµ + gdµ

We suppose that f (x) ≤ g(x) holds at each x ∈ X,then g − f ∈ J+ ,so,


Z Z Z Z Z
gdµ = {f + (g − f )}dµ = f dµ + (g − f )dµ ≥ f dµ

Proposition 2.2. Let (X, A , µ) be a measure space. let f belongs to J+ and let {fn } be a non
decreasing sequence of functions in J+ for which f (x) = limn fn (x) holds at each x in X. Then
Z Z
f dµ = lim fn dµ
n

Proof. since
f1 ≤ f2 ≤ f3 ≤ ... ≤ f

therefore Z Z Z Z
f1 dµ ≤ f2 dµ ≤ f3 dµ ≤ ... ≤ f dµ
R R R
Hence, limn fn dµ exists and satisfies limn fn dµ ≤ f dµ
R R
Now we shall prove that limn fn dµ ≥ f dµ.
Let  be a positive number such that 0 <  < 1.
suppose that a1 , a2 , ..., ak are the non zero values of f and that A1 , A2 , ..., Ak are the sets on which

2.2
Chapter 2 Functions and integrals

these values occur. Thus


k
X
f= ai χAi
i=1

For each n and i, let


A(n, i) = {x ∈ Ai : fn (x) ≥ (1 − )ai }

then each A(n, i) ∈ A and for each i the sequence {A(n, i)}∞ n=1 is non decreasing and satisfies
Ai = ∪n A(n, i).
we assume
Xk
gn = (1 − )ai χA(n,i)
i=1

Then gn ∈ J+ and gn ≤ fn

Z k
X
∴ gn dµ = (1 − )ai µ(A(n, i))
i=1

Z k
X
∴ lim gn dµ = lim (1 − )ai µ(A(n, i))
n n
i=1

k
X
= (1 − )ai lim µ(A(n, i))
n
i=1

k
!
X [
= (1 − )ai µ(A(n, i))
i=1 n

k
X
= (1 − )ai µ(Ai )
i=1
Z
= (1 − ) f dµ

Since gn ≤ fn Z Z
⇒ gn dµ ≤ fn dµ
Z Z
⇒ lim gn dµ ≤ lim fn dµ
n n
Z Z
⇒ (1 − )f dµ ≤ fn dµ
Z Z
⇒ f dµ ≤ lim fn dµ
n

[Since  is arbitrary]
Hence comparing the two inequalities we have the result.

2.3
Chapter 2 Functions and integrals

Definition 2.1. We define the integral of an arbitrary [0, ∞] valued A -measurable function f on
X. Z Z
f dµ = sup{ gdµ : g ∈ J+ , g ≤ f }

Proposition 2.3. Let (X, A , µ) be a measure space, let f be [0, ∞] valued A -measurable function
on X, and let {fn } be non decreasing sequence of functions in J+ for which limn fn (x) = f (x)
holds at each x in X.Then Z Z
f dµ = lim fn dµ
n

Proof. Since f1 ≤ f2 ≤ ... ≤ fn ≤ ..... which is convergent [as limn fn (x) = f (x)].Therefore the
monotonic increasing sequence must converge to its least upper bound i.e.
f1 ≤ f2 ≤ ... ≤ fn ≤ ..... ≤ f .therefore by the previous theorem
Z Z Z Z
f1 dµ ≤ f2 dµ ≤ ... ≤ fn dµ ≤ ..... ≤ f dµ

R
Hence limn fn dµ exists and satisfies
Z Z
lim fn dµ ≤ f dµ (2.1)
n

We turn to the reverse inequality.


Let g ∈ J+ and g ≤ f . Then g ∧ fn is a non decreasing sequence of functions in J+ for which
R R
g = limn (g ∧ fn ).Therefore, previous theorem suggests that gdµ = limn (g ∧ fn )dµ
Since g ∧ fn ≤ fn ,so Z Z
(g ∧ fn )dµ ≤ fn dµ

so Z Z
lim (g ∧ fn )dµ ≤ lim fn dµ
n n

i.e. Z Z
gdµ ≤ lim fn dµ
n
Z Z
∴ sup gdµ ≤ lim fn dµ
g≤f n
Z Z
∴ f dµ ≤ lim fn dµ (2.2)
n

Therefore, from (2.1) and (2.2) we have


Z Z
f dµ = lim fn dµ
n

2.4
Chapter 2 Functions and integrals

Proposition 2.4. Let (X, A , µ) be a measure space, let f and g be [0, ∞]-valued A -measurable
functions on X, and let α be a non-negative real number.Then
R R
(a) αf dµ = α f dµ
R R R
(b) (f + g)dµ = f dµ + gdµ
R R
(c) If f (x) ≤ g(x) holds at each x in X, then f dµ ≤ gdµ

Proof. We know that if f is a [0, ∞] valued measurable functions on X, then there is a sequence
{fn } of increasing [0, ∞) valued simple measurable functions such that f (x) = limn fn (x).Therefore
we can choose non decreasing sequence {fn } and {gn } of functions in J+ such that f (x) =
limn fn (x),g(x) = limn gn (x).Also,{αfn } and {fn + gn } are non decreasing sequences of functions
in J+ that satisfies αf (x) = limn αfn (x) and lim(fn + gn ) = limn fn + limn gn = f + g
Therefore, Z Z
αf dµ = lim αfn dµ
n
Z Z Z
= lim α fn dµ = α lim fn dµ = α f dµ
n n

and Z Z
(f + g)dµ = lim (fn + gn )dµ
n
Z Z Z Z
= lim fn dµ + lim gn dµ = f dµ + gdµ
n n

Let h ∈ J+ s.t. h ≤ f This implies h ≤ g[Since f ≤ g] Therefore,


Z Z
{ hdµ : h ∈ J+ andh ≤ f } ⊆ { h0 dµ : h ∈ J+ andh0 ≤ g}

Therefore Z Z
sup{ hdµ : h ∈ J+ andh ≤ f } ≤ sup{ h0 dµ : h ∈ J+ andh0 ≤ g}
Z Z
∴ f dµ ≤ gdµ

Definition 2.2. Let f be an arbitrary [−∞, ∞] valued A -measurable functions on X.If


R
f + dµ
and f − dµ are both finite, then f is called integrable and
R

Z Z Z
f dµ = +
f dµ − f − dµ

f − dµ is finite.
R R
The integral of f is said to exist if at least one of f + dµ and

2.5
Chapter 2 Functions and integrals

Let f : X → [−∞, ∞] is A -measurable and that A ∈ A . Then f is integrable over A if the


function f χA is integrable, the integral of over A is defined as the integral (if it exists) of the
function on X that agrees with f on A and vanishes on Ac
Lemma 2.1. Let (X, A , µ) be a measure space, and let f1 , f2 , g1 , g2 be non negative real valued
integrable functions on X such that f1 − f2 = g1 − g2 .Then
Z Z Z Z
f1 dµ − f2 dµ = g1 dµ − g2 dµ

Proposition 2.5. Let (X, A , µ) be a measure space.Let f and g be real valued integrable functions
on X, and let α be a real number .Then
(a) αf and f + g are integrable.
R R
(b) αf dµ = α f dµ
R R R
(c) (f + g)dµ = f dµ + gdµ
R R
(d) If f (x) ≤ g(x) holds at each x in X, then f dµ ≤ gdµ
R R
Proof. For α = 0 clearly αf is integrable and αf dµ = α f dµ.
If α > 0
(αf )+ = max{αf (x), 0} = α max{f (x), 0}

= αf +

(αf )− = − min{αf (x), 0} = −α min{f (x), 0} = α(− min{f (x), 0})

= αf −

Thus (αf )+ , (αf )− are integrable. Hence (αf ) is integrable and


Z Z Z Z Z Z
+ − + −
αf dµ = (αf ) dµ − (αf ) dµ = α f dµ − α f dµ = α f dµ

If α < 0

(αf )+ = max{αf (x), 0} = − min{−αf (x), 0} = −α(− min{f (x), 0}) = −αf −

(αf )− = − min{αf (x), 0} = max{−αf (x), 0}

= (−α) max{f (x), 0} = −αf +

Thus (αf )+ , (αf )− are integrable. Hence (αf ) is integrable and


R R
αf dµ = α f dµ
Next we consider the case of f + g

(f + g)+ = max{(f + g)(x), 0}) = max{f (x) + g(x), 0}

2.6
Chapter 2 Functions and integrals

≤ max{f (x), 0} + max{g(x), 0} = f + + g +

(f + g)− = − min{(f + g)(x), 0}) = − min{f (x) + g(x), 0}

≤ − min{f (x), 0} + (− min{g(x), 0}) = f − + g −


Z Z Z
∴ (f + g) dµ ≤ f dµ + g + dµ < ∞
+ +

and Z Z Z
− −
(f + g) dµ ≤ f dµ + g − dµ < ∞

Therefore f + g is integrable.
Now (f + g)dµ = {(f + − f − ) + (g + − g − )}dµ
R R

Z Z Z Z Z

(f + g)dµ = +
f dµ − f dµ + +
g dµ − g − dµ

Z Z

= +
(f − f )dµ + (g + − g − )dµ
Z Z
= f dµ + gdµ

If f (x) ≤ g(x) holds at each x in X. Then

g−f ≥0
Z
∴ (g − f )dµ ≥ 0
Z Z
∴ gdµ − f dµ ≥ 0

[As g −f has becomes non negative function,all the properties of non negative functions hold good]
Z Z
⇒ gdµ ≥ f dµ

Proposition 2.6. Let (X, A , µ) be a measure space.Let f be [−∞, ∞] valued A -measurable func-
tion on X.Then f is integrable iff |f | is integrable.If these functions are integrable ,then
Z Z
| f dµ| ≤ |f |dµ

Proof. f is integrable
⇒ f + , f − are integrable.Since |f | = f + + f − ,so,|f | is integrable.

2.7
Chapter 2 Functions and integrals

Conversely, |f | is integrable
⇒ f + , f − are integrable.Since f = f + − f − ,so,f is integrable.
Z Z Z Z

| f dµ| = | +
(f − f )dµ| = | +
f dµ − f − dµ|
Z Z Z Z Z
+ − + −
≤| f dµ| + | f dµ| ≤ |f |dµ + |f |dµ = |f |dµ

Proposition 2.7. Let (X, A , µ) be a measure space, and let f and g be [−∞, ∞] valued A -
R R
measurable function on X that agree almost everywhere.Prove that if either f dµ or gdµ ex-
R R
ists,then both exist and f dµ = gdµ

Proof. First we consider that f and g are non negative. Let

A = {x ∈ X : f (x) 6= g(x)}

Let h be defined by

h(x) = ∞ if x ∈ A

= 0 if x ∈
/A

We consider hn = nχA ,then {hn } is a non decreasing sequence of functions in J+ and limn→∞ hn = h
Therefore, Z Z
hdµ = lim hn dµ
n→∞

Now Z Z
hn dµ = nχA dµ = nµ(A) = 0

[µ(A) = 0] Z Z
∴ hdµ = lim hn dµ = 0
n

Since
f ≤g+h
Z Z Z
∴ f dµ ≤ gdµ + hdµ
Z Z
∴ f dµ ≤ gdµ

Again
g ≤f +h

2.8
Chapter 2 Functions and integrals

Z Z
∴ gdµ ≤ f dµ

Comparing the two inequalities we have


Z Z
f dµ = gdµ

If f and g are not non negative, then


f = f + −f − and g = g + −g − . Since f (x) = g(x) almost everywhere,so, f + = g + almost everywhere
and f + = g + almost everywhere
Therefore, f + dµ = g + dµ and f − dµ = g − dµ
R R R R

Z Z Z Z

∴ f dµ = +
(f − f )dµ = +
f dµ − f − dµ

Z Z Z Z
+ − + −
= g dµ − g dµ = (g − g )dµ = gdµ

Proposition 2.8. Let (X, A , µ) be a measure space and let f be [0, ∞] valued A -measurable
function on X. If t is a positive real number and if At is defined by

At = {x ∈ X : f (x) ≥ t}

,then Z Z
1 1
µ(At ) ≤ f dµ = f dµ
t At t
Let (X, A , µ) be a measure space and let f be [−∞, ∞] valued A -measurable function on X
R
that satisfies |f |dµ = 0.Then f vanishes almost everywhere.

Proof. From the previous theorem applied to the function |f | implies that
Z
1
µ({x ∈ X : |f (x)| ≥ }) ≤ n |f |dµ = 0
n

holds for each positive integer n.Since


[ 1
{x ∈ X : f (x) 6= 0} = {x ∈ X : |f (x)| ≥ }
n
n

X 1
∴ {x ∈ X : f (x) 6= 0} ≤ {x ∈ X : |f (x)| ≥ }=0
n
n

Hence f vanishes almost everywhere.

2.9
Chapter 2 Functions and integrals

Let (X, A , µ) be a measure space and let f be [−∞, ∞] valued integrable function on X.Then
|f (x)| < ∞ holds almost everywhere.

2.10

You might also like