Measure Theory

MEASURE THEORY
Oscar Blasco
Contents
1 Abstract Measure
1.1 Basic notions on sets . . . . . .
1.2 Basic notions on set functions. .
1.3 Outer measures. . . . . . . . . .
1.4 Extension of measures. . . . . .
1.5 Borel-Stieltjes measures on R. .
1.6 Measurable and non-measurable
1.7 Exercises . . . . . . . . . . . . .
. . .
. . .
. . .
. . .
. . .
sets.
. . .
2 Measurable and Integrable functions

2.1 Measurable functions . . . . . . . . .
2.2 Some types of convergence. . . . . . .
2.3 Integrable functions. . . . . . . . . .
2.4 Exercises . . . . . . . . . . . . . . . .
3 The
3.1
3.2
3.3
3.4
product measure and Fubinis

The product measure . . . . . . .
Fubini theorem . . . . . . . . . .
Applications . . . . . . . . . . . .
Exercises . . . . . . . . . . . . . .
4 The
4.1
4.2
4.3
4.4
Radon-Nikodym Theorem
Complex and real measures. .
The theorem and its proof. . .
Applications . . . . . . . . . .
Exercises . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
theorem
. . . . . .
. . . . . .
. . . . . .
. . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
5
9
11
15
17
23
26
.
.
.
.
29
29
35
38
50
.
.
.
.
57
57
63
64
68
.
.
.
.
73
73
80
84
88
Chapter 1
Abstract Measure
1.1
Basic notions on sets
Definition 1.1.1 A non empty family of subsets of X, say A P(X), is

called an algebra if
(i) A,
(ii) If A A then X \ A A,
(iii) If A, B A then A B A.
Definition 1.1.2 A non empty family of subsets of X, say P(X), is
called a -algebra if
(i) ,
(ii) If A then X \ A ,
(iii) If An for all n N then n An .
Definition 1.1.3 A non empty family of subsets of X, say M P(X), is
called a monotone class if for all monotone sequence of sets An M,
increasing An An+1 (respect. decreasing An+1 An ), then n An M
(respect. n An M).
Definition 1.1.4 A non empty family of subsets of X, say R P(X), is
called a ring if
(i) R,
(ii) If A, B R then A \ B R,
(iii) If A, B R then A B R.
5
Chapter 1. Abstract measure
Remark 1.1.1 Let P(X). is a -algebra if and only if is a monotone class and an algebra.
Let A P(X). A is an algebra if and only if A is a ring containing X.
Example 1.1.1 (1) The trivial -algebras are P(X) and = {, X}.
(2) Let M = {An : n N} X where A1 = , An An+1 for all n N
and n An = X. This is a monotone class but not necessarily a ring.
(3) Let X = [0, 1) and A = {finite unions of intervals [a, b), 0 a b
1} is an algebra but not -algebra.
If the intervals in the previous family are assumed to have 0 a b < 1
then it is a ring but not an algebra.
(4) Let X be non empty and numerable. R = {A X : card(A) < } is
a ring but not algebra.
Definition 1.1.5 Let An P(X) for n N. The upper limit (respect. lower
limit) of the sequence is defined by
lim sup An =
n=1 k=n Ak
(respect.
lim inf An =
n=1 k=n Ak .)
A sequence is said to have limit if lim sup An = lim inf An . Such a set is
called lim An .
Remark 1.1.2 Any monotone sequence has a limit. If An is increasing
(respect. decreasing) then lim An = n An (respect. lim An = n An ).
Proposition 1.1.6 Let R be a ring. If A, B R then A M B R and
A B R.
Proof: Note that A M B = ((A B) \ A) ((A B) \ B) and A B =
A B \ (A M B).
Proposition 1.1.7 Let be a -algebra. If An for all n N then
n An , lim sup An and lim inf An
Proof: Write n An = X \ n (X \ An ) and apply the properties of an
-algebra.
1.1. Basic notions on sets
Definition 1.1.8 A pair (X, ) given by a non-empty set X and a -algebra

over X is called a measurable space.
We shall give several methods of constructing measurable spaces. The
proofs are rather straightforward and left to the reader.
Definition 1.1.9 (Induced -algebra) Let (X, ) be a measurable space and
let Y X. Then Y = {A Y : A } is a -algebra over Y .
If Y then Y = {A Y, A }.
Definition 1.1.10 (Image of a -algebra ) Let (X, ) be a measurable space
and let f : X Y be a function. We define
f () = {B Y : f 1 (B) }.
Then (Y, f ()) is a measurable space.
Definition 1.1.11 (-algebra generated by a family) Let F P(X). We
denote (F) the smallest -algebra containing F, which is called the -algebra
generated by F.
It is elementary to see that (F) = { : -algebra , F }.
Remark 1.1.3 (1) If F1 F2 (F1 ) then (F1 ) = (F2 ).
(2) F1 (F2 ) and F2 (F1 ) if and only if (F1 ) = (F2 ).
Definition 1.1.12 (Borel -algebra ) Let (X, ) be a topological space and
let G be the collection of open sets for the topology . The -algebra (G)
is called the Borel -algebra and denoted B(X). The elements in B(X) are
called Borel sets.
Remark 1.1.4 Closed sets, G sets (numerable intersection of open sets) or
F sets (numerable union of closed sets) are examples of Borel sets.
Using Remark 1.1.3 one easily sees the following facts:
If F denotes the collection of closed sets in then B(X) = (F).
If (X, d) is a separable metric space (or a metric space where any open
set is a numerable union of balls) and E = {B(x, r) : x X, r > 0} then
B(X) = (E).
Proposition 1.1.13 Let X = R and let us consider the following collections
E1 = {(a, b) : a b}, E2 = {(a, b] : a b}, E3 = {[a, b) : a b} and
E4 = {[a, b] : a b}. Then B(R) = (Ei ) for i = 1, 2, 3, 4.
Proof: Notice that any open set is a numerable union of open intervals.
Hence Remark 1.1.3 shows that B(R) = (E1 ).
Let us observe that (a, b] = n (a, b + n1 ), [a, b) = n (a n1 , b) and [a, b] =
n (a n1 , b + n1 ) to get the other cases.
Proposition 1.1.14 Let n N and X = Rn and consider E = {(a1 , b1 ]
(a2 , b2 ] ...(an , bn ] : ai bi , i = 1, ..., n}. Then B(Rn ) = (E).
Proof: We sketch the proof for the case n = 2.
Consider J0 = {(n, n + 1] (m, m + 1] : n, m Z} and Jn the collection
of intervals resulting of dividing each square of the previous family into four
of the same area.
Now for each x R2 there exists a unique sequence of intervals Ik (x) Jk
such that x Ik (x) for all k N, Ik+1 (x) Ik (x) and Area(Ik (x)) 0 as
k .
Given an open set G we can consider the family F = {J k Jk : J G}.
Of course G = JF J and with a little effort it can be seen that only a
numerable number of sets is needed. Since F E then the proof is finished.
Definition 1.1.15 Let E P(X). M(E) stands for the smallest monotone
class containing E, which is called the monotone class generated by E.
Theorem 1.1.16 (The monotone class theorem) Let A be an algebra over
X. Then (A) = M(A).
Proof: It suffices to see that M(A) is -algebra . Since M(A) is a monotone
class we have only to show that it is an algebra.
Clearly A M(A). Now given A M(A), to see that X \A M(A)
let us define = {A M(A) : X \ A M(A)} and show that = M(A).
For such a purpose we shall see that is a monotone class and contains A.
Indeed, if An is a monotone sequence then lim An . Clearly if A A
then X \ A A M(A). Therefore A .
Given now A, B M(A), we need to show that A B M(A). Let us
define A = {B M(A) : A B M(A)}. Note that A is a monotone class
(since (Bn )n monotone sequence in A has limit in A ).
Now let us deal first with the case A A. In this case A contains A
(since B A implies A B A M(A)). Therefore A = M(A) for all
A A.
1.2. Basic notions on set functions.
Now let us consider A M(A) \ A. To see that A A observe that

B A belongs to A since we have that B A if and only if A B and
this was shown to be true in the previous case.
1.2
Basic notions on set functions.
Definition 1.2.1 Let A be an algebra over X, a set function : A [0, ]

is called additive (or finitely additive) if (A B) = (A) + (B) for all
A, B A such that A B = .
P
is called a measure over A if () = 0 and (n An ) =
n=1 (An ) for
any sequence (An ) of pairwise disjoint sets in A such that n An A.
is a finite measure if (X) < . is -finite if there exist Xn A
such that (Xn ) < and X = n Xn .
Proposition 1.2.2 Let : A [0, ] be a additive set function.
(1) If there exists A A with (A) 6= then () = 0.
(2) is monotone, i.e. if A, B A and A B then (A) (B).
(3) If A, B A and A B with (A) < then (B \A) = (B)(A).
(4) If is a measure then is subadditive, i.e. if (An ) A and n An A
P
then (n An )
n=1 (An ).
(5) If is a measure and (An ) A is increasing and n An A then
(lim An ) = (n An ) = lim (An ).
n
(6) If is a measure and (An ) A is decreasing, (A1 ) < and

n An A then
(lim
An ) = (n An ) = lim
(An ).
n
n
Proof: (1) Observe that (A ) = (A) + ().
(2) and (3) Note that (B) = (A) + (B \ A).
(4) Define B1 = A1 and Bn = An \ n1
k=1 Ak for n 2. Now (Bn ) is a
sequence in A of pairwise sets such that n An = n Bn and
(n An ) = (n An ) =
X
n=1
(Bn )
(An ).
n=1
(5) Put B1 = A1 and Bn = An \ An1 for n 2. Note that (Bn ) is a

sequence in A of pairwise sets such that An = nk=1 Bk . Hence
(An ) =
n
X
k=1
(Bk )
X
n=1
(Bn ) = (n Bn ) = (n An ).
10
(6) Put Bn = A1 \ An . Then (Bn ) is an increasing sequence in A, n Bn =

A1 \ n An and (Bn ) = (A1 ) (An ). Now applying (5) we get the corresponding result.
Definition 1.2.3 A triplet (X, , ) given by a non-empty set X, a -algebra

over X and a measure over is called a measure space.
It is said to be finite or -finite if the measure is finite or -finite over .
It is said to be complete if for any A with (A) = 0 and B A then
B and hence (B) = 0.
Remark 1.2.1 If (X, , ) is a -finite measure space then X can be splited
into a sequence (Xn ) of either disjoint or increasing sets in such that
X = n Xn and (Xn ) < .
Example 1.2.1 (Counting measure) Let X 6= , = P(X) and (A) =
card(A).
Example 1.2.2 (Dirac mass) Let X 6= and a X. Put = P(X) and
a (A) = 1 if a A and a (A) = 0 if a
/ A.
Example 1.2.3 (Induced measure) If (X, , ) is a measure space and Y
, we define Y (B) = (B Y ) for all B . Hence (Y, Y , Y ) is a
measure space.
Example 1.2.4 (Image measure) If (X, , ) is a measure space and f :
X Y is a function, we define f ()(B) = (f 1 (B)) for all B f ().
Hence (Y, f (), f ()) is a measure space.
Theorem 1.2.4 Let (X, , ) be measure space. Let us define
N = {N P(X) : N B, B , (B) = 0}.
= N and
= (A) for A = A N for some A and
Put
(A)
and
N N. Then (X, ,
) is a complete measure space such that
(A) = (A) for all A .

(X, ,
) is called the completion of (X, , ).
1.3. Outer measures.
11
is -algebra observe first that N. Now if

Proof: To see that
A = A N for some A and N N where N B for B and
(B) = 0 then we have
X \ A = (X \ (A B)) (B \ (N A)) .
If An = An Nn where An and Nn N for all n N. Hence
n An = (n An ) (n Nn ) .
Indeed, if A N = A0 N 0
Let us now show that
is well defined on .
0
0
where A, A and N, N N where N B and N 0 B 0 for B, B 0
and (B 0 ) = (B) = 0 then
(A) (A0 B 0 ) (A0 ) and (A0 ) (A B) (A).
If H A N with
It is elementary to see that
is a measure over .
(A N ) = (A) = 0 then H N . This shows the completeness and

the proof is finished.
Next proposition is immediate and left to the reader.
if and only if
Proposition 1.2.5 Let (X, , ) be a measure space. A
there exist A1 , A2 such that A1 A A2 and (A2 \ A1 ) = 0.
1.3
Outer measures.
We would like to be able to extend measures defined in an algebra A

to a bigger family, for instance to the generated -algebra or even to P(X).
Let us give a procedure which allows to get an extension preserving some
properties.
Proposition 1.3.1 Let (X, , ) be a measure space. We define
(A) = inf{(E) : A E, E }
for any A P(X). Then
(i) is an extension of ,
(ii) is monotone, i.e. if A B then (A) (B) and
P
(iii) is subadditive, i.e. (n An )

n=1 (An ).
12
Proof: (i) is immediate.

(ii) If A B and E with B E then A E. Thus, (A) (B).
(iii) Given (An ) P(X) we may assume that (An ) < for all n N.
Given > 0 there exist En , An En such that (An ) + 2n > (En )
(An ). Hence
(n An ) (n En )
(En )
(An ) + .
Motivated by these properties we give the following definition.

Definition 1.3.2 Let X 6= . A monotone and subadditive set function
: P(X) [0, ] with () = 0 is called an outer measure.
Proposition 1.3.3 Let : P(X) [0, ] be an outer measure. Then is
a measure if and only if is additive.
Proof: Assume is additive, and take (An ) a sequence of pairwise disjoint
sets. Since
n
X
(Ak ) = (nk=1 Ak ) (n An )
(An ),
k=1
passing to the limit we get the result.

Let us now give a procedure of constructing outer measures from measures
defined on algebras which generalize the method in Proposition 1.3.1.
Proposition 1.3.4 Let A be an algebra over X and : A [0, ] a measure on A. Let us define
(A) = inf{
(Ei ) : A i Ei , Ei A, i N}.
i=1
Then is an outer measure which extends .

Proof: The facts () = 0 and that is monotone are immediate.
Given (An ) P(X) we may assume that (An ) < for all n N.
Given n N and > 0 there exist En,k , An k En,k such that
P
(An ) + 2n > k (En,k) (An ). Hence
(n An ) (n,k En )
X
n,k
(En,k )
X
n
(An ) + .
1.3. Outer measures.
13
To see that extends note first that if A A then (A) (A). On

the other hand if A i Ei for (Ei ) A we have A = i (A Ei ). Hence
(A)
X
i
(A Ei )
(Ei ).
This gives that (A) (A) and the proof is finished.

Remark 1.3.1 Observe that in the case of A being a -algebra the procedures of Propositions 1.3.1 and 1.3.4 coincide.
Remark 1.3.2 The procedure in Proposition 1.3.4 does not need neither A
to be an algebra nor to be a measure. This actually can be also done for F
semiring, which is an family of sets with the properties
(i) F,
(ii) If A, B F then A B F and
(iii) If A, B F, A B then B \ A nk=1 Ck for a finite family of
pairwise disjoint sets Ck F and for premeasures : F [0, ], which are
set functions with the properties
(i) () = 0,
(ii) If A, B F are disjoint then (A B) = (A) + (B) and
P
(iii) If A, An F for n N and A n An then (A) n (An ).
Let us now give a procedure to obtain measures from outer measures.
Definition 1.3.5 Let be an outer measure on X. A set A is called measurable if, for all E P(X),
(E) = (E A) + (E (X \ A)),
(or equivalently (E) (E A) + (E \ A) .)
We denote by the family of all -measurable sets.
Theorem 1.3.6 (Caratheodorys theorem) Let be an outer measure on
X. Then (X, , ) is a complete measure space, where denotes the
restriction of to .
Proof: We first prove that is an -algebra .
Obviously and if A if and only if X \ A .
14
Assume A, B . Let us see that A B . For each E P(X) we

have, since (A B) E = (A E) (E (X \ A) B), that
((A B) E) + ((X \ (A B)) E)
(A E) + (E (X \ A) B) + (E (X \ A) (X \ B))
(A E) + (E (X \ A)) = (E).
Therefore is an algebra.
Before proving that is an -algebra let us observe that for any E
P(X) we have that E (A) = (E A) is a measure on .
Indeed, if A1 and A2 and A1 A2 = then, denoting E 0 = (E
A1 ) (E A2 ), we have
E (A1 A2 ) =
=
=
=
((E A1 ) (E A2 )
(E 0 A1 ) + (E 0 (X \ A1 ))
(E A1 ) + (E A2 )
E (A1 ) + E (A2 )
If (An ) is a sequence of pairwise disjoint -measurable sets we have for

all n N
E (n An ) E (N
k=1 Ak ) =
N
X
E (An ).
k=1
Hence passing to the limit as n we get the result.

In particular, for E = X we have that is countably additive on .
To see that is -algebra we need to show that countably union of
pairwise disjoint of -measurable sets (since any union of elements in an
algebra can be written as a union of pairwise disjoint sets in the algebra) is
-measurable. Given now such a sequence (An ) we have for each N N
N
(E) = (E (N
k=1 Ak )) + (E (X \ k=1 Ak ))
(N
k=1 (E Ak )) + (E (X \ k=1 Ak ))
N
X
(E Ak ) + (E (X \
k=1 Ak ))
k=1
Taking limits as N we get

(E)
(E Ak ) + (E (X \
k=1 Ak ))
k=1
= (E (
k=1 Ak )) + (E (X \ k=1 Ak )).
1.4. Extension of measures.
15
To finish the proof we need to show that is complete. Note that

(A) = 0 implies A . Hence the monotonicity of gives also the
completeness.
1.4
Extension of measures.
Theorem 1.4.1 (Hahns extension theorem) Let A be an algebra on X and

a measure on A. There exists a measure
defined on the -algebra (A)
which extends .
Moreover if is -finite the extension is unique and -finite.
Proof: Let us consider the outer measure from Proposition 1.3.4, that is
(A) = inf{
(Ai ) : A i Ai , Ai A, i N}.
i=1
We now apply Caratheodorys method to get the -algebra of . If we

prove that (A) and we define
the restriction of to (A) we have
the result.
Actually we only need to show that A . Let A A and E P(X).
For each > 0 we have a sequence (An ) in A such that E n An and
(E)
(An ) (E) + .
Hence
(E A) + (E (X \ A)) (n (An A)) + (n (An (X \ A))
X
X
(An A) +
(An (X \ A))
n
X
n
(An A) +
(An (X \ A))
(An ) (E) + .
We show now that

es unique if is -finite on A, say X = n Xn where
Xn A, Xn Xn+1 and n (Xn ) < . Assume that there are two measures
1 and 2 on (A) which extend . Denote by 1,n (A) = 1 (A Xn ) and
2,n (A) = 2 (A Xn ) for any A (A). For each n N denote
Mn = {A (A) : 1,n (A) = 2,n (A)}.
16
Using that i,n are finite we have that Mn is a monotone class (see (5) and
(6) in Proposition 1.2.2) which clearly contains A. Therefore 1,n = 2,n and
hence 1 = 2 on (A).
Now
is -finite because it extends .
Remark 1.4.1 Take X = [0, 1) and the algebra A given by finite unions
of intervals [a, b), 0 a b 1. Consider the measure on A such that
() = 0 and (A) = for any 6= A A.
We can get two different extensions to (A). The counting measure and
the measure 1 defined by 1 () = 0 and 1 (A) = for any 6= A (A).
They are different since ({1/2}) = 1 and 1 ({1/2}) = .
Theorem 1.4.2 Let A be an algebra on X and : (A) [0, ) a measure
such that there exists Xn A of finite measure such that X = n Xn . Then
for each > 0 and each A (A) with (A) < there exists B A such
that (A4B) < .
Proof: Using the uniqueness in Hahns theorem we have that
(A) = inf{
(An ) : A n An , An A}
for all A (A).

Now given > 0 and A (A) with (A) < there exists a sequence
P
(An ) in A such that A n An and (A) n (An ) < (A) + 2 .
P
N
Take now N N such that
k=N +1 (Ak ) 2 and define B = k=1 Ak A.
Clearly (A4B) (
k=N +1 Ak ) + (n An \ A) .
Lemma 1.4.3 Let be a measure on and (A) = inf{(E) : A E, E
}. If A P(X) then there exists E such that A E and (E) = (A).
Proof: In the case (A) = take E = X. If (A) < we can select,
for each n N, En such that (En ) < (A) + n1 . Then E = n En
and (E) = (A).
Theorem 1.4.4 Let (X, , ) be a finite measure space. Then the com
pletion (X, ,
) coincides with (X, , ).
1.5. Borel-Stieltjes measures on R.
17
Proof: Recall that from Proposition 1.3.1 we have that

(A) = inf{(E) : A E, E }.
Let us see that .
For each A and E P(X), select, using Lemma 1.4.3, E1 with
E E1 and (E1 ) = (E) . Hence
(E A) + (E (X \ A)) (E1 A) + (E1 (X \ A)) = (E1 ) = (E).
On the other hand if B N for some N and (N ) = (N ) = 0 then
(B) = 0. Since any -nul set is -measurable then the set N associated
.
to (X; , ) is contained in . This shows that
Since restricted to coincides with we get also that

= on .
Conversely, assume first that A and (A) < . Take E such

that A E and (E) = (A). Hence
(E) = (E A) + (E (X \ A))
= (A) + (E \ A)
= (E) + (E \ A)
Therefore (E \ A) = 0. This allows to get B with E \ A B and
(B) = 0. So E \ B A E and this gives that A .

The case (A) = can be done by writting A = n (A Xn ) where
Xn and (Xn ) < . Since A Xn for all n N we obtain, from
the previous case, that A .
1.5
Borel-Stieltjes measures on R.
Throughout this section I denotes an interval in R with not empty interior, int(I) 6= , and let x0 int(T ).
Definition 1.5.1 Given a measure on B(I) which is finite on bounded
intervals contained in I we define F : I R as the function F (x) =
((x0 , x]) for x I, x x0 and F (x) = ((x, x0 ]) for x I, x < x0 .
Proposition 1.5.2 If : B(I) [0, ] is a measure such that (J) <
for all bounded interval J I, then F is increasing, right continuous and
F (x0 ) = 0.
Moreover F is continuous at x if and only if ({x}) = 0.
18
Proof: We shall see that if x1 , x2 I with x1 < x2 then F (x2 ) F (x1 ) =

((x1 , x2 ]) 0. Hence F is increasing.
Indeed, note that if x0 x1 < x2 then
F (x2 ) F (x1 ) = ((x0 , x2 ]) (x0 , x1 ]) = ((x1 , x2 ]),
if x1 x0 < x2 then
F (x2 ) F (x1 ) = ((x0 , x2 ]) + (x1 , x0 ]) = ((x1 , x2 ]),
and if x1 < x2 x0 then
F (x2 ) F (x1 ) = ((x2 , x0 ]) + (x1 , x0 ]) = ((x1 , x2 ]).
To see that F is right continuous we fix x I and a sequence decreasing
(xn ) I such that xn > x and limn xn = x. Since limn (x, xn ] = then
limn F (xn ) F (x) = limn ((x, xn ]) = 0.
To see the last part, observe that for an increasing sequence (xn ) I such
that xn x and limn xn = x we have limn F (xn )F (x) = limn ((x, xn ]) =
({x})
Definition 1.5.3 An increasing and right continuous function F : I R is
called a distribution function over I. For such a function we define for each
A I the set function
F (A) = inf{
F (bk ) F (ak ) : A k (ak , bk ], ak , bk int(I)}.
k=1
Our aim is, given a distribution function F to find a measure such that
F = F . For such a purpose we first need the following lemma.
Proposition 1.5.4 Let F : I R an increasing function. If we have(a, b]
nk=1 (ak , bk ] I where a1 a2 ...an and a, b, ak I for k = 1, ..., n, then
F (b) F (a)
n
X
F (bk ) F (ak ).
k=1
Proof: We may assume that (a, b] (ak , bk ] 6= for all k and,since (a, b]
must be contained in one of its connected components, also assume that
nk=1 (ak , bk ] = (ak0 , bk0 ], .
19
We shall prove it using induction. The case n = 1 is obvious.

Let us assume the result is true for any family of m intervals for m < n.
Put no = max{k : ak a} and n1 = min{j : b bj }.
We may assume that n1 > n0 , since n0 n1 implies that (a, b] (an0 , bn0 ]
and the result is clear.
We now have two possibilities:
If (an0 , bn0 ] (an1 , bn1 ] 6= then bn0 > an1 what gives
F (b) F (a) F (bn1 ) F (an0 ) F (bn1 ) F (an1 ) + F (bn0 ) F (an0 ).
If (an0 , bn0 ] (an1 , bn1 ] = then (bn0 , an1 ] k6=no ,n1 (ak , bk ]. Using now the
induction assumption we have
X
F (an1 ) F (bn0 )
F (bk ) F (ak )
k6=n0 ,n1
and therefore
F (b) F (a) F (bn1 ) F (an0 )
F (bn1 ) F (an1 ) + F (an1 ) F (bn0 ) + F (bn0 ) F (an0 )
n
X
F (bk ) F (ak ).
k=1
Theorem 1.5.5 If F is a distribution function over I then

(i) F is an outer measure,
(ii) F ((a, b]) = F (b) F (a) for all a, b int(I) and
(iii) any borel set is F -measurable.
Proof: (i) Using that (a, a] then F () = 0.
Clearly F (A) F (B) if A B.
P
Let (An ) I with n F (An ) < . For each > 0 we find (ank , bnk ] such
that An k (ank , bnk ] and
F (bnk ) F (ank ) < F (An ) +
k=1
Hence
F (n An )
X
k,n=1
F (bnk ) F (ank )
X
n=1
.
2n
F (An ) + .
20
(ii) Clearly if A = (a, b] then F (A) F (b) F (a). Assume now that
(a, b] k (ak , bk ] where ak , bk int(I).
Given > 0, since F is right continuous there exists a0 > a so that
F (a0 ) F (a) < 2 and there exist b0k > bk such that F (b0k ) F (bk ) < 2k+1
for
all k .N.
0
Since [a0 , b] k (ak , b0k ), using compactness, we have [a0 , b] N
k=1 (ak , bk )
for some N N. Now we can apply Proposition 1.5.4 to get
F (b) F (a0 )
N
X
k=1
F (b0k ) F (ak )
N
X
F (bk ) F (ak ) + .
2
k=1
Hence F (b) F (a) N

k=1 F (bk ) F (ak ) + .
(iii) It suffices to see that (a, b] is F -measurable.
Let E I with F (E) < (the other case is obvious) and > 0.
P
Let us take E k (ak , bk ] where ak , bk int(I) and
k=1 F (bk ) F (ak ) <
F (E) + . Now
P
F ((a, b] E) + F ((I \ (a, b]) E)
F ((a, b] (ak , bk ]) + F ((I \ (a, b]) (ak , bk ]) =
Ak .
Observe now that if (a, b] (ak , bk ] 6= then coincides with the interval
(max(a, ak ), min(b, bk )]. We have four situations, namely
Case a ak and b bk .
Hence (a, b] (ak , bk ] = (ak , b] and (I \ (a, b]) (ak , bk ] = (b, bk ] what
implies that
Ak = F (b) F (ak ) + F (bk ) F (b) = F (bk ) F (ak ),
Case a ak and bk < b.
Hence (a, b] (ak , bk ] = (ak , bk ] and (I \ (a, b]) (ak , bk ] = what implies
that
Ak = F (bk ) F (ak )
Case ak < a and b bk .
Hence (a, b] (ak , bk ] = (a, b] and (I \ (a, b]) (ak , bk ] = (ak , a] (b, bk ]
what implies that
Ak = F (b) F (a) + (F (a) F (ak ) + F (bk ) F (b)) = F (bk ) F (ak ),
Case ak < a and bk < b.
21
Hence (a, b] (ak , bk ] = (a, bk ] and (I \ (a, b]) (ak , bk ] = (ak , a] what
implies that
Ak = F (bk ) F (a) + F (a) F (ak ) = F (bk ) F (ak ).
Therefore
F ((a, b] E) + F ((I \ (a, b]) E)
X
F (bk ) F (ak )
k
< F (E) +
Definition 1.5.6 Given a distribution function F over I we denote by MF

and mF the family of F -measurable sets and the measure F restricted to
MF obtained by using the Caratheodory method. These are called LebesgueStieltjes measurable sets and the Lebesgue-Stieltjes measure respectively. The
case I = R and F (x) = x corresponds to the Lebesgue measure space and it
is simply denoted (R, M, m).
Theorem 1.5.7 Let F : I R be a distribution function. Then there exists
a unique Borel measure mF : B(I) [0, ] verifying that mF ((a, b]) =
F (b) F (a) for all a, b int(I). (This is called the Borel-Stieltjes measure
associated to the distribution function F .)
Proof: Consider the outer measure F and let be the measure F restricted to B. Note that any interval can be decomposed as I = n An where
An = (an , bn ] (and eventually A0 = {a} in the case I = [a, b] or I = [a, ])
and, of course (An ) < . Therefore we can use the uniqueness of the Hahn
theorem to conclude the result.
Remark 1.5.1 Let us list some properties of mF which are left to the reader.
(1) F is continuous at x if and only if mF ({x}) = 0.
(In particular m(N ) = 0 for all numerable set in R. )
(2) mF is always -finite and mF is finite if and only if F is bounded.
(3) For any a, b int(I) we have mF ((a, b)) = F (b )F (a), mF ([a, b)) =
F (b )F (a ), and mF ([a, b]) = F (b)F (a ), where F (a ) = limxa F (x).
22
Let us now characterize the Legesgue measure on R by means of its invariance under translations.
Theorem 1.5.8 Let be a Borel measure over R which is invariant under
translations and finite over bounded intervals in R. Then = cm where
c = (0, 1]) and m is the Lebesgue measure restricted to B.
Proof: Let us first see that ((a, b]) = c(b a) where c = ((0, 1]).
Since (a, b] = a + (0, b a] it suffices to see that ((0, x]) = cx for any
x > 0. Since x = limn qn where qn Q, and (qn ) is an increasing sequence,
then it is enough to prove that ((0, q]) = cq for any q > 0 and q Q.
where m, n N and observing that
Now writing q = m
n
(0,
m
1
1 2
m1 m
] = (0, ] ( , ] ... (
, ]
n
n
n n
n
n
k1 k
1
we have that ((0, m
]) = m
k=1 (( n , n ]) = m(0, n ]). Finally to see that
n
(0, n1 ]) = nc simply note that 1 = nn and then ((0, 1]) = n(0, n1 ]).
Now let A be the algebra generated by {(a, b] : a b} which is easily
seen to coincide with finite unions of intervals (a, b] or (a, ) where
a b < . Using that R = (n, n] and (n, n] = 2nc we can apply
the uniqueness of the Hanh theorem to conclude that = cm (because both
measures coincide on A).
Let us now use general theory to get information on Lebesgue measurable

sets. Recall that Theorem 1.4.4 one has that the complection of (R, B, mF )
coincides with (R, M, mF ). In particular, E is a Lebesgue measurable set if
and only if there exist Borel sets A, B such that A E B and m(B \ A) =
0. We can improve this result as follows.
Theorem 1.5.9 Let E R, F (x) = x and write F = . The following are
equivalent:
(1) E is a Lebesgue measurable set.
(2) For any > 0 there exists an open set G with E G and (G\E) < .
(3) For any > 0 there exists a closed set F with F E and (E\F ) < .
(4) There exist A F (countable union of closed sets) and B G
(countable intersection of open sets) with A E B and m(B \ A) = 0.
Proof: (1)=(2) Assume m(E) = (E) < . There exist (ak , bk ] such
P
that E k (ak , bk ] and k bk ak < (E) + 2 .
1.6. Measurable and non-measurable sets.

Define G = k (ak , bk +
).
2k
23
Hence
(G) = m(G)
X
k
b k ak +
< m(E) + .
2
Therefore (G \ E) = m(G \ E) m(G) m(E) < .

Assume now m(E) = (E) = . First consider E (n, n) find open
sets Gn with En Gn and m(Gn \ En ) < 2n and then define G = n Gn
which satisfies E G and m(G \ E) < .
(2)=(1) For each n N, let Gn be an open set with E Gn and
m(Gn \ E) < n1 . Define G = n Gn and N = G \ E. Clearly (N ) = 0 and
E G.
Therefore F = R \ G B and R \ E = F N . Hence R \ E M and
also E M.
(1)(3) It follows from the equivalence between (1) and (2)
(1)=(4) Using that (2) and (3) holds we have a sequence of open sets Gn
1
and closed sets Fn such that Fn E Gn , m(Gn \ E) < 2n
and m(E \ Fn ) <
1
. Define A = n Fn and B = n Gn . We have m(B \ A) m(Fn \ Gn ) n1
2n
for all n N.
(4)=(1) It follows from the fact that M is the complection of B.
1.6
Measurable and non-measurable sets.
Let us now construct some examples of sets in B and M.

Example 1.6.1 A non numerable Borel set with measure zero: The Cantor
set in [0, 1].
Proof: Let us define the following family of open sets: First divide the
interval [0, 1] into three subintervals of the same measure. Select the open
interval in the middle. Now divide each of the remaining intervals into three
equal parts and select each open intervarl in the middle. Repeat this process.
Hence in the first step we select J1 = ( 31 , 23 ), in the second step we select
1 2
7 8
J2 = ( 91 , 29 ) ( 79 , 89 ), and in the third step we get J3 = ( 27
, 27 ) ( 27
, 27 )
25 26
17 18
( 27 , 27 ) ( 27 , 27 )
2k1
In this way we obtain a sequence of sets Jk = l=1
Il,k where Il,k are
disjoint open intervals of length 3k .
24
Now define the Cantor set by C = [0, 1] \ k Jk . Another description of C

P
n
is given by C = {x =
n=1 3n : n {0, 2}}.
Let us list some of its properties:
(1) C is non empty and compact.
(2) int(C) = .
(3) C is non numerable.
(4) m(C) = 0.
Clearly 13 C and [0, 1] \ C is an open set contained into [0, 1] what gives
(1).
To see (2) notice first that for each m N we have that C \ m
k=1 Jk is
a union of closed intervals of length 3m . Now if x int(C) then we would
have B(x, ) C for some > 0. Taking m such that 3m < we would
contradict the first observation.
Using the description in terms of ternary decompositions one gets that
card(C) = card(2N ) = 0 . This shows (3).
Finally
k1
m([0, 1] \ C) =
X
k
m(Jk ) =
X 2X
k
l=1
m(Ik,l ) =
X
k
2k1
1
= 1.
3k
Remark 1.6.1 card(B) 20 and card(M) = card(P(R)) = 21

Example 1.6.2 A non Lebesgue measurable set: The Vitali set in [0, 1].
Proof: Let us introduce in [0, 1] the equivalence relation: xRt if and only
if x y Q. Consider the set of equivalence classes [0, 1]/R.
Let us select a representant of each equivalence class and define the set
E = {e [0, 1] : e is a representant }.
The set E verifies that
[0, 1] rQ[1,1] (E + r) [1, 2]
and rQ[1,1] (E + r) is a countable union of pairwise disjoint sets.
Indeed, on the one hand, if x [0, 1] there exists e E such that x e
Q [1, 1] and if x E + r for some r Q [1, 1] then 1 x 2.
On the other hand r1 6= r2 implies that (E + r1 ) (E + r2 ) 6= , since
e1 + r1 = e2 + r2 gives [e1 ] = [e2 ] and then e1 = e2 .
1.6. Measurable and non-measurable sets.
25
Let us now show that E is not Lebesgue measurable. In case it is measurable one has
X
1
m(E + r) 3,
rQ[1,1]
and this leads to a contradiction, since m(E + r) = m(E) for all r.

Remark 1.6.2 For any A M with m(A) > 0 there exists E A such
that E is not Lebesgue measurable.
Proof: Same argument as in the case [0, 1] works for the set A[0, 1].
Example 1.6.3 A Lebesgue measurable set which is not a Borel set.
Proof: Let us first construct the Cantor function in [0, 1]. Define the
following sequence of piecewise linear functions:
f0 (x) = x,
f1 (x) = 21 if x ( 13 , 23 ) and then linearly with f1 (0) = 0 and f1 (1) = 1,
f2 (x) = 41 if x ( 19 , 29 ), f2 (x) = 34 if x ( 79 , 98 ) and then linearly with
f1 (0) = 0 and f1 (1) = 1,
In this way, inductively we construct fn (x). This take the constant value
l
in the set In,l and it is piecewise linear in the complement of Jn .
2k
Define f (x) = limn fn (x).
We now prove that it is increasing, non constant, continuous in [0, 1] and
derivable with f 0 (x) = 0 for all x [0, 1] \ C, where C is the Cantor set.
Since f is pointwise limit of increasing functions is increasing. Let x
/ C.
Then there exists n0 for which x In0 ,l then fn (y) = f (y) for all n n0 and
y In0 ,l . So f 0 (x) = 0.
To see that f is continuous we shall show that f is uniform limit of
continuous functions. Due to the construction we have
sup |fn (x) fm (x)|
0x1
m1
X
sup |fk (x) fk+1 (x)|
k=n 0x1
m1
X
sup |fk (x) fk+1 (x)|
1
k=n 0x 3k
m1
X
1
2k+1
k=n
1
1
.
6 k=n 2k
3
1
( )k k+1 |
2 3
26
Now we construct g : [0, 1] [0, 2] given by g(x) = x + f (x). We have

that g is a homeomorphism. Of course g is strictly increasing and continuous,
and hence there exists g 1 continuous and strictly increasing.
Taking into account that g(Ik,l ) = ck,l + Ik,l we get that
m(g(C)) = m(g([0, 1] \ k Jk ))
= m([0, 2] \ g(k Jk ))
X
= 2
m(g(Jk ))
k
k1
= 2
X 2X
k
m(ck,l + Ik,l )
l=1
k1
= 2
X 2X
k
m(Ik,l )
l=1
= 2 m([0, 1] \ C) = 1.
We then have that g(C) is a Borel set (using the fact, to be proved in
Exercise 1.7.5, that g(B([0, 1])) B([0, 2])) and m(g(C)) > 0. So we can find
E
/ M([0, 2]) with E g(C). Now consider A = g 1 (E)). This is a Lebesgue
measurable set (since A C) but A
/ B([0, 1]) (since g(A) = E
/ B([0, 2])).
1.7
Exercises
Exercise 1.7.1 (i) Let R be a ring. Show that (R, 4, ) is a ring in the
algebraic sense.
(ii) Let M be a non-empty family of sets in X and R(M) the ring generated by M. Show that any set in R(M) can be covered by a finite union of
sets in M.
Exercise 1.7.2 Let f : X Y be a function, A an algebra over Y . Let
(A) denote the -algebra generated by A and f 1 (A) = {f 1 (A) : A A}.
Then (f 1 (A)) coincides with f 1 ((A)).
Exercise 1.7.3 Let (X, , ) be a measure space and let (An ) be a sequence
in such that Aj intersects at most one other set in the sequence. Show that
(
j=1 Aj )
X
n=1
(Aj ) 2(
j=1 Aj ).
1.7. Exercises
27
Exercise 1.7.4 Study whether or not the following set functions are outer
measures.
(i) Let X = (ai,j ) be a matrix in M10 and define (A) = card{j : aij A}.
.
(ii) Let X = N and define (A) = limsup card(A{1,2,...,k})
k
(iii) Let X = Z and define (A) = 0 for A = ,
(A) =
a
if A is finite, where a = sup{|n| : n A} ,
a+1
(A) = 1 if A is infinite.
(iv) Let X be a metric space with distance d and let > 0. Define
= sup>0 inf{
((Ak )) : A = Ak , (Ak ) < }},
k=1
where (Ak ) = diam(Ak ) = sup{d(x, y) : x, y Ak }.

Exercise 1.7.5 Let X, Y be topological spaces and let g : X Y be continuous. Show that the B(Y ) g(B(X)).
Exercise 1.7.6 Let I be an open interval in R and F : I R a continuous
and strictly increasing function. Show that the Borel-Stieljes measure defined
by F coincides with the measure image by F 1 of the Lebesgue measure on
the Borel sets in F (I).
Exercise 1.7.7 Describe the Lebesgue-Stieltjes measure mF associated to the
following functions:
(i) F (x) = [x],
(ii) F (x) = [0,) (x),
(iii) F (x) = (x 1)+ .
+ (log(x) 1)[1,) .
Exercise 1.7.8 Let F (x) = 1
x (0,1)
(i) Find an unbounded Borel set A with 0 < mF (A) < .
(ii) Find an open set G such that 0 G0 and mF (G) < .
Exercise 1.7.9 Let mF be the Lebesgue-Stieltjes measure (0, ) associated
n2 +1
to F (x) = x for some > 0 and let A =
n=1 (n, n ).
Find the values of so that mF (A) < .
28
Exercise 1.7.10 Let (t) =

n=1 n(n, 2n+1
) (t).
2
Find F for the Lebesgue-Stieltjes measure associated to F to coincide with
the measure image by of the Lebesgue measure m, that is mF = (m).
P
Exercise 1.7.11 Let F : (0, ) R be given by F (t) = log(t). Show that

(i) mF = exp(m) where m is the Lebesgue measure and exp(x) = ex .
(ii) mF is invariant under dilations.
(iii) If : B((0, )) [0, ] is a measure finite over bounded intervals
and invariant under dilations then = CmF for some constant C > 0.
(iv) Find an unbounded open set G (0, ) such that 0 G0 and
mF (G) < .
Chapter 2
Measurable and Integrable
functions
2.1
Measurable functions
Definition 2.1.1 Let (X, ) be a measurable space. A function f : X

[0, ] is called measurable if the sets E = {x X : f (x) > } for all
> 0.
Some equivalent formulations are given in the following proposition.
Proposition 2.1.2 Let (X, ) be a measurable space and f : X [0, ].
The following are equivalent
(i) f is measurable.
(ii) {x X : f (x) } for all > 0.
(iii) {x X : f (x) < } for all > 0.
(iv) {x X : f (x) } for all > 0.
(v) {x X : f (x) = } and f 1 (B) for all Borel set B R+ .
(vi) {x X : f (x) = } and f 1 (G) for all open set G R+ .
(vii) {x X : f (x) = } and f 1 ([a, b)) for all 0 a < b < .
Proof: (i) (ii) It follows from {x X : f (x) } = X \ {x X :
f (x) > }.
(ii) (iii) It follows since {x X : f (x) < } = nN {x X : f (x)
n1 } and {x X : f (x) } = nN {x X : f (x) < + n1 }
(iii) (iv) It follows from {x X : f (x) } = X \ {x X : f (x) <
}.
29
30
Chapter 2. Measurability and Integrability

(i) = (v) Write {x X : f (x) = } = nN {x X : f (x) > n} .
Define R = {B B([0, )) : f 1 (B) }.
This is -algebra and contains (a, b] for all 0 a < b since
f 1 ((a, b]) = {x X : f (x) > a} {x X : f (x) b} .
(v) = (vi) It is obvious.

(vi) = (vii) Note that [0, b) is open in [0, ) and [a, b) = nn0 (a
1/n, b) for 0 < 1/n0 < a.
(vii) = (i) It follows since
{x X : f (x) > } = {x X : f (x) = }(n {x X : f (x) [+1/n, n)}).
Definition 2.1.3 Let (X, ) be a measurable space. A function f : X

[0, ) is called a simple function if it takes a finite number of different values.
P
In particular, f = ni=1 i Ai where Ai = f 1 ({i }) are disjoint sets and
i > 0 are different for i = 1, 2, ..., n.
Note that f is measurable if and only if Ai for i = 1, 2, ..., n..
Remark 2.1.1 Let us denote by S the set of simple measurable functions.
This is a vector space and also we have that if s1 , s2 S then s1 .s2 S,
max(s1 , s2 ) S and min(s1 , s2 ) S. In general limn sn
/ S for a sequence
P
(sn ) S, for instance f = n=1 n[n,n+1/2] is not simple, but it is limit of
simple functions.
Theorem 2.1.4 Let f : X [0, ] be a measurable function. There exists
a sequence sn of simple functions, such that 0 s1 s2 ... f and
limn sn (x) = f (x) for all x X.
Moreover, if f a bounded function then limn sn (x) = f (x) uniformly in
xX .
, k )), for k =
Proof: For each n N we denote by En,k = f 1 ([ k1
2n 2n
n
1
1, 2, ..., n2 and En,n2n +1 = f ([n, ]) and define
sn =
n +1
n2X
k=1
k1
En,k .
2n
2.1. Measurable functions
31
Let us show first that sn (x) is increasing for all x X. Fix n N and
x X.
If f (x) n + 1 then sn (x) = n sn+1 (x) = n + 1.
If n f (x) < n + 1 then sn (x) = n and sn+1 (x) = 2k1
n+1 for some
n+1
k n2
+ 1. Hence sn (x) sn+1 (x).
j
, k ) [ 2j1
If f (x) < n then there exist k, j so that f (x) [ k1
n+1 , 2n+1 ).
2n 2n
Hence either j1
= k 1 or j1
= k. In the first case sn (x) = k1
= 2j1
n+1 =
2
2
2n
j3
j1
k1
sn+1 (x) and in the second case sn (x) = 2n = 2n+1 2n+1 = sn+1 (x) .
Let us now show that lim sn (x) = f (x).
n
The case f (x) = we have sn (x) = n for all n N.
Assume f (x) < . Take n0 = n0 (x) N so that f (x) n0 . For all
n > n0 we have that there exists k {1, 2, .., n2n } such that f (x) En,k .
Hence for all n n0 we have
f (x) sn (x) = f (x)
1
k1
<
.
2n
2n
Note that if f is bounded n0 is the same for all x and the previous estimate
holds uniformly in x X. This finishes the proof.
Definition 2.1.5 Let (X1 , 1 ) and (X2 , 2 ) be measurable spaces. A function f : X1 X2 is called (1 , 2 )-measurable if f 1 (A) 1 for all A 2 .
For 2 = B(X2 ) where (X2 , ) is a topological space, a function f : X1
X2 is (1 , B(X2 ))-measurable (usually called simply measurable) if and only
if f 1 (A) 1 for all A .
If X and Y are topological spaces, f : X Y is said to be Borel measurable if f 1 (G) B(X) for all open set G X.
For X = Y = R a function f : R R is called Lebesgue measurable if it
is (M, B)-measurable, or equivalently f 1 ((a, b]) M for all a < b.
Remark 2.1.2 Given a measurable space (X, ) and a function f : X
[0, ]. Considering [0, ] as the compactification of the topological space
R+ then f is measurable (according to Definition 2.1.1) if and only if f is
(, B([0, ])-measurable (according to Definition 2.1.5).
Equivalently f is measurable if and only if X = {x X : f (x) = }
and f |X\X is (, B(R+ ))-measurable.
Let us establish some simple results about measurability on composition
of functions.
32
Lemma 2.1.6 (i) Let (X1 , 1 ), (X2 , 2 ) and (X3 , 3 ) be measurable spaces
and functions f : X1 X2 and g : X2 X3 . If f is (1 , 2 )-measurable
and g is (2 , 3 )-measurable then g f is (1 , 3 )-measurable.
(ii) Let (X1 , 1 ) be a measurable space,let X2 and X3 be topological spaces.
If f : X1 X2 is measurable and g : X2 X3 is continuous then g f is
measurable.
(iii) Let (X1 , 1 ) and (X2 , 2 ) be measurable spaces and let Y be a topological space. If f1 : X1 Rn and f2 : X1 Rm are measurable and :
Rn Rm Y is continuous then h : X Y given by h(x) = (f1 (x), f2 (x))
is measurable.
Proof: (i) and (ii) are immediate.
(iii) Consider : X Rn Rm given by (x) = (f1 (x), f2 (x)) and note
that
1 ((a1 , b1 ] ... (an , bn ] (a01 , b01 ] ... (a0m , b0m ])
= (f1 )1 ((a1 , b1 ] ... (an , bn ]) (f2 )1 ((a01 , b01 ] ... (a0m , b0m ]).
Since any open set in Rn Rm = Rn+m is numerable union of sets (a1 , b1 ]
... (an , bn ] (a01 , b01 ] ... (a0m , b0m ] we have that is measurable and then
h = too.
Let us recollect several operations on functions which are stable under
measurability.
Proposition 2.1.7 Let (X; ) be a measurable space and let 0 and
f, g, {fn } be measurable functions.
Then (with the convections 0 = 0, 1/0 = and 1/ = 0) f , f + g,
f.g, max{f, g}, min{f, g} 1/f , supn fn , inf n fn , lim supn fn and lim inf n fn
are measurable functions.
Proof: Write X = f 1 ({}), X 0 = g 1 ({}), f1 = f |X\X and
g1 = g|X\X 0 Hence
0
{f + g > } = X X
{f1 + g1 > },
{f g > } = {f = , g > 0} {g = , f > 0} {f1 g1 > }

Now, using (iii) in Lemma 2.1.6 we have that (f1 (x), f2 (x)) for (t, s) = t+s
or (t, s) = ts are measurable. Hence we obtain the measurability of f + g
or f g. The other cases follow from
{f > } = {f > }.
2.1. Measurable functions

{
33
1
1
> } = {f = 0} {0 < f < }.
f
{max(f, g) } = {f } {g }.
{min(f, g) > } = {f > } {g > }.
{sup fn } = n {fn }.
n
{inf fn > } = n {fn > }.

n
The limsup and liminf follows from the previous ones.

Remark 2.1.3 The supremum of measurable functions needs not be measurable.
Let E be the Vitali set in [0, 1] and write E = sup {x} .
xE
Definition 2.1.8 Let f : X R be a function, then f + = max(f, 0) and

f = min(f, 0).
Clearly, f + = f {f 0} , f = f {f 0} , f = f + f and |f | = f + + f .
Proposition 2.1.9 Let (X, ) be a measurable space and f : X R. The
following are equivalent.
(i) f is (, B(R))-measurable.
(ii) {x : f (x) > } for all R.
(iii) f + and f are measurable.
Proof: (i) = (ii) {x : f (x) > } = f 1 ((, )) .
(ii) = (iii) Let > 0 then {x : f + (x) > } = {x : f (x) > } and
{x : f (x) > } = {x : f (x) < } = X \ n {x : f (x) > 1/n} .
(iii) = (i) Let G be an open set. Then f 1 (G) = (f + )1 (G (0, ))
1
(f ) (G [0, )) .
Proposition 2.1.10 Let (X, ) be a measurable space and f : X C. The
following are equivalent.
(i) f is (, B(C))-measurable.
(ii) <f and =f are measurable.
(iii) |f | is measurable and there exists a measurable function : X
T = {z : |z| = 1} such that f = |f |.
34
Proof: (i) = (ii) It follows from composition since z <(z) and z

=(z) are continuous.
(ii)q= (iii) If follows from Lemma 2.1.6 that |f | is measurable, since
|f | = (<f )2 + (=f )2 . Take now E = {x : f (x) = 0} . Consider
z
: C \ {0} T given by (z) = |z|
and define (x) = (f (x) + E (x)). We
can apply Lemma 2.1.6 to get that is measurable and, of course, f = |f |.
(iii) = (i) Using : CC C given by (z1 , z2 ) z1 z2 we get that the
product of measurable functions is measurable and gives the result.
Let us list several operations which preserve measurability, whose proofs
are left to the reader.
Proposition 2.1.11 Let (X; ) be a measurable space and let K and
f, g, {fn } be measurable functions from X into K where K = R or K = C.
Then
(i) f , f + g, f.g, 1/f are measurable.
(ii) For K = R, max{f, g}, min{f, g} supn fn , inf n fn , lim supn fn and
lim inf n fn are measurable functions.
(iii) If limn fn (x) = F (x) for all x X then F is measurable.
Definition 2.1.12 Let (X, ) be a measurable space. A measurable function
f : X C is called a simple function if it takes a finite number of different
P
values. In particular, f = ni=1 i Ai where Ai = f 1 ({i }) are disjoint
sets and i are the non zero values for i = 1, 2, ..., n.
Corollary 2.1.13 Let f : X C. Then f is measurable if and only if
there exists a sequence (sn ) of simple functions such that |sn | |f | and
f (x) = lim sn (x) for all x X.
n
Proof: Since the limit of measurable functions is measurable we simply

need to prove the only if part. Assume f = u + iv be measurable and write
also u = u+ u and v = v + v . We can apply Theorem 2.1.4 and find
increasing sequences (rn ), (tn ), (pn ) and (qn ) of non-negative simple functions
converging to u+ , u , v + and v respectively.
The functions sn = rn tn +i(pn qn ) are simple complex-valued functions
and converge to f . Also we have
|sn | =
(rn tn )2 + (pn qn )2
(rn + tn )2 + (pn + qn )2
2.2. Some types of convergence.
2.2
35
(u+ + u )2 + (v + + v )2
u2 + v 2 = |f |.
Some types of convergence.
Definition 2.2.1 Let (X, , ) a measure space. We say that a property P

holds almost everywhere with respect to (X, , ) (in short -a.e.) if there
exists the set A (A) = 0 such that the property holds in X \ A.
Let us point out some facts to see the difference between complete and
not complete measure spaces regarding the behaviour a.e.
Remark 2.2.1 (X, , ) a measure space. Let f be a measurable function
and f = g -a.e. Then g needs not be measurable.
Indeed, the example can be produced for non complete measure space.
Consider (R, B, m) the Borel measure space on R. There exists A M \ B
with m(A) = 0 such that there is B B satisfying A B and m(B) = 0.
Now define f = 1 and g = R\B + 2B\A . We have that f = g m-a.e since
{f (x) 6= g(x)} = B, but g 1 ({0}) = A
/ B.
Remark 2.2.2 (X, , ) a measure space. Let fn be measurable functions
and limn fn = f -a.e. Then f needs not be measurable.
Indeed, let (R, B, m) be the Borel -algebra on R. As above tak A M\B
with m(A) = 0 and B B such that A B and m(B) = 0. Now define
fn = 0 and f = A . We have that limn fn = f m-a.e. since {f (x) 6= 0} B.
Proposition 2.2.2 Let (X, , ) be a complete measure space.
(i) If f : X [0, ]( or C) is measurable and f = g -a.e. then g is
measurable.
(ii) If fn : X [0, ]( or C) is a sequence of measurable functions and
lim fn = f -a.e. then f is measurable.
Proof: (i) Assume f : X [0, ] and f (x) = g(x) for x
/ A and A
with (A) = 0. Given > 0 we have that
{x : g(x) > } = {x A : g(x) > } {x
/ A : f (x) > }.
36
Since {x A : g(x) > } A then {x A : g(x) > } and also

{x
/ A : f (x) > } , which gives the result.
The case f : X C is done in a similar way.
(ii) Assume f (x) = lim fn (x) for all x
/ A and A with (A) = 0.
Denote X1 = X \ A and 1 the induced -algebra. The restrictions of fn
to X1 are measurable with respect to 1 and then f restricted to X1 is
measurable with respect to 1 . Define g(x) = f (x) for x X1 and g(x) = 0
for x A. Obviously g is measurable and f = g -a.e., hence using (i) we
get that f is measurable.

Theorem 2.2.3 Let (X, , ) be a measure space and (X, ,
) its complection. Then f : X [0, ] (respect. f : X C) is measurable with respect to
if and only if there exists g : X [0, ] (respect. g : X C) measurable
with respect to such that f = g -a.e.

Then
Proof: Assume that f : X [0, ] is measurable with respect to .
using Theorem 2.1.4 we get (sn ) simple functions measurable with respect to
and we can write, with s0 = 0,
f=
X
n=1
(sn sn1 ) =
ci Ei
i=1
for all i N.
where ci > 0 and Ei
We now choose Ai , Bi so that Ai Ei Bi and (Bi \ Ai ) = 0.
P
Define g =
i=1 ci Ai . It is measurable with respect to . If N = i (Bi \ Ai )
we have that f (x) = g(x) for x X \ N (since x
/ N implies that for each i
one gets x
/ Bi or x Ai ) and (N ) = 0.
The case f : X C follows from the previous one in the usual way and
it is left to the reader.
Assume now f = g -a.e. for some g : X [0, ] (respect. g : X C)
measurable with respect to . Using that g is also measurable with respect
and Proposition 2.2.2 we get that f is measurable with respect .
Corollary 2.2.4 f : R C is Lebesgue measurable if and only if there

exists g : R C Borel measurable such that f = g m-a.e.
Definition 2.2.5 Let (X, , ) be a measure space and B .
2.2. Some types of convergence.
37
A sequence (fn ) is said to converge uniformly in B to f if

lim sup |f (x) fn (x)| = 0.
n xB
A sequence (fn ) is said to converge almost uniformly to f if there exists

A with (A) = 0 such that fn converges to f uniformly in X \ A, i.e.
lim sup |f (x) fn (x)| = 0.
n xA
/
Remark 2.2.3 The almost uniform convergence is weaker than the uniform
convergence.
It suffices to take a sequence uniformly convergent to a function and modify the limit function in a set of measure cero to get an almost uniform limit
which is not uniform.
Remark 2.2.4 The pointwise convergence is weaker that the almost uniform
convergence.
Take fn = [n,n+1] in (R, M, m). Of course fn (x) converges pointwise to
zero, but for any set A M with m(A) = 0 we have that for each n N there
is xn [n, n + 1]
/ A (otherwise there exists k N such that [k, k + 1] A).
Hence supxA
/ |fn (x)| supn |fn (xn )| = 1.
Theorem 2.2.6 (Egorovs theorem) Let (X, , ) be a finite measure space.
Let fn : X [0, ] be finite a.e. measurable functions (i.e. ({fn = }) =
0) and let f (x) = limn fn (x) a.e is measurable and finite a.e. (i.e.({f =
}) = 0 and there exists C , (C) = 0 such that limn fn (x) = f (x), x
/
C). Then for any > 0 there exists A with (A) < such that fn
converges to f uniformly in X \ A.
Proof: We can assume that fn , f take values in R+ and that fn (x)
converges to f (x) for all x X. In other case, we put An = {fn = },
B = {f = } and C with (C) = 0 such that lim fn (x) = f (x) if x
/ C.
We work in X1 = X \ N where N = (n An ) B C.
Let us first prove that , > 0 there exists A, and n, N such
that (A, ) < and |fm (x) f (x)| < , for all x
/ A, and m n, .
Indeed, write
An () = {x X : |fm (x) f (x)| for some m n}.
38
It is a decreasing sequence An+1 () An () and n An () = .

Since (X) < then limn (An ()) = 0. Therefore there exist n0 =
n, N and A, = An0 () such that (An0 ()) < . Note that |fm (x)
f (x)| < if x
/ A, and m n0 .
Now, given k N and > 0, we apply the previous result for and
being = k1 and 2k to get Bk , nk N for which (Bk ) < 2k and
|fm (x) f (x)| < k1 for x
/ Bk and m nk . Finally take A = k Bk . We
have (A) < and fn converges uniformly in X \ A.
Indeed, given > 0 we first get k0 for which 1/k < for k k0 . Now for
all m nk0 and x
/ A this implies x
/ Bk and then |fm (x)f (x)| < 1/k < .
Remark 2.2.5 If (X) = then Egorovs theorem does not hold.
Take (R, B, m) and fn = [n,) . Clearly fn 0 pointwise, but if A
B and (A) < then A [n, ) 6= for all n N. This shows that
supxA
/ |fn (x)| 1 for all n N and (fn ) does not converge uniformly in A.
2.3
Integrable functions.
Definition 2.3.1 Let (X, , ) a measure space and let s : X [0, ] be a

simple function. We define
Z
sd =
n
X
i (Ei )
i=1
where i are the different non zero values of s.

R
We say that s is -integrable (or simply integrable) if X sd < , which
is equivalent to (Ei ) < for i > 0.
P
We use the convection 0 = 0 and then we can write ni=1 i (Ei ) even
when i = 0.
Let us see that actually the definition is independent of the decomposition
of s.
Proposition 2.3.2 Let s be a simple function with non zero values i for
P
P
i = 1, ..., n, say s = ni=1 i Ei . Assume that also s = m
j=1 j Fj for j 0
and Fj for j 1, ..., m. Then
Z
X
sd =
m
X
j=1
j (Fj ).
2.3. Integrable functions.
39
Proof: Assume first that Fj are pairwise disjoint. For each i 1, ..., n,
consider Mi = {j {1, ..., m} : j = i } Hence Mi are pairwise disjoint,
{1, ..., m} = ni=1 Mi and Ei = jMi Fj . Therefore
n
X
i (Ei ) =
i=1
=
=
n
X
i=1
n
X
(Fj )
jMi
j (Fj )
i=1 jMi
m
X
j (Fj )
j=1
Now we show the general case.

For s = 1 F1 + 2 F2 where F1 , F2 not necessarily disjoint.
We can write
s = (1 + 2 )F1 F2 + 1 F1 \F2 + 2 F2 \F1 .
Applying the previous case
Z
X
sd = (1 +2 )(F1 F2 )+1 (F1 \ F2 )+2 (F2 \ F1 ) = 1 (F1 )+2 (F2 ).
For s = 1 F1 + 2 F2 + 3 F3 where F1 , F2 , F3 not necessarily

pairwise disjoint.
We can write
s = (1 + 2 + 3 )F1 F2 F3
+ (1 + 2 )(F1 F2 )\F3 + (1 + 3 )(F1 F3 )\F2 + (2 + 3 )(F2 F3 )\F1
+ 1 F1 \(F2 F3 ) + 2 F2 \(F1 F3 ) + 3 F3 \(F1 F2 ) .
Applying the case of disjoint sets
Z
X
sd = (1 + 2 + 3 )(F1 F2 F3 )
+
+
+
+
=
(1 + 2 )((F1 F2 ) \ F3 )
(1 + 3 )((F1 F3 ) \ F2 )
(2 + 3 )((F2 F3 ) \ F1 )
1 (F1 \ (F2 F3 )) + 2 (F2 \ (F1 F3 )) + 3 (F3 \ (F1 F2 ))
1 (F1 ) + 2 (F2 ) + 3 (F3 ).
40
Now if s = ni=1 i Fi where Fi not necessarily pairwise disjoint. This

general case follows same argument with a rather more complicate notation.
Set Fi,j = Fi Fj for i 6= j and Fi1 ,i2 ,...,ij = Fi1 Fi2 ... Fij for different
i1 , ..., ij .
We can write
P
n
X
s =
i Fi \(j6=i Fj )
i=1
(i + j )Fi,j \(i0 ,j0 )6=(i,j) Fi0 ,j0
i6=j
(i1 + i2 + i3 )Fi1 ,i2 ,i3 \(i0 ,i0 ,i0 )6=(i

1 2 3
(i1 ,i2 ,i3 )
F0 0
1 ,i2 ,i3 ) i1 ,i2 ,i3
+ ....
n
X
+ (
i )F1 ...Fn .
i=1
Now using the result for disjoint sets and adding up the measures the
result is complete.
Definition 2.3.3 Let (X, , ) a measure space and let f : X [0, ] be a
measurable function. We define
Z
f d = sup{
sd : 0 s f, s simple }.
We say that f is -integrable

if X f Rd < .
R
For each E we define E f d = X f E d.
R
Proposition 2.3.4 Let f, g : X [0, ] be measurable functions, > 0

and E, F .
R
R
(i) If f g then XR f d XR gd.
(ii) If E F then E f d F gd.
(iii) If s is a simple function, E then
Z
sd =
(iv)
R
X
n
X
i (E Ei ) = sup{
i=1
f d =
R
X
f d.
Proof: (i) and (iv) are obvious.

(ii) and (iii) follow from (i).
td : 0 t s : t simple }.
41
Theorem 2.3.5 (The Lebesgue monotone convergence theorem) Let fn :

X [0, ] be measurable functions such that fn (x) fn+1 (x) and limn fn (x) =
f (x) for all x X. Then
Z
f d = lim
n X
fn d.
Proof: Using Proposition 2.3.4 we have

Z
X
fn d
Z
X
fn+1 d
f d.
Define M = supn XR fn d = limn X fn d.

R
Of course M X f d. Let us show that X f d M .
P
We may assume M < . For each simple function s = ni=1 i Ei such
that 0 s f and each 0 < c < 1 we consider the sequence
An = {x X : fn (x) cs(x)}
which form an increasing sequence of measurable sets in .
Observe that X = n An . Indeed, there exists n0 so that fn (x) cs(x)
for all n n0 . If s(x) = 0 we can take n0 = 1 and if s(x) > 0 then
f (x) = limn fn (x) > cs(x) and we can apply the definition of limit.
Since
Z
Z
Z
fn d M,
fn d
csd
X
An
An
and lim
(An Ei ) = (Ei ) and therefore limn An csd = X csd we get
R n
that
X csd M . Since this holds for all s f and 0 < c < 1 we have
R
X f d M .
R
Corollary 2.3.6 (The Fatou Lemma) Let fn : X [0, ] be measurable

functions. Then
Z
Z
(lim inf fn )d lim inf
fn d.
X
Proof: Define gn = inf{fk : k n}. They are measurable functions,

gn gn+1 and gn fn for all n N.
Applying Theorem 2.3.5 we have
Z
X
(lim
inf fn )d =
n
( lim gn )d = n
lim
X n
Z
= lim inf
n
gn d
XZ
gn d lim inf
n
fn d.
42
Proposition 2.3.7 Let f, g : X [0, ] be measurable functions, , > 0

and E,RF .
R
R
(i) X (f + g)d = RX f d + RX gd. R
(ii) If E F = then
d = E f d + F f d.
EF f
R
R
(iii) If f g then X f d X gd.
1 R
(iv) ({x X : f (x) >
})
X f d for all > 0.
R
(v) IfR (E) = 0 then E f = 0.
(vi) X f d = 0 if and only if f = 0 -a.e.
(vii) If f is -integrable then ({x X : f (x) = }) = 0.
Proof: (i) follows from Proposition 2.3.2 for simple functions. The general
case then follows by combining Theorem 2.1.4 and the Lebesgue monotone
convergence theorem.
(ii) Note that f EF = f E + f F and apply (i).
(iii) For each s simple function with 0 s f one has 0 s g. Then
the result follows from the definition.
(iv) Note that E f E where E = {x X : f (x) > } and
integrate.
R
P
(v)R Let s simple with 0 s f . Then E sd = ni=1 i (E Ei ) = 0.
Then E f d = 0.
(vi) Assume f = 0 -a.e. Hence ({x : f (x) > 0} = 0 and then, using
(iv) we get
Z
Z
Z
f d = 0
f d +
f d =
{f >0}
{f =0}
Conversely, assume ({x : f (x) > 0} > 0. Since {x : f (x) > 0} =

nN {x : f (x) > n1 } we have ({x : f (x) > n10 }) > 0 for some n0 . Hence
Z
X
f d
Z
{f > n1 }
0
f d >
1
1
({x : f (x) > }) > 0.
n0
n0
This gives the direct implication.

Conversely, assume f = 0 -a.e. then if E0 = {xR : f (x) >R0} we have
(E0 ) = 0 and f = f E0 . Now using (iv) we get that X f d = E0 f d = 0.
(vii) The set {x X : f (x) = } = nN {x X : f (x) > n}. Hence
(iii) gives that
1Z
f d
({x X : f (x) = })
n X
for all n N. Passing to the limit we get ({x X : f (x) = }) = 0.
43
Corollary 2.3.8 Let fn : X [0, ] be measurable functions. Then

Z
X n=1
fn )d =
Z
X
n=1 X
Proof: Apply Theorem 2.3.5 for gn =
Pn
k=1
fn d.
fk .
Proposition 2.3.9 Let f : X [0, ] be a measurable funtion.

Define
Z
(E) =
f d, E .
E
(i) is a measure on .
(ii) is finite if and only if f is -integrable.
(iii) If (E) = 0 then (E) = 0.
(iv) If is finite then lim (E) = 0.
(E)0
Proof: (i) Let {En } be a sequence of pairwise disjoint measurable sets.

P
Since f En =
n=1 f En we have that
(nN En ) =
Z
X
Z
X
f En d =
n=1 X
f En d =
(En ).
n=1
(ii) It is obvious.
(iii) follows from (iv) in Proposition 2.3.7.
(iv) Note that lim (E) = 0 means that for all > 0 there exists > 0
(E)0
so that (E) < implies (E) < .

If (X) < then for each > 0 there exist a simple function s such that
Z
X
Now for s =
Pn
i=1
Z
E
Take < 2 Pn
i=1 i
f d
sd +
i Ei we have that
sd =
n
X
n
X
i (Ei E) (
i=1
. Now if (E) < then
i )(E).
i=1
R
E
f d < .
44
Definition 2.3.10
Let f : X C be a measurable function. f is said to be
R
-integrable if X |f |d < .
Moreover if f = u + iv is -integrable we define
Z
f d =
u+ d
u d + i
v + d i
v d.
Remark 2.3.1 In particular

Z
f d =
ud + i
vd.
Lemma 2.3.11 Let f : X R be a measurable function f = f1 f2 for

some fi ; X R+ which are -integrable. Then f is -integrable and
Z
f d =
f1 d
Z
X
f2 d.
Proof: Using that |f | |f1 | + |f2 | one gets the integrability of f .

Now notice that f + f = f1 f2 . Hence f + + f2 = f + f1 and using
the linearity of the integral for non-negative functions one obtains
Z
f + d +
Therefore
f2 d =
f d =
f d +
Z
X
f1 d
Z
X
f1 d.
f2 d.
Proposition 2.3.12 Let f and g be -integrable functions and let , be

complex numbers. Then f + g is -integrable and
Z
(f + g)d =
f d +
gd.
Proof: Since |f + g| |||f | + |||g| and

Z
|f + g|d ||
|f |d + ||
|g|d < .
It suffices to show the result for real valued functions. The case of complex
values follows immediately from the previous one using that if f = u + iv
then
Z
Z
Z
f d =
ud + i vd.
X
45
Hence for f = u + iv, g = u0 + iv 0 , = a + ib and = a0 + ib0 one has

f + g = (au bv) + (a0 u0 b0 v 0 ) + i(av + bu) + i(a0 v 0 + b0 u0 ).
Now use the result for real valued functions to get
Z
(f + g)d =
+ i
(au bv)d +
X
Z
(av + bu)d + i
(a0 v 0 + b0 u0 )d
ud b
ZX
+ ia
(a0 u0 b0 v 0 )d
XZ
= a
vd + ib
vd + a
u d b
ud + ia
= (a + ib)
v 0 d
vd + ib
(u + iv)d + (a0 + ib0 )
ud
(u0 + iv 0 )d.
Assume first that R and f : X R is -integrable. Since g + = 21 (|g|+g)

and g = 21 (|g| g) for any real-valued function, f = (+ )(f + f )
and |f | = (+ + )(f + + f ) then we conclude that
(f )+ = + f + + f ,
(f ) = + f + f + .
These give
Z
(f )d =
+ f + + f d
ZX
+ f + f + d
= +
(f + f )d
(f + f )d
f d.
Let f, g : X R be -integrable. Then f + g = (f + + g + ) (f + g )

and we can apply Lemma 2.3.11 to conclude that
Z
X
(f + g)d =
f d +
Proposition 2.3.13 If f is -integrable then |
gd.
R
X
f d|
R
X
|f |d.
46
Proof: Assume that f : X R, and write f = f + f . Then

|
f d| = |
f d
f d|
f d| =
f d =
f d +
In the general case, assume

observe that
|
f d =
|f |d.
X
|z|
z
f d = z C \ {0}. Take =
f d =
<(f )d
|f |d =
and
|f |d.
Theorem 2.3.14 (The Lebesgue dominated convergence theorem) Let fn :

X C a sequence of measurable functions and f (x) = limn fn (x) for all
x X. If there exists g 0 -integrable such that |fn (x)| g(x) for all
x X then f is -integrable and
lim
n X
In particular limn
R
X
fn d =
|fn f |d = 0.
R
X
f d.
Proof: Since |f | g we have that f is -integrable. Observe now that

|fn f | 2g and write hn = 2g |fn f |. Using Fatous Lemma we have
Z
2gd lim inf
Z
X
hn d =
Z
X
2gd lim sup
|fn f |d.
This shows that limRsup |fn R f |d = 0. Now using Proposition 2.3.13 we

obtain that limn X fn d = X f d.
R
Corollary 2.3.15 (Bounded convergence theorem) Let (X, , ) be a finite

measure space. Let fn be a sequence of measurable functions such there exists
M > 0 for which |f
(x)| M for all n k and x X. If limn fn = f then
R
Rn
f
d
=
lim
n X fn d.
X
Proof: Note that constant functions are -integrable for finite measures .
So the result follows from the dominated convergence theorem.
Remark
2.3.2 If (X) = then supn |fn | < and limn fn = f do not
R
imply X f d = limn fn d.
Take fn = n1 (0,n) . We have that fn 1, fn 0 as n but
R
(0,) fn dm = 1 for all n N.
47
Definition 2.3.16 We define the equivalence relation f g if f = g a.e. and we denote by L1 () the set of equivalence classes of -integrable
functions.
R
Let us denote ||f ||1 = X |f |d.
Proposition 2.3.17 (L1 (), ||.||1 ) is a Banach space.
Proof: Let us first show that L1 () is a vector space.
This actually shows that ||f + g||1 ||f ||1 + ||g||1 .
Obviously ||f ||1 = ||||f ||1 .
R
The other property to get a norm is that X |f |d = 0 implies that f = 0
-a.e.
P
Let us show the completeness. It is equivalent to show that
n=1 ||fn ||1 <
P
implies that the series n=1 fn converges in L1 ().
R P
P
Now, using that
n=1 |fn |d < we get that the series
n=1 ||fn ||1 = X
P
|f
|
<
is
convergent
-a.e,
and
hence there exists A such that
n=1 n
P
for all x
/ A.
n=1 fn (x) is convergent
P
/ A andR f (x) = 0 for x A. Observe
Define f (x) = n=1 fn (x) forR x
P
first that f is -integrable, since X |f |d X
n=1 |fn |d < .
Now
lim
(
n
n
X
fk (x) f (x)) = 0, x
/ A and
k=1
n
X
fk (x) f (x)| 2|f (x)|, x

/ A.
k=1
Therefore, from the Lebesgue dominated convergence theorem,

lim
n X
n
X
k=1
fk f |d = lim
n
X
n X\A
k=1
fk f |d = 0.
Corollary 2.3.18 Simple integrable functions are dense in L1 ().

Proof: Using Proposition 2.1.13 we get that if f is -integrable there exists
a sequence of simple functions such that sn f pointwise
and |sn | |f |.
R
Now, the dominated convergence theorem gives limn X |f sn | = 0.
48
PropositionZ2.3.19 Let f : X C be -integrable. Then

(i) lim
|f |d = 0, i.e. for all > 0 there is > 0 such that (E) <
(E)0 E
R
implies E |f |d < .
R
(ii) For all > 0 there exists B with (B) < and X\B |f |d < .
Proof:
(i) Note that, using the dominated Lebesgue theorem, for each > 0
there exist a simple function s such that
|f s|d .
2
X
Now for s =
Pn
i=1
i Ei we have that
|s|d
n
X
|i |(Ei E) ( sup |i |)(E).

i=1,...n
i=1
Hence if < 2 sup |i | and (E) < we get that E |f |d < .

i=1,...n
(ii) Take sn simple functions such that sn converges to f and |sn | |f |
and apply
the dominated convergence theorem. Given > 0 there is n0 so
R
Pn
that X |f sn0 |d < . Take B = m
i=1 Ei where sn0 =
i=1 i Ei . One has
that (B) < since sn0 is -integrable and
R
|f |d
X\B
Z
X\B
|f sn0 |d < .
Definition 2.3.20 A sequence {fn } of -integrable functions is called equiintegrable if for every > 0
(a) there exist > 0 such that implies that
(E) < = sup
nN
Z
E
|fn |d <
(b) and there exists B with (B) < and

sup
nN
Z
X\B
|fn |d < .
Remark 2.3.3 In particular if |fn | g for some -integrable function g the

sequence {fn } is equi-integrable from (iii) and (iv) in 2.3.19.
49
Next theorem provides a converse to the Dominated convergence theorem.

Theorem 2.3.21 (Vitalis theorem) Let fn be a sequence of -integrable
functions converging to f . The following are equivalent:
(i) {fn } is equi-integrable
R
(ii) f is -integrable and limn X |fn f |d = 0.
Proof: (i) =)(ii) Given 0 < < 1 there exist > 0 and B with
(B) < such that
sup
X\B
|fn |d < ,
sup
|fn |d < whenever (E) < .
Hence, using Fatous

lemma, X\B |fR |d lim inf n X\B |fn |d .
R
Therefore X |f fn |d 2 + B |f fn |d.
Now from Egorovs theorem (since (B) < ) we can find A B, A
and (A) < such that fn converges to f uniformly in B \ A.
Putting everything together we obtain
R
Z
B
|f fn |d
Z
B\A
Z
B\A
|f fn |d +
Z
A
|f fn |d
|f fn |d + sup
n
Z
A
|f fn |d
Applying Fatous Lemma again and (A) < we can say that
Z
B
|f fn |d
Z
B\A
|f fn |d + 2.
Using that limn supxB\A |fn (x) f (x)| = 0 we have |fn f | M B\A
for some constant
M > 0. So we can use the bounded convergence theorem
R
and get limn B\A |f fn |d = 0.R
Finally this shows
that limRn X |f fn |d =
0. This also gives that f is
R
R
-integrable Rsince X |f |d X |f fn |d + X |fn |d < after choosing
n such that X |f fn |d < 1.
(ii) =(i) Note that
|
Z
E
fn d|
Z
E
|fn f |d + |
f d|
for all n
N and E . Hence for each > 0 there
exists n0 N, 0 > 0 for
R
R
which X |fn f |d < /2 for all n n0 and | E f d| < /2 if (E) < 0 .
50
For each for k n0 , Proposition 2.3.19 provides k > 0 such that

| E fk d| < /2 if (E) < k . Now take < min{k , 0 } to conclude part (a)
in the definition of equi-integrability.
On the other hand,
from Proposition 2.3.19 again we choose B0 with
R
(B0 ) < and R X\B0 |f |d < /2 and Bk for 1 k n0 such that
0
Bk . Therefore
(Bk ) < and X\Bk |fk |d < . Denote B = nk=0
R
sup
nn0
Z
X\B0
|fn |d sup
nn0
and
sup
n
2.4
Z
X\B
Z
X\B0
|f fn |d + /2 <
|fn |d < .
Exercises
Exercise 2.4.1 Let (X, ) be a measurable space y (An ) be a sequence of

measurable sets such that nN An = X.
i) Given a functions f defined on X such that fn = f An is measurable
with respect to the -algebra n = {E An : E } for all n N. Show
that f is - measurable.
ii) Assume An are pairwise disjoint and let fn be n -measurable functions
defined on An . If f is defined on X in such a way that f (x) = fn (x) for
x An . Show that f is - measurable.
iii) Assume An is increasing sequence of measurable sets and let fn be
n -measurable functions defined in An and such that fn (x) = fn+1 (x) for
x An . If f is defined in X such that f (x) = fn (x) for x An . Show that
f is -measurable.
Exercise 2.4.2 Let (X, ) be a measurable space, S and S the algebra induced over S. Let f be a map from X into a topological space
Y and let y Y . Show that f is S -measurable if and only if f given by
f = f S + yX\S is -measurable.
Exercise 2.4.3 Let (X, ) be a measurable space and let f : X [0, ] be
P
-measurable. Show that f =
n=1 cn An for certain cn 0 and An .
2.4. Exercises
51
Exercise 2.4.4 Let (Y, ) be a topological space and (f )J a family of maps

from X into Y . Show that there exists a minimum -algebra over X such
that f are measurable and get its description. (It is called the -algebra
generated by (f ) and denoted (f : J).)
Exercise 2.4.5 Let (R, ) be a measurable space where = {B R :
B numerable or X\B numerable}. Find a characterization of the -measurable
functions f : R R.
Exercise 2.4.6 Let (X, ) be a measurable space and f, g : X [0, ] be
-measurable. Show that the set {x X : f (x) = g(x)} belongs to .
Exercise 2.4.7 Show that the set of points where converge a sequence of
complex measurable functions is measurable.
Exercise 2.4.8 Let I be an open interval in R and f : I R.
i) If f is piecewise monotone then f is measurable.
ii) Assume f is derivable. Show that f 0 is Borel-measurable.
Exercise 2.4.9 Let (X, ) be a measurable space and f : X Rn given by
f = (f1 , f2 , ..., fn ).
Show that f is -measurable if and only if fi are -measurable for all
1 i n.
Exercise 2.4.10 (i) Show that f : R R is Lebesgue-measurable if and only
if f 2 is Lebesgue-measurable and {x : f (x) > 0} is a Lebesgue-measurable set.
(ii) Give a non-measurable function f such that f 2 is measurable.
Exercise 2.4.11 Let (X, , ) a measure space, A , A the -algebra
induced over A and A measure concentrated on A.
(i) Let f be -measurable (from X into either
[0, ] or C).
R
Show that
f is A -integrable
if and only if A |f |d < . Moreover if
R
R
E then E f dA = EA f d.
(ii) Let f a A -measurable function and f0 = f A the -measurable
extension defined on X.
Show that Rf is A -integrable
if and only if f0 is -integrable. Moreover if
R
E A then E f dA = EA f0 d.
Exercise 2.4.12 Let (X, ) be a measurable space and let (n ) be a sequence
P
of measures on it. Define (E) = nN n (E).
Describe the -integrability in terms of the n -integrability.
52
Exercise 2.4.13 Let (X, , ) be a measure space and let : X Y be

a map. Show that g : Y R is ()-integrable if and only if g is
-integrable. Moreover
Z
g d =
gd().
Exercise 2.4.14 Let (X, ) a measurable space, a X and a the Dirac

mass concentrated in a.
(i) Describe the a -integrability and find the integral with respect to a .
P
n
n .
(ii) Let be defined over P(N) given by =
n=1 2
Describe the -integrability and the integral with respect .
R
Calculate N f d para f (n) = n.
Exercise 2.4.15 Let (X, , ) be a measure space and let fn : X [0, ]
be measurable functions. Show that
R
R
P
i) X (supkN fk )d kN X fk d.
ii) If fj1 .fj2 ...fjn+1 = 0 for any (j1 , j2 , ...jn+1 ) then
XZ
kN X
fk d n
Z
X
(supkN fk )d.
What does it mean for fk = Ak .?

Exercise 2.4.16 Let f : Rn [0, ) Rbe Lebesgue-measurable. let be the
measure defined on M(Rn ) by (A) = A f (x)dx. Show that g : Rn C
is
-integrable
if and only if g f is Lebesgue integrable. In such a case,
R
R
gd = g(x)f (x)dx.
Exercise 2.4.17 Study the -integrability of f , calculating its integral whenever it does exist:
a) (Z, P(Z), ), the counting measure and f (n) = e|n| for n Z.
n
for n N.
b) (N, P(N), ), the counting measure and f (n) = (1)
n+1
P
x
c) ((0, ), B((0, )), ), d(x) = e dx and f = n=1 n[n1,n) .
d) (R, B(R), ), d(x) = |sen x|dx and f (x) = x1 R\{0} .
e) (R2 , B(R2 ), ), = (m) where (t) = (e|t| cos t, e|t| sen t) and
f (x, y) = xy.
f ) ([0, 2 ], B([0, 2 ]), m) and f (x) = sen x[0, 2 ]Q + cos x[0, 2 ]R\Q .
g) ([0, 2 ], B([0, 2 ]), m) and f (x) = sen x[0, 2 ]{x:cos xQ} +sen2 x[0, 2 ]{x:cos
xR\Q} .
2.4. Exercises
53
Exercise 2.4.18 Study the mF -integrability de f for the Lebesgue-Stieltjes

measure mF , computing its integral when possible:
a) I = (0, ), F (x) = (x 1)+ , f (x) = x ( R).
b) I = (0, 1), F (x) = [ x1 ], f (x) = x ( R).
P
1
1
1 (x), f (x) = x.
c) I = (0, 1), F (x) =
n=1 n [ n+1
,n
)
R
d) I = R, F (x) = (0,) |sen t|dt, f (x) = x1 R\{0} .
P
e) I = (0, ), F (x) = ex , f (x) =
n=1 n(n1,n) (x).
f ) I = (0, ), F (x) = log x, f (x) = x (0,1) + x (1,) , (, > 0).
Exercise 2.4.19
Find the following limits:
Z
x
nx
e cos nxsen dx.
a) lim
n Z 0
n
x
2
n
b) n
lim
e x dx.
Z1n
e n x 2 log xdx.
n 0
Z 1
n
sen(
))dx.
d) lim
n 0
2n + x
Z n
x
e) lim
eax (1 + )n dx, (a > 0).
n 0
n
1
t
1X
f ) lim+
arctag .
t0 t
n
n=1 n
c) lim
Exercise 2.4.20 Let (X, , ) be a measure space and f, g : X R integrable

functions. Discuss whether or not the following functions are also -integrable.
(Give conditions to get afirmative answers and also counterexamples for the
negative ones).
q
q
q
1
f
1
),
|f |2 + |g|2 , 1+|g|
,
f 2 , f 3 , arctag f , |f | + |g|, f.g, f g, sen( 1+|f
|
|f | ( R).
Exercise 2.4.21 Let (X, , ) be a measure space and let f : X [0, ] be
integrable. Show that {x X : f (x) > 0} is a countable union of sets of
finite measure.
Exercise 2.4.22 Give an example where the inequality of Fatous lemma is
strict.
Exercise 2.4.23 Let ((, ], B((, ]), m) be the Lebesgue measure space
over B((, ]). Consider T = {z C : |z| = 1}, B(T) and the measure
= (m) where (t) = eit . Show that
54
(i) is an invariant under rotations probalility measure, i.e. ((T) = 1)

and (A) = (A) for || = 1.
(ii) A measurable function f : TR C is -integrable
if and only if g(t) =
1 R
f ((t)) is m-integrable. Moreover T f d = 2 gdm.
Exercise 2.4.24 Let be a C 1 -curve defined in Rn and let : [a, b] Rn
be a parametrization of . On the borelians of we define the measure m
given by m = () where corresponds to the measure over [a, b] with
density ||0 ||, i.e. d = ||0 ||dm.
(i) Show that m is independent of the chossen parametrization.
(ii) Characterize the m -integrability and show that for integrable functions one has
Z
Z b
f dm =
f ((t))||0 (t)||dt.
Exercise 2.4.25 Let s C and f : (0, ) C given by f (x) = xs1 ex .

Show that f is integrable
if and only if <s > 0.
R
(Recall that (s) = (0,) xs1 ex dx).
Exercise 2.4.26 Let a, s C and f : (0, ) C given by f (x) = xs eax .
(i) Find the values of a and s for f to be integrable.
Find the integral in the cases a > 0 and integrable f .
(ii) Show that if <s > 1 then
X
1
1 Z
xs1
=
dx.
(s) =
s
(s) (0,) ex 1
n=1 n
2
Exercise 2.4.27 Let z C and f : (0, ) C given by f (t) = cos ztet .

Show that f is integrable for all z C and compute its integral.
Exercise 2.4.28 Prove, justifying the computations, that
Z
x dx =
nn .
n=1
Exercise 2.4.29 Show, using the image measure, the following well-known
change of variable result. Let X : R R be of class C 1 , strictly increasing
and bijective and let g : R R be integrable. Then
Z
R
(g X)(t)dt =
Z
R
g(x)(X 1 (x))0 dx.
2.4. Exercises
55
Exercise 2.4.30 Let f : X [0, ] be measurable and let be a finite

P
measure. Show that f is -integrable if and only if n ({x X : f (x)
n}) < .
Exercise 2.4.31 Let f, fn be non-negativeintegrable functions such that
a) lim Rfn = f a.e.
R
b) lim fn d =R f d
Show that lim |fn f |d = 0.
Chapter 3
The product measure and
Fubinis theorem
3.1
The product measure
Definition 3.1.1 Let (X, 1 ) and (Y, 2 ) be measurable spaces. We define

R = {A B : A 1 , B 2 } the family of measurable rectangles and
A = {E = nk=1 Ak Bk : Ak Bk R, (Ak Bk ) (Al Bl ) = , k 6= l}
the elementary sets in the product X Y .
We denote by 1 2 = (A) = (R) the product -algebra over X Y .
Proposition 3.1.2 The family A is an algebra over X Y .
Proof: Of course A.
Let E R, say E = A B where A 1 and B 2 . Note that
(X Y ) \ (A B) = ((X \ A) B) (X \ A) (Y \ B) (A (Y \ B)) A.
Observe now that the intersection of two rectangles is a rectangle, since
(A B) (A0 B 0 ) = (A A0 ) (B B 0 ), what gives that
(X Y ) \ (nk=1 (Ak Bk )) = nk=1 (X Y ) \ (Ak Bk ) A.
Finally any finite union of pairwise disjoint elementary sets is elementary.
Using ni=1 Ai = i=1 (Ai \ i1
j=1 Bj ) one concludes the result.
57
58
Chapter 3. Product Measure
Proposition 3.1.3 Let n = k + l, k, l N. Then B(Rn ) = B(Rk ) B(Rl ).

Proof: Recall that B(Rn ) = (E) where
E = {(a1 , b1 ] ... (an , bn ] : ai < bi , i = 1, 2, ..., n}
and B(Rk ) B(Rl ) = (R) where
R = {A B : A B(Rk ), B B(Rl )}.
Hence it suffices to see that E B(Rk ) B(Rl ) and R B(Rn ).
Obviously E R.
Now, given A B R we have
A B = (A Rl ) (Rk B).
We shall only see that A Rl B(Rn ) for all A B(Rk ). Define
= {A B(Rk ) : A Rl B(Rn )}.
Clearly it is a -algebra and contains the open sets, so the proof is complete.
Definition 3.1.4 Let E be a subset of X Y , x X and y Y . We

call x-section (respect. y-section) of E the sets Ex = {y Y : (x, y) E}
(repect. E y = {x X : (x, y) E}.)
Let f : X Y Z be function, x X and y Y . We call the x-section
(respect. y-section) of f the functions fx : Y Z (respect. f y : X Z)
where fx (y) = f y (x) = f (x, y).
Proposition 3.1.5 Let Z be a topological space or Z = [0, ], E X Y ,
x X and y Y . Then
(i) if E 1 2 then Ex 2 and E y 1 ,
(ii) if f : X Y Z is 1 2 -measurable then fx is 2 -measurable
and f y is 1 -measurable.
Proof: (i) Let = {E 1 2 : Ex 2 }. It is clear that (n En )x =
(En )x and ((X Y ) \ E)x = Y \ Ex . Hence is -algebra.
On the other hand, if E = AB R then Ex = B for x A and Ex =
for x
/ A. Hence R . This shows that = 1 2 .
3.1. The product measure
59
Similarly the case of y-sections.

To see (ii) note that, for each open set G Z,
{y Y : fx (y) G} = ({(x, y) X Y : f (x, y) G})x
and apply part (i).
Theorem 3.1.6 Let (X, 1 , ) and (Y, 2 , ) be finite measure spaces and
let E 1 2 . Then
(i) xR (Ex ) is R1 -measurable and y (E y ) is 2 -measurable, and
(ii) X (Ex )d = Y (E y )d.
Proof: Case 1.- Assume E = A B R. Obviously
(E y ) = (A)B (y),
(Ex ) = (B)A (x),

and
(A)(B) =
Z
X
Case 2.- E =
nk=1 Ak
(Ex ) =
(Ex )d =
Z
Y
Bk where Ak Bk are pairwise disjoint. Then
n
X
(E y ) =
(Bk )Ak (x),
n
X
n
X
(Ak )Bk (y),
k=1
k=1
and
(E y )d.
(Ak )(Bk ) =
k=1
Z
X
(Ex )d =
(E y )d.
In general, define
= {E 1 2 : satisfies (i) and (ii)}.
We shall prove that is a monotone class. Since it contains A we will have
M(A) = = 1 2 and the proof will be finished.
Let {En } be an increasing sequence in . Taking into account that
((En )x ) (respect. ((En )y )) are increasing sequences of 2 -measurable (respect. 1 -measurable) functions and converges to ((n En )x ) (respect. to
((n En )y )) we get that n En verifies (i).
Using also the monotone convergence theorem we have that
lim
n
Z
X
(Ex )d =
Z
X
((n En )x )d,
60

lim
n
(E )d =
((n En )y )d.
Hence Y ((n En ) )d = X ((n En )x )d and n En .

Let {En } be an decreasing sequence in . Taking into account that
((En )x ) (respect. ((En )y )) are decreasing sequences of 2 -measurable (respect. 1 -measurable) functions and, using that and are finite, converges
to ((n En )x ) (respect. to ((n En )y )) we get that n En verifies (i).
Since ((En )x ) (Y ) for all x X and all n N (respect. ((En )y )
(X) for all y Y and all n N), then, using the bounded convergence
theorem we have that
R
lim
n
lim
n
Hence
R
Y
(Ex )d =
y
(E )d =
Z
X
((n En )y )d =
R
X
((n En )x )d,
((n En )y )d.
((n En )x )d and n En .
Corollary 3.1.7 Let (X, 1 , ) and (Y, 2 , ) be -finite measure spaces and
let E 1 2 . Then
(i) xR (Ex ) is R1 -measurable and y (E y ) is 2 -measurable.
(ii) X (Ex )d = X (E y )d.
Proof: Write X =
n=1 Xn with Xn 1 for all n N, (Xn ) < and
0
Xn Xn = if n 6= n0 , and Y =
m=1 Ym with Ym 2 for all m N,
(Ym ) < and Ym Ym0 = if m 6= m0 .
Let n, m N and define n (A) = (A Xn ) and m (B) = (B Ym ).
Since (X, 1 , n ) and (Y, 2 , m ) are finite measure spaces, then x m (Ex )
is 1 -measurable and y n (E y ) is 2 -measurable and also
Z
Xn
(Ex Ym )d =
Z
Ym
(E y Xn )d.
y
y
Now (Ex ) =
n=1 n (E ) and hence x
m=1 m (Ex ) and (E ) =
y
(Ex ) is 1 -measurable and y (E ) is 2 -measurable.
Moreover
Z
X
(Ex )d =
=
Z
X
m=1 X
X
m (Ex )d
Z
m=1 n=1 Xn
(Ex Ym )d
3.1. The product measure
61
X
Z
X
n=1 m=1 Ym
Z
X
y
n=1 Y
(E y Xn )d
(E Xn )d
n (E y )d.
Definition 3.1.8 Let (X, 1 , ) and (Y, 2 , ) be -finite measure spaces.

We define the measure
(E) =
Z
X
(Ex )d =
(E y )d
for all E 1 2 .
Then (X Y, 1 2 , ) is a -finite measure space.
Remark 3.1.1 (A B) = (A)(B) for all A 1 and B 2 .
Remark 3.1.2 Let E 1 2 . Then
Z
XY
E (x, y)d =
Z
X
Ex d)d(x) =
Z
Y
E y d)d(y).
Remark 3.1.3 The -finiteness is necessary for the iterated integrals to coincide.
Indeed, let X = Y = [0, 1], = m be the Lebesgue measure and let be
the counting measure. Take E = {(x, x) : 0 x 1}. Note that Ex = {x}
and E y = {y},R and hence (E
) = 1 and (E y ) = 0, for all x, y [0, 1].
R x
Therefore 1 = X (Ex )d 6= Y (E y )d = 0.
Definition 3.1.9 In the case X = Y = R, 1 = 2 = B(R) and =
= m the Lebesgue measure on B(R), we define the Lebesgue measure on
B(R) B(R) = B(R2 ) as the product measure m m and it is denoted m2 .
Inductively mn = m mn1 is defined on B(Rn ) for all n N. It follows
from the unicity of the Hahn theorem that mn = mk ml for any n = k + l.
Remark 3.1.4 The space (R2 , B(R2 ), m2 ) is a -finite not complete measure
space.
Indeed, let E and C be the Vitali and the Cantor sets in [0, 1] respectively.
Then E C
/ B(R2 ) (due to the fact (E C)y = E for any y C) but
E C [0, 1] C and m2 ([0, 1] C) = 0.
62
Definition 3.1.10 We denote by (R2 , M(R2 ), m2 ) the complection of the

space (R2 , B(R2 ), m2 ) and it is called the Lebesgue measure space in R2 .
Similarly (Rn , M(Rn ), mn ) for all n N.
Remark 3.1.5 Due to the fact that m m is not complete in M(R) M(R)
(see Remark 3.1.4) then M(R) M(R) 6= M(R2 ).
Theorem 3.1.11 Let T : Rn Rn be a linear transformation and A
M(Rn ). Then T (A) M(Rn ) and mn (T (A)) = |det(T )|mn (A).
Proof: Case 1: T is a bijection and that A B(Rn ).
Note that, since kT (x) T (y)k kT kkx yk, one get the continuity.
Define now = {A B(Rn ) : T (A) B(Rn )}. Clearly it is a -algebra
which contains the open sets. Hence = B(Rn ).
On the other hand (A) = mn (T (A)) defines a measure on B(Rn ) such
that ((a1 , b1 ] ... (an , bn ]) = mn (T ((a1 , b1 ] ... (an , bn ]))
Note that (a1 , b1 ]...(an , bn ] = (a1 , ..., , an )+(0, b1 a1 ]...(0, bn an ].
Hence its measure equals mn (T ((0, b1 a1 ] ... (0, bn an ])).
Since T is continuous we may assume that bi ai Q for i = 1, ..., n.
Standard arguments show that mn (T ((0, b1 a1 ] ... (0, bn an ])) = (b1
a1 )...(bn an )mn (T ((0, 1] ... (0, 1])).
Hence we simply need to see that if Qn = (0, 1] ... (0, 1] then
mn (T (Qn )) = |det(T )|.
(3.1)
Any automorphism in Rn can be decomposed into products of automorphisms of the following types:
(i) (T (e1 ), ..., T (en )) is a permutation of (e1 , ..., en ),
(ii)T (e1 ) = e1 and T (ei ) = ei for i = 2, ..., n, or
(iii) T (e1 ) = e1 + e2 and T (ei ) = ei for i = 2, ..., n.
Hence it suffices to show (3.1) for these cases.
In the first one T (Qn ) = Qn and det(T ) = 1, in the second case T (Qn ) =
[0, ) Qn1 and det(T ) = and in the third case T (Qn ) = {(x1 , x1 +
x2 , x3 , ..., xn ) : 0 xi < 1} and det(T ) = 1. Since T (Qn ) = A2 Qn2 where
clearly m2 (A2 ) = 1 we get the result.
Case 2: T is not bijection and A B(Rn ).
The image T (A) would lie in a proper subspace. Hence the result follows
by showing that any subspace has measure zero. Actually, due to the previous
case, by composing with automorphims we may assume that T (A) {x
Rn : x1 = 0} = {0} Rn1 which is clearly of measure zero.
3.2. Fubini theorem
63
Case 3: A M(Rn ). Now the measure and |det(T )|mn are -finite
measures such that coincide on the generating family and then coincide on
B(Rn ). Now their complections must coincide and the result extends then
to Lebesgue measurable sets.
3.2
Fubini theorem
Theorem 3.2.1 Let (X, 1 , ) and (Y, 2 , ) be -finite measure spaces and
let f : X YR [0, ] be 1 2 -measurable.R Then
(i) xR Y fx d is 1R-measurable
and Ry
f y d is 2 -measurable.
R
R Xy
(ii) XY f d = X ( Y fx d)d = Y ( X f d)d.
Proof: For f = E this coincides with Corollary 3.1.7.
P
P
For simple functions f = ni=1 i Ei we have fx = ni=1 i ((Ei )x ) and
P
f y = ni=1 i ((Ei )y ). Hence f y and fx are 1 and 2 -measurable respectively. Moreover
Z
f d =
XY
coincides with
n
X
and
n
X
((Ei )x )d =
i=1
i (Ei )
i=1
i=1
n
X
((Ei ) )d =
fx d)d
f y d)d.
For a general 1 2 -measurable function f : X Y [0, ], take

a sequence of simple functions sn which increases to f . For each x X
and y Y we have that (sn )x and (sn )y are 2 and 1 -measurable simple
functions which increase to fx and f y respectively. To obtain (i) and (ii) we
simply need toR apply the monotone
convergence theorem
several Rtimes.
R
R
y
Note that Y (sn )x dR and X (sn ) d increase
to Y fx d and X f y d reR
y
spectively. Hence x Y fx d and y X f d are 2 and 1 -measurable
respectively. Moreover
lim
n
lim
n
Z
Y
(sn )x d)d =
y
(sn ) d)d =
Z
Y
fx d)d,
f y d)d,
64
which both coincide with

lim
n
Z
XY
sn d =
f d .
XY
Theorem 3.2.2 (Fubini theorem) Let (X, 1 , ) and (Y, 2 , ) be -finite

measure spaces and let f : X Y C be -integrable. Then
(i) fx is -integrable -a.e. and f y is -integrable -a.e.
R
R
(ii) x Y fx d is -integrable and y X f y d is -integrable.
R
R R
R R
(iii) XY f d = X ( Y fx d)d = Y ( X f y d)d.
Proof: Assume first that f : X Y [0, ) is -integrable. We can
apply Theorem 3.2.1 and get
Z
f d =
XY
fx d)d =
f y d)d < .
Hence
f d < -a.e. and X f y d < -a.e. and x Y fx d and
R Yy x
y X f d are and -integrable respectively.
The general case follows from the previous one applied to the decompositon f = u+ u + iv + iv .
R
Corollary 3.2.3 Let (X, 1 , ) and (Y, 2 , ) be -finite measure spaces and
let f : X Y C be 1 2 -measurable.
R R
If X ( Y |f (x, y)|d(y))d(x) < then f is -integrable and
Z
f d =
XY
Z
X
fx d)d =
Z
Y
f y d)d.
Proof: Apply Theorem 3.2.1 to the function |f | to get that it is integrable and then apply Theorem 3.2.2.
3.3
Applications
q
Definition 3.3.1 Let n N, n 2, and denote ||x|| = x21 + ...x2n , the

open unit ball by Bn = {u Rn : ||u|| < 1} and the unit sphere by Sn1 =
{u Rn : ||u|| = 1}.
3.3. Applications
65
Clearly any x 6= 0, x Rn can be decomposed as x = ru where r = ||x|| >

0 and u = x/||x|| Sn1 , what allows us to say that Rn \{0} = (0, )Sn1 .
Let us denote by : Bn \ {0} Sn1 the continuous projection defined
x
by (x) = ||x||
.
We denote by (Sn1 , B(Sn1 ), n1 ) the measure space defined by the image measure of n-times the Lebesgue measure on the B(Bn \ {0}), that is
where A = {ru : 0 < r < 1, u A}.
n1 (A) = nmn (A)
Theorem 3.3.2 (Integration in polar coordinates) Let n N, n 2 and let
f : Rn [0, ] be Borel measurable.
Then the Lebesgue measure space (Rn \ {0}, B(Rn \ {0}), mn ) coincides
with ((0, ) Sn1 , B((0, )) B(Sn1 ), rn1 dr n1 ) and
Z
Rn
f (x)dmn (x) =
(0,)
Sn1
f (ru)dn1 (u))rn1 dr.
Proof: To see the coincidence between B((0, )) B(Sn1 ) and B(Rn \

{0}) it suffices to show that rectangles A Sn1 and (0, ) B belong to
B(Rn \ {0}) for A B((0, )) and B B(Sn1 ), and that for any open set
G Rn one has that G \ {0} B((0, )) B(Sn1 ) .
Since Rn \ {0} is homoeomorphic to (0, ) Sn1 then any G Rn open
is a product of to open sets and hence belongs to B((0, )) B(Sn1 ). On
the other hand, defining
1 = {A B((0, ) : A Sn1 B(Rn \ {0})},
1 = {B B(Sn1 ) : (0, ) B B(Rn \ {0})},
we have -algebras containing the open sets, so 1 = B((0, ) and 2 =
B(Sn1 ).
To see that mn = rn1 dr dn1 it suffices to see that both measures
coincide on rectangles [r1 , r2 ) B where B is open in Sn1 (due to the fact
that both are -finite and they would coincide on A) and the extension to
the (A) is unique.
\ r1 B
and
Given [r1 , r2 ) B (0, ) Sn1 one gets (r1 , r2 ) B = r2 B
then
mn ([r1 , r2 ) B) = (r2n r1n )mn (B)

Z
r2
= (
r1
rn1 dr)n1 (B)
= rn1 dr dn1 ([r1 , r2 ) B).
66
Since both measures coincide then for simple functions f =

we have
Z
Rn
f dmn =
Rn \{0}
m
X Z
i=1
m
X
i=1
Z
Pm
i=1
i Ei
f dmn
Ei dmn
Rn \{0}
n1
Sn1
n1
Ei (ru)dn1 (u))dr
f (ru)dn1 (u))dr
Sn1
The standard argument using the monotone convergence theorem gives

the general case.
Corollary 3.3.3 Let : [0, ) [0, ] be a measurable function and f :
Rn [0, ] be the radial function f (x) = (||x||). Then
Z
Rn
f dmn = nvn
(r)rn1 dr
where vn = mn (Bn ).
Proof: Applying the previous theorem we get
Z
Rn
f dmn =
n1
Z
Sn1
Proposition 3.3.4 ( 12 ) =
R
R
Proof: Note that (1/2) =

have
Z
t2
f (ru)dn1 = nm(Bn )
et dm1 (t) =
R
0
dm1 (t))
(r)rn1 dr
R
0
et dt. Using Fubini we
e||x|| dm2 (x)
R2
= 2v2
= v2
t1/2 et dt = 2
Z
Z
0
rer dr
er dr = v2 =
3.3. Applications
67
Proposition 3.3.5 Let n N. Then vn =
n/2
.
( n
+1)
2
Proof: Write
Bn = {x Rn : ||x|| < 1} = {(t, x0 ) RRn1 : ||x0 || < 1, |t| <
1 ||x0 ||2 }.
Hence
vn =
=
Z
||x0 ||<1
m1 ({t R : |t| <
1 ||x0 ||2 })dmn1 (x0 )
||x0 ||<1
2 1 ||x0 ||2 dmn1 (x0 )
= (n 1)vn1
= (n 1)vn1
rn2 2 1 r2 dr
n3
2
1 tdt
n1 3
, )
2
2
3
)(
)
( n1
2
2
= (n 1)vn1
n
( 2 + 1)
= (n 1)vn1 B(
Applying this formula again one gets

vn = (n 1)vn1
)( 32 )
)( 32 ) ( n2
)( 32 )
( n1
( n1
2
2
2
=
(n
1)(n
2)v
.
n2
( n2 + 1)
( n2 + 1) ( n1
+ 1)
2
Since (p + 1) = p(p) we obtain

)2 ( 23 )
( n2
2
vn = 2(n 2)vn2
.
( n2 + 1)
Repeating the process we finally have for 1 k n 1
vn = 2k1 (n k)vnk
)k ( 32 )
( nk
2
,
( n2 + 1)
which, for k = n 1 and using Proposition 3.3.4 gives vn =
n/2
.
( n
+1)
2
Definition 3.3.6 Let (X, , ) be a -finite measure space. Given a measurable function f : X [0, ] we define the distribution function of f by
f (t) = ({x X : f (x) > t}).
68
Theorem 3.3.7 Let : R+ R+ be a derivable function with (0) = 0 and

0 (t) 0. If f : X [0, ] is a measurable function then
Z
(f (x))d(x) =
Z
0
0 (t)f (t)dt.
Proof: Consider the function F (x, t) = f (x) t defined on X (0, ). It

is measurable with
respect to the product -algebra. Consider the density
R
measure (E) = E 0 (t)dm(t) on B((0, )). Then
({(x, t) : F (x, t) > 0}) =
=
({t : f (x) > t > 0})d
ZX
({f (x) > t})d.
(0,)
Hence
Z
(f (x))d(x) =
Z
X
(0,f (x))
(t)dm(t))d =
Z
0
f (t)0 (t)dt.
Corollary 3.3.8 LetR (X, , ) beR a -finite measure space, 1 p < and
f measurable. Then X |f |p d = 0 ptp1 f (t)dt.
3.4
Exercises
Exercise 3.4.1 Let X, Y be non empty sets, and let M P(X) y R P(Y )
such that X M y Y R. Show that (M R) = (M) (R).
Exercise 3.4.2 Let (X, 1 , ) and (Y, 2 , ) be -finite measure spaces. De e Y their complections and by X Y
the completion of X Y with
note by X
Y = X Y
.?
respect the product measure. Does it hold that X
Exercise 3.4.3 Let (X, 1 , ) and (Y, 2 , ) be -finite complet meausure
2 , )
spaces and let (X Y, 1

be the complection of (X Y, 1
2 , ).
2 and (A)
Show that if A 1
= 0,then (Ax ) = 0 -a. e. and
(Ay ) = 0 a.e.
1
2Deduce then that Fubinis theorem holds true for non negative
1
2 -integrable functions.
measurable functions or for
3.4. Exercises
69
Exercise 3.4.4 (Integration by parts) Let be a -finite Borel measure on

[a, b] for a < b . Given -integrable functions f, g, we define, for
x a,
Z
Z
F (x) =
f d,
G(x) =
gd.
[a,x]
[a,x]
Show that , if we write F (a ) = 0, then

Z
f (x)G(x)d(x) = F (b)G(b)
[a,b]
F (x )g(x)d(x).
[a,b]
Exercise 3.4.5 Let (X, , ) be -finite measure space and f : X [0, ]

be a measurable function. For E we define
R(f, E) = {(x, y) E R :
0 y < f (x)}
and F (y) = ({x E : f (x) > y}) , y > 0 (called the distribution function
f over E).
Show that, for the Lebesgue measure m over R, we have
Z
f d = ( m)(R(f, E)) =
F (y)dm(y)
and for 0 < p < one has

Z
f p d =
ptp1 F (t)dm(t).
Exercise 3.4.6 Let (X, , ) be -finite measure space, I = (a, ) where

a < and f : X I measurable with distribution function F .
(i) Let : I R be C 1 non decreasing function with (a+ ) = 0. Then
Z
(f )d =
0 (t)F (t)dt.
(ii) Let : I R be a non decreasing continuous function such that

(a+ ) = 0. Then
Z
Z
(f )d =
F (t)dm (t).
Exercise 3.4.7 Let (N, P(N), ) be the measure space for the counting measure . Let (X, , ) be a measure space. Define, for E P(N), the
measure
(E) =
X
n=1
(En ).
70
Show that f defined from N X into [0, ] (or into C) is measurable if

and only if the sections fn are -measurables for all n N.
R
P
Show that f is -integrable if and only if the series
n=1 X |fn |d is
convergent. In such a case
Z
Z
X
f d( ) =
n=1 X
NX
fn d =
X n=1
fn d
Exercise 3.4.8 Let f, g : [0, /2] R be given by f (x) =

sin2 (x).
(i) Describe = f (m) and = g(m).
(ii) Compute ({(x, y) R2 : y < 4x2 }).
1
2
and g(x) =
Exercise 3.4.9 Let f : Rk N R be given by

f (x, y) = nx1 {(x,n):||x|| 1 } .
n
Let be a measure over N such that ({n}) = n1 . Find the values of

for f to be mk -integrable, where
mk is the Lebesgue measure over Rk . For
R
such values calculate the integral Rk N f dmk .
Exercise 3.4.10 Study the integrability on R2 the following functions:
[0,)[0,a] (x, y) for a > 0.
(i) f (x, y) = sin(x)cos(xy)
x
x2 y 2
(ii) f (x, y) = (x2 +y2 )2 .
(iii) f (x, y) =
(iv)f (x, y) =
y 2 sin2 (x)
x2 (x2 +y 2 )(x2 +y 2 +m2 )
y 2 sen2 x
x2 (x2 +y 2 )(x2 +y 2 +m2 )
R2 {(0,0)} (x, y).
R2 \{(0,0)} (x, y) for m > 0.
Exercise 3.4.11 Let X = Y = N and the counting measure. Study the

P
P
-integrability of f = n (2 2n ){(n,n)} n (2 + 2n ){(n+1,n)} .
sen(
)1
||x||
Exercise 3.4.12 Let f : Rk \ {0} R be given by f (x) = ||x||k (1||x||)
.
R
Show thatR it does not exist {||x||<1} f (x)dx but it does exist the principal
value lim0 {<||x||<1} f (x)dx. Compute such a value.
Exercise 3.4.13 Find the values of for the following functions to ve integrable Rand compute the value of their integrals.
dx
(i) Rk (1+||x||
2 ) .
R
(ii) {||x||<r} ||x|| dx.
x21 x22 +x23 +...+(1)k+1 x2k
dx.
{||x||<1}
||x||
R
|x1 |+...+|xk |
(iv) {||x||<1}
dx.
||x||
(iii)
3.4. Exercises
71
Exercise
3.4.14 Find out the integrals Ik = Bk |x1 ...xk |dmk (x) and Jk =
R
k
Sk1 |u1 ...uk |dk1 (u), where Bk is the closed unit ball of R and Sk1 the
sphere ||x|| = 1.
R
Exercise 3.4.15 Compute the integral A ( + 1 x1 + ... + k xk )dmk where

A = {x Rk : ||x a|| < r}, a Rk , i , R y r > 0.
ExerciseP3.4.16 Find out in terms of the function the value of the integral
k
R
n ( i=1 ai x2i )
dmk (x) for ai > 0 and n N .
Rk x 1 e
Exercise 3.4.17 Compute the Lebesgue measure of the set An = {x Rn :
P
xj > 0, nj=1 xj < 1}.
R
2
Use it to show that {xi >0,i=1,..,n} e(x1 +...+xn ) dx = (n/2+1)
.
n!
Exercise 3.4.18 Compute the Lebesgue measure of the following sets.
P
(i) An = {x Rn : nj=1 |xj | 1}.
P
(ii)Bn = {x Rn : nj=1 |xj |2 1}.
(iii)Cn = {x Rn : max|xj | 1}.
(iv) Dn = {x Rn : |xj | + |xn | a, j = 1, 2, .., n 1} for a > 0.
Exercise 3.4.19 Compute the measure of the following sets:
(i) A = {(x, y, z, u) : (x + y)2 + (z + u)2 < 1, |x y| + |z u| < 1}.
(ii) B = {x = (x0 , x) Rk+j : ||x0 || 1, ||x0 ||||x|| 1}.
(iii) A = {1 v1 + ... + n vn : 0 j 1, j = 1, 2, ..., n}, where v1 , ..., vn
are linearly independent vectors in Rn .
(iv) B = {x Rn : j < x.vj < j , j = 1, 2, ..., n} where v1 , ..., vn are
linearly independent vectors in Rn and x.v denotes the scalar product and
j < j for all j.
Chapter 4
The Radon-Nikodym Theorem
4.1
Complex and real measures.
Let us start with the following result to motivate the next definitions.
Proposition 4.1.1 Let (X, , ) be a measure space and f : X C be
-integrable.
R
Then (E) = E f d for E defines a complex-valued set function on
with the following properties:
P
(i) (
n=1 (En ) for any sequence {En } of pairwise disjoint
n=1 En ) =
measurable sets.
(ii) If (E) = 0 then (E) = 0.
(iii) lim (E) = 0, i.e. for all > 0 there is > 0 such that (E) <
(E)0
implies |(E)| < .

Proof: (i) Let {En } be a sequence of pairwise disjoint measurable sets.
P
P
Let us write f En =
f En = limn nk=1 f Ek . We have that gn =
n=1
Pn
Pn
E and |gn | =
E . Since |gn |
k=1 f Ek = f n
k=1 |f |Ek = |f |n
k=1 k
k=1 k
|f |, the dominated convergence theorem implies that
(nN En ) =
Z
X
f En d =
Z
X
n=1 X
f En d =
(En ).
n=1
(ii) It is obvious.
R
(iii) Given > 0 take a simple function s such that X |f s|d 2 .
73
74
Chapter 4. The Radon-Nikodym Theorem

Pn
Now for s =
|
i=1
i Ei we have that
sd|
Hence if <
n
X
|i |(Ei E) ( sup |i |)(E).

i=1,...n
i=1
2 supi=1,...n |i |
and (E) < we get that |
R
E
f d| < .
Definition 4.1.2 Let (X, ) be a measurable space. A complex measure is a

set function : C such that for any sequence {En } of pairwise disjoint
sets in
(
n=1 En ) =
(En ).
n=1
Remark 4.1.1 (i) Observe that the convergence of the series

n=1 (En )
for any complex measure and any sequence {En } of pairwise disjoint sets
in is unconditional, since any permutation converges to the same sum.
(ii) The condition () = 0 follows from the definition.
(iii) If (X, , ) is a measure space
and f : X C is -integrable then
R
Proposition 4.1.1 shows that (E) = E f d is a complex measure.
Actually there exists 1 a non negative Rmeasure such that |(E)| 1 (E)
for all E . It suffices to take 1 (E) = E |f |d.
P
Let us see first that this last remark holds true for any complex measure,
that is there exists always a non negative measure such that |(E)| (E)
for all E .
Definition 4.1.3 Let : C be a complex measure and E , we
define the variation of on E, as
||(E) = sup{
n
X
|(En )| : E =
n=1 En , En , En Em = for n 6= m}.
n=1
Remark 4.1.2 If is a complex measure, A B and A, B then

||(A) ||(B).
Theorem 4.1.4 Let : C be a complex measure. Then
(i) (X, , ||) is a measure space.
(ii) If is a non-negative measure such that |(E)| (E) for all E
then || .
4.1. Complex and real measures.
75
Proof: (i) Let us see first that || is a non-negative measure. Of course

||() = 0 since () = 0.
Let {Am } be a sequence of pairwise disjoint sets in and A =
m=1 Am .
E
where
{E
}
are
pairwise
disjoint
sets
in .
Assume now that A =
n
n=1 n
Note that En = m=1 En Am and Am = n=1 En Am . Therefore
|(En )|
n=1
X
n=1 m=1
X
|(En Am )|
|(En Am )|
m=1 n=1
||(Am )|
m=1
This shows that ||(A)

m=1 ||(Am ).
Let us now see the converse. We may assume ||(A) < and then, from
Remark 4.1.2, ||(Ak ) < for all k.
Given k N and > 0 there exist En,k pairwise disjoint such that
P
k
Ak =
n=1 En,k and ||(Ak ) <
n=1 |(En,k )| + /2 .
Therefore, since n,k En,k = A, we have
P
||(Ak ) <
|(En,k )| +
k=1 n=1
k=1
/2k < ||(A) + .
k=1
(ii) Observe that for any partition E = n En we have
X
n=1
|(En )|
(En ) = (E).
n=1
Hence ||(E) (E).

Lemma 4.1.5 Let 1 , 2 be complex measures and 1 , 2 C and E .
Then 1 1 + 2 1 is a complex measure. Moreover
|1 1 + 2 2 |(E) |1 ||1 |(E) + |2 ||2 |(E).
Proposition 4.1.6 Let R(X, , ) be a measure Rspace and f : X C be
-integrable and (E) = E f d. Then ||(E) = E |f |d for all E .
76
Proof: From Theorem 4.1.4 we have ||(E) E |f |d for all E .

P
Assume now that
s is a simple -integrable function, say s = ni=1 i Ai and
R
denote s (E) = E sd.
Therefore for E , choosing the partition, E = ni=1 (E Ai ) (E
(X \ ni=1 Ai ), we get
R
n
X
|s (Ai E)| =
i=1
n
X
|i |(Ai E) =
i=1
Z
E
|s|d |s |(E).
In general, given > 0 take a -integrable simple function s so that

X |f s|d < /2. Now for any E ,
||(E) | s |(E) + |s |(E)
Z
ZE
|f s|d +
|f s|d +
|s f |d +
|f |d
|s|d
ZE
|f |d
This concludes the result.

Lemma 4.1.7 Let z1 , z2 , ..., zn C. There exists S {1, 2, ..., n} such that
n
n
X
X
1X
|zk |.
|zk | |
zk |
k=1
k=1
kS
Proof: Let zk = |zk |eik . For each [, ) define S() = {k

P
{1, 2, ..., n} : cos(k ) > 0} and f () = nk=1 |zk |cos+ (k ).
P
P
Clearly f () = kS() |zk |cos(k ) = <( kS() ei zk ).
P
P
Hence f () | kS() ei zk | = | kS() zk |.
Integrating over [, ) we have
Z
f ()d =
n
X
k=1
n
X
|zk |
|zk |
k=1
n
X
= 2
k=1
|zk |.
cos+ (k )d
cos+ ()d
77
On the other hand, since f is continuous, say that is a point where it

attains the maximum, we have
Z
f ()d 2f () = 2|
zk |,
kS()
which concludes the proof.

Lemma 4.1.8 Let be a complex measure. If ||(E) = there exist A, B
such that A B = E with ||(A) = and |(B)| > 1.
Proof: There exists {En } such that E = n En with
N N be such that
N
X
|(En )| = . Let
|(En )| > (1 + |(E)|).
n=1
Applying Lemma 4.1.7 to zk = (Ek ) one can take A = kS Ek and B =

E \ A.
Hence
|(A)| = |
X
kS
(Ek )| >
N
1X
|(Ek )| > 1 + |(E)|.
k=1
Also |(B)| |(E) (A)| > 1.

Now, since || is a measure, then ||(A) = or ||(B) = .
Theorem 4.1.9 If is a complex measure then ||(X) < .
Proof: If ||(X) = then, applying consecutively Lemma 4.1.8, we can
find a sequence of pairwise disjoint sets En with |(En )| > 1. Hence
P
E = n En and (E) = n (En ) but the series can not converge since
(En ) does not converges to zero.
Given a -integrable function f : X R we have
that (E) = E f d
R
+
+
can be decomposed as = where (E) = E f + d and (E) =
R
E f d.
We shall see that this decomposition is actually true for any real measure.
R
78
Definition 4.1.10 A complex measure such that (E) R for all E

is called a real measure. For such a measure we define + = ||+
and
2
||
+
= 2 . Both are non-negative finite measures and
= + and || = + .
Proposition 4.1.11 Let (X, , )Rbe a measure space and let f : X R be
a -integrable function. If (E) = E f d for all E then
+
(E) =
f d,
(E) =
f d.
Proof: This follows from Proposition 4.1.6.

Theorem 4.1.12 (Jordan decomposition theorem) Let : C be a function set. Then is a complex measure if and only if there exist {i }4i=1 finite
non-negative measures such that = 1 2 + i3 i4 .
Theorem 4.1.13 Let be a real measure and E . Then
+ (E) = sup{(F ) : F E, F },
(E) = inf{(F ) : F E, F }.
Proof: Since = verifies that + = and
sup{(F ) : F E, F } = inf{(F ) : F E, F },
it suffices to see the part corresponding to + .
Let F and F E, (F ) + (F ) + (E). Hence
+ (E) sup{(F ) : F E, F }.
Given > 0 there exists {An } pairwise disjoint such that E = n An and
P
||(E) <
n=1 |(An )| + .
Write S = {n : (An ) 0}, F = nS An and G = nS
/ An . Of course
E = F G and F G = . Observe that
||(E) <
X
n=1
|(An )| + =
X
nS
(An )
(An ) + = (F ) (G) + .
nS
/
Since (E) = (F ) + (G) we get that + (E) < (F ) + , what finishes

the proof.
79
Lemma 4.1.14 Let be a real measure and E . Then there exists F

and F E such that + (E) = (F ).
Proof: Take Fn with Fn E such that + (E) (Fn ) > + (E) 21n .
Define F = lim sup Fn =

n=1 k=n Fk E.
+
Of course (F ) (E), (F ) = limn (
k=n Fk ) and
(
k=n Fk )
= (Fn ) +
(Fn+k \ k1
j=0 Fn+j ).
k=1
Now observe that
1
2n+k
+ + (E) (Fn+k )
k1
= (Fn+k \ j=0
Fn+j ) + (k1
j=0 Fn+j )
k1
(Fn+k \ j=0
Fn+j ) + + (E)
1
Therefore 2n+k
(Fn+k \ k1
j=0 Fn+j ) and then
(
k=n Fk )
X
1
1
1
= (E + ) n1 .
(E ) n
n+k
2
2
k=1 2
+
The proof is completed by taking limits.

Theorem 4.1.15 (Hahn decomposition theorem) Let be a real measure.
There exist A, B with X = A B, A B = such that (E) 0 for
all E A and (E) 0 for all E B.
Proof: Let us apply Lemma 4.1.14 for E = X and find A such that
+ (X) = (A). Write B = X \ A.
If E A then
(A) = (E) + (A \ E) (E) + + (X) = (E) + (A).
This shows that (E) 0 for all E A.
On the other hand, if E B,
(A) = + (X) (A E) = (A) + (E).
This shows that (E) 0 for all E B.
80
4.2
The theorem and its proof.
In this chapter we want to find out conditions on two measures and

on
R
a -algebra to get that has a density with respect to , i.e. (E) = E f d
for all E and some non-negative measurable function f .
Definition 4.2.1 Given two measures and on a measurable space (X, )
we say that is absolutely continuous with respect to (or -continuous ),
to be denoted << if (E) = 0 implies (E) = 0.
If is a measure and is a complex measure we also say that is continuous if (E) = 0 for all E such that (E) = 0.
Theorem 4.2.2 Let and be measures with (X) < . Then <<
if and only if lim(E)0 (E) = 0, i.e. for any > 0 there exists > 0 such
that if (E) < then (E) < .
Proof: Assume that there exists > 0 and En so that (En ) < 1/2n
but (En ) .
Take E = lim sup En . Let us see that (E) = 0 and (E) .
Since (X) < and
k=n Ek is decreasing then
(E) = lim (
k=n Ek ) .
n
On the other hand (

k=n Ek )
for all n, and then (E) = 0.
k=n
(Ek ) < 1/2n . Hence (E) 1/2n
Remark 4.2.1 The finiteness of is necessary.

R
Consider the Lebesgue measure on [0, 1] and (E) = E 1t dm(t). Clearly
En = [0, 1/n] verifies that (En ) converges to 0, but (En ) = for all n.
Theorem 4.2.3 Let (X, , ) and (X, , ) be finite spaces. Then <<
if and only ifR there exists a -integrable function f (and -a.e. unique) such
that (E) = E f d for all E .
Proof: Let us see first the uniqueness.
Assume
that there are f, g meaR
R
surables
and
non-negative
such
that
f
d
=
gd
for all E . Then
E
E
R
E (f g)d = 0 for all E and f g is -integrable. Hence f = g -a.e.
To see the existence let us consider
G = {g : X [0, ] measurable :
Z
E
gd (E), E }.
4.2. The theorem and its proof.
81
Define A = { X gd : g G}. Clearly is a non-empty

set bounded by (X).
R
Put M = supA and select gn G such that X gn d M 1/n.
Note first that fn = max{g1 , g2 , ..., gn } G. By induction, assume that
fn1 G and since fn = max{gn , fn1 } we have
R
Z
E
fn d =
Z
E{fn1 gn }
gn d +
Z
E{fn1 <gn }
fn1 d
(E {fn1 gn }) + (E {fn1 gn }) = (E).

Define
now f = Rsupn fn . Using the monotone convergent theorem we get
R
that E f d = limn E fn d (E). This proves that f G.
R
R
R
Since M
1/n
g
n d X fn d X f d M for all n N, we
X
R
obtain M = X f d.
R
R
Our aim is to show that E f d = (E) for all E . Since (E) E f d
is a non-negative measure, it suffices to see that (X) = M .
Assume
that (X) > M . Using that (X) <R we get > 0 so that
R
(X) > X (f + )d. Define now (E) = (E) E (f + )d. We have that
is a real measure. Consider
A, B the Hahn decompositionRof the measure
R
. Therefore (E) E (f + )d, for E A and (E) E (f + )d for
E B.
Define g = f B + (f + )A . We have that g G. Indeed,
Z
E
gd =
f d +
EB
(f + )d (E B) + (E A) = (E).
EA
This gives that X gd = X f d + (A) M . Hence (A) = 0 and, using

that << , we have that (A)
= 0 which implies 0. This leads to a
R
contradiction because (X) X (f + )d.
R
Remark 4.2.2 The Radon-Nikodym theorem does not hold without assumptions on the measures and .
Take the counting measure on [0, 1] and the Lebesgue measure onR [0, 1].
Since there are no -null sets besidesR we have << , but m(E) = E f d
for all E B([0, 1]) leads to f (x) = {x} f d = m({x}) = 0 for all x [0, 1].
Theorem 4.2.4 Let (X, , ) and (X, , ) be -finite spaces. Then <<
if and only if there exists aR measurable function f : X [0, ] (and -a.e.
unique) such that (E) = E f d for all E .
82
Proof: Let us write X = m Xm = n Yn where Xm are pairwise

disjoint and Yn are pairwise disjoint, (Xm ) < and (Yn ) < for
all n, m N. Hence if Xn,m = Xn Ym we have X = (n,m)N2 Xn,m where
(Xn,m ) < and (Xn,m ) < .
For fixed n, m N we can consider (X, , m,n ) and (X, , m,n ) where
m,n (E) = (E Xm,n ) and m,n (E) = (E Xm,n ) for all E . Note that
n,m << n,m because (E Xm,n ) = 0 implies (E Xm,n ) = 0. We can
then apply Theorem 4.2.3 to get fn,m such that, for all E ,
(E Xn,m ) =
Define f =
(n,m)N2
Z
E
fn,m dn,m .
fn,m Xn,m . It is measurable and

X
(E) =
(E Xn,m )
(n,m)N2
(n,m)N2
Z
E
fn,m dn,m
fn,m Xn,m d
(n,m)N2
f d.
The uniqueness follows from the fact that f = g -a.e. in Xn,m for all
(n, m) N2 and hence f = g -a.e.
Theorem 4.2.5 Let (X, , ) be a -finite space and (X, , ) any measure
space. Then << if and only if there exists aR measurable function f :
X [0, ] (and -a.e. unique) such that (E) = E f d for all E .
Proof: We may assume that there is E0 such that (E0 ) < , otherwise f = . We first deal with (X) < . Define
C = {E : : E [0, ] -finite},
where E = {E A : A }. Observe that C 6= since E0 C.
Let S = sup{(E) : E C}. Note that S (X). Take En C with
limn (En ) = S. Consider X1 = n En . Clearly X1 C, hence (X1 ) = S
4.2. The theorem and its proof.
83
since (En ) (X1 ) S for all n N. Using Theorem 4.2.4 there exists
f1 : X1 [0, ] measurable with respect X1 such that
(A X1 ) =
Z
AX1
f1 d
for all A .
Let us define f (x) = f1 (x) for x X1 and f (x) = for x X \ X1 . We
have that f is -measurable and
(E) =
Z
EX1
f1 d + (E (X \ X1 )).
Now if (E (X \ X1 )) > 0 then (E (X \ X1 )) = = E f d, since

(E (X \ X1 )) < implies that X2 = X1 E (X \ X1 ) C and
(X2 ) > S.
On the other
hand if (E
(X \ X1 )) = 0 then (E (X \ X1 )) = 0 and
R
R
then (E) = EX1 f1 d
= E f d.
R
Therefore (E) = E f d for all E .
To see
the uniqueness,
assume f, g are measurable functions satisfying
R
R
(E) = E f d = E gd for all E . Since is -finite on X1 , using the
uniqueness of Theorem 4.2.4 then there exists A X1 with (A) = 0 such
that f (x) = g(x) for all x
/ A.
Note that for all E we have
R
(E (X \ X1 )) =
f d =
E(X\X1 )
gd.
E(X\X1 )
Hence if (X \ X1 ) = 0 then f = g -a.e. and in the case (X \ X1 ) > 0

we have as above that
(X \ X1 ) = =
Z
X\X1
f d =
gd.
X\X1
This implies that f = g = -a.e. in X \ X1 .

Indeed, if ({x X \ X1 : f (x) < } > 0 gives n N so that ({x
X \ X1 : f (x) n} > 0. Hence
({x X \ X1 : f (x) n} n({x X \ X1 : f (x) n} <
and then X2 = X1 {x X \ X1 : f (x) n} C and (X2 ) > S.
The case is -finite follows from the previous case in the usual way.
84
Theorem 4.2.6 Let (X, , ) be a -finite space and a complex measure

on . Then << if and only if there exists
a -integrable function f :
R
X C (and -a.e. unique) such that (E) = E f d for all E .
Proof: Using the Jordan decomposition theorem we can write = 1 2 +
i3 i4 where i are non-negative finite measures. Note that << implies
i << for i = 1, 2,R 3, 4. Hence using Theorem 4.2.4 we find fi -integrable
such that i (E) = E fi d for all E . Therefore f = f1 f2 + if3 if4
verifies the result.
The uniqueness follows from the same argument as in the previous cases.
4.3
Applications
Definition 4.3.1 Let (X, ) be a measurable space and a measure (or a

complex measure) over X. Given A , we say that is concentrated in A
if (E) = (E A) for all E .
In other words, for non-negative measures, is concentrated in A if and
only if = A , or (X \ A) = 0.
R
Example 4.3.1 (i) A density measure (E) = E f d is concentrated in

supp(f ) = {x X : f (x) 6= 0} or in X \ N for any N with (N ) = 0.
(ii) Let be a measure, A and let : X Y be a function with
(A) (). If is concentrated in A then the image measure () is
concentrated in (A).
(iii) x is concentrated in {x}.
Definition 4.3.2 Let and be measures (or complex measures) on a measurable space (X, ). We say that they are mutually singular, denoted ,
if there exist disjoint sets A, B such that is concentrated in A and
is concentrated in B.
Example 4.3.2 (i) If f1 and f2 are measurable functions and (supp(f1 )
supp(f2 )) = then the density measures d1 = f1 d and d2 = f2 d verify
that 1 2 .
(ii) Let be a real measure, then + .
(iii) x m for any x [0, 1] where m stands for the Lebesgue measure
in [0, 1].
4.3. Applications
85
Theorem 4.3.3 (The Lebesgue decomposition theorem) Let (X, ) be a measurable space. If and are -finite measures then there exist two unique
measures a and s such that = a + s where a << and s .
Proof: Let us begin assuming (X) < . Using that << + we can
apply the Radon -Nikodym theorem and find a -integrable function f such
that, for all E ,
Z
Z
(E) =
f d + f d.
E
Define s (E) = (E A) where A = {x X : f (x) 1} and a (E) =

(E B) where B = {x X : f (x) < 1}.
Obviuosly = a + s .
If (E) = 0 then
a (E) =
f d +
EB
f d =
EB
f d.
EB
Hence EB (1 f )d = 0 since (E B) = EB f d, which shows that

(E B) = a (E) = 0 since 1 f > 0 in E B. Therefore a << .
On the other hand,
R
(A) =
Z
A
f d +
f d (A) + (A).
Hence s is concentrated in A and (A) = 0, which shows that s .

Let us show now the uniqueness. Assume = a + s = a0 + s0 where
a << , s , a0 << and s0 .
Consider the real measure = a a0 = s s0 . It is clear that << .
Let us see that .
If s is concentrated in A and in B for some A, B and AB = and
also s0 is concentrated in A0 and in B 0 for some A0 , B 0 and A0 B 0 =
then we obtain that is concentrated in A A0 and in B B 0 . Hence,
since (E (A A0 )) = 0, we get that (E) = (E (A A0 )) = 0 for all
E , which gives that a = a0 and s = s0 .
Let us now show the -finite case. Write X = n Xn where Xn are
P
pairwise disjoint measurable sets with (Xn ) < . Let us write =
n=1 n
where n (E) = (E Xn ). Using the previous case we can find (n )a and
(n )s such that n = (n )a + (n )s where (n )a << and (n )s .
P
P
Define now a =
n=1 (n )a and s =
n=1 (n )s .
Clearly a << and s since s is concentrated in n An and is
concentrated in n Bn where An Bn = for all n N.
86
To show the uniqueness observe that if = a + s = a0 + s0 where

a << , s , a0 << and s0 then we also would have that
(a )n (E) = a (E Xn ) and (a0 )n (E) = a0 (E Xn ) verify that (a )n << ,
(s )n , (a0 )n << and (s0 )n . Therefore (a )n = (a0 )n and (s )n =
(s0 )n . This gives that a = a0 and s = s0 .
Using the Jordan decomposition theorem and the Lebesgue decomposition theorem for non-negative measures we easily get the following corollary.
Corollary 4.3.4 Let (X, ) be a measurable space. If is a -finite measure and is a complex measure on then there exist two unique complex
measures a and s such that = a + s where a << and s .
Definition 4.3.5 Let (X, , ) be a measure space and let f : X [0, ]
be measurable. A non-negative number M is called an essential bound for f
if f M -a.e., that is ({x X : f (x) > M }) = 0.
It is said to be essentially bounded if it has some essential bound.
A measurable function f : X C is said to be essentially bounded if |f |
is.
Recall thaf f g if f = g -a.e. Hence if f is essentially bounded and
g f then g is essentially bounded as well. We denote by L () the space
of equivalent classes of complex-valued essentially bounded functions.
We write ||f || = inf{M 0 : f M a.e.}.
Proposition 4.3.6 Let (X, , ) be a measure space. Then (L (), ||.|| )
is a Banach space.
Proof: Clearly ||f || = 0 gives f = 0 -a.e. and ||f || = ||||f || .
Let us see the triangular inequality. Observe that
{|f + g| > ||f || + ||g|| } {|f | > ||f || } {|g| > ||g|| },
hence ({|f + g| > ||f || + ||g|| }) = 0 and then ||f + g|| ||f || + ||g|| .
To see the completeness. Let {fn } be a Cauchy sequence in L ().
Given > 0 there exists n0 N and Nn,k with (Nn,k ) = 0 such that
|fn (x) fk (x)| < /2 for x
/ Nn,k and n, k n0 .
Hence {fn (x)} is a Cauchy sequence in C for all x
/ n,kn0 Nn,k . Define
f (x) = limn fn (x) for x
/ n,kn0 Nn,k and f (x) = 0 for x n,kn0 Nn,k
where (n,kn0 Nn,k ) = 0.
4.3. Applications
87
Therefore, taking limits as k , one gets |fn (x) f (x)| /2 < for
x
/ n,kn0 Nn,k and n n0 .
This implies that ||fn f || < for n n0 and the proof is complete.
Theorem 4.3.7 Let (X, , ) be a finite measure space. Then the dual of
L1 () is isometrically isomorphic to L ().
Proof:
Let : L () (L1 ()) be defined by (f ) = f where f (g) =
R
X f gd.
Observe first that if f L () and g L1 () then f g L1 () since
|f g| ||f || |g|, which says that f is well defined. From the properties of
the integral, it is linear and verifies |f (g)| ||f || ||g||1 . Therefore f
(L1 ()) .
Also it follows from the properties of the integral that is linear and
from the previous estimate ||(f )|| ||f || .
Therefore is a continuous injective linear map with |||| 1.
Let us show that it is surjective and |||| = 1.
Assume first that (X) < . Given (L1 ()) we can define (E) =
(E ) for any E (since E L1 ()).
We first see that is a complex measure. Indeed, if {Ek } are pairwise
P
disjoint sets in then k Ek converges in L1 (). Hence
X
(k Ek ) = (
Ek ) =
X
k
(Ek ) =
(Ek ).
On the other hand | (E)| ||||(E), which implies that << . Using
now
the Radon-Nikodym theorem we obtain f L1 () such that (E) =
R
X E f d = (E ).
R
Note that | (E)| = | E f d| ||||(E) for all E . This gives
| |(E) ||||(E) for all E .
Let us define E0 = {x RX : |f (x)| > ||||}. Observe that (E0 ) = 0,
since otherwise | |(E0 ) = E0 |f |d > ||||(E0 ). This shows that f
L () and ||f || ||||.
R
Since is linear we get (s) = X sfR d for all simple functions s and by
approximation we actually
get (g) = X f gd for all g L1 (), due to the
R
facts that and g X f gd are both continuous maps. This shows that
(f ) = f = and that ||f || = ||f || = ||||.
88
To deal with the -finite case, we split X = n Xn where Xn are pairwise

disjoint and (Xn ) < .
Given (L1 ()) consider n (g) = (gXn ). Applying the previous
argument to n one gets fn L () such that ||fn || ||n || |||| and
n = (fn ) = fn for all n N.
P
Defining f = nN fn Xn we obtain a function such that f = and the
proof is finished.
4.4
Exercises
Exercise 4.4.1 Let be the measure on (R, B(R)) given by (A) =

Show that << m but lim|(E)0 (E) 6= 0.
R
A
|x|dx.
Exercise 4.4.2 Let identify Q with (rn )nN and R define, for n N, fn :
R R the non-negative Borel function such that fn dx = 1 which vanishes
in the exterior of the closed interval of length 21n centered on rn . Let (A) =
R P
fn dx for a Borel set A.
A
P
i) Show that fn (x) < m-a.e. x R.
ii) Show that is -finite, << m and that every non empty open set A
verifies that (A) = .
Exercise 4.4.3 Let and be -finite measures over (X, A), such that
<< and let g be the Radon-Nikodym derivative of with respect to and
let f be A-measurable.
R
Show
that
f
is
-integrable
if
and
only
if
f
g
es
-integrable
and
f d =
R
f gd.
Exercise 4.4.4 Let X be a non numerable set, M the class of all numerable
or co-numerable sets in X and let be the counting measure. Let (E) = 0
for numerable sets E and (E) = otherwise.
Show that, although << , one cannot define the Radon-Nikodym
derivative in this case.
d
, g =
Exercise 4.4.5 Let , , be -finite measures and let f = d(+)
d
d
, F = d(++) .
d(+)
Justify their existence and get an expression of F in terms of f and g.
4.4. Exercises
89
Exercise 4.4.6 Let be d = e ax +by dxdy defined on the Borel sets of

R2 . Let v : R2 {0} S1 the projection on the unit sphere and denote by
= v() the image measure.
Show that is absolutely continuous with respect to the Lebesgue measure
d
.
on S1 and compute d
Exercise 4.4.7 Let (X, M) be a measurable space and let (Pn ) be a sequence
of probabilities over M. Find a probability P such that Pn << P for all n N.
Exercise 4.4.8 Let be the counting measure over (N, P(N)). Show that a
measure over (R, B(R)), is absolutely continuous with respect to if and
only if there exists a sequence {an } of non-negative real numbers such that
P
d
=
n=1 an n . Compute d in this case.
Exercise 4.4.9 Let be the restriction of the Lebesgue measure m to the
-algebra F generated by the vertical strips in the plane. If (A R) =
m(A (0, 1)). Show that << but it does not have integral representation.
Exercise 4.4.10 Let be a probability measure and let be a -finite measure over R such that << . Show that the Radon-Nikodym derivative of
f verifies
(x h, x + h]
= f (x)
lim
h0 (x h, x + h]
on a set of -measure 1.
Exercise 4.4.11 Let (X, ) be a measurable space. Denote by L0 (X) the
space of complex measurable functions and by M(X) the space of complex
measures over .
(i) Let M(X). Show that there exists a function h L1 (||), essentially unique, such that d = hd||. Moreover |h(x)| = 1 a.e..
We say that f L0 (X) is integrable
(denoted f L1 ()) if f.h
R
R
L1 (||) and, in this case, we define E f d = E f.hd|| for all E .
Show that
R
R
(ii) If M(X), f L1 () and E then E f d = X E f d.
(iii) IfR M(X), f, g L0 (X), f L1 () y f = g || a.e. then
R
X f d = X gd.
R
(iv) If M(X) then T : L1 () C given by T (f ) = X f d is linear.
90
Exercise 4.4.12 Let , be complex measures which are absolutely continuous with respect to a -finite measure . Show that for all a, b C one
has
d
d
d(a + b)
=a
+b .
d
d
d
Exercise 4.4.13 Let , , be -finite measures over (X, ) such that <<
and << . Show the following chain rule
d
d d
=
. .
d
d d
Exercise 4.4.14 Let , be -finite measures over (X, ) such that <<
y << . Show that
d
6= 0 a.e.
d
d
1
=
d
d/d
a.e.
Exercise 4.4.15 Let 1 , 1 be -finite measures over (X1 , 1 ) and let 2 , 2

be -finite measures over (X2 , 2 ).
(i) If i << i (i = 1, 2) then 1 2 << 1 2 .
d(1 2 )
(ii) Compute d(
.
1 2 )
(iii) Describe, in the general case, the Lebesgue decomposition of 1 2
with respect to 1 2 .
(iv) Show that 1 2 << 1 2 if and only if 1 << 1 and 2 << 2 .
(v) Show that 1 2 is mutually singular to 1 2 if and only if 1 is
mutually singular to 1 or 2 is mutually singular with 2 .
Exercise 4.4.16 Let , be real measures defined over (X, ) and let be
a -finite measure. Show that
(i) | + | || + ||, ( + )+ + + + and ( + ) + .
(ii) | + | = || + || if and only if + , are mutually singular with
respect to + , respectively.
(iii) If is absolutely continuous with respect to and is mutually
singular with respect to then is mutually singular with respect to .
(iv) If is absolutely continuous with respect to and is also mutually
singular with respect to then = 0.
Exercise 4.4.17 Let , be real measures defined over (X, ) and (Y, R)
respectively.
4.4. Exercises
91
(i) Show that there exists a real measure over R such that
(A B) = (A)(B) for A and B R.
(ii) Find the Hahn decomposition of from the Hanh descompositions
of each factor.
(iii) Compute ( )+ , ( ) and | | in terms of those of y .
Exercise 4.4.18 Let = B([0, 1]) and (E) = m(E) + im(E [0, 21 ]).
(i) Describe || in terms of m.
(ii) Show that
(E) (Re)+ (E) + (Re) (E) + (Im)+ (E) + (Im) (E)
and that the inequality can be strict.
R
(iii) Find a Borel function h such that |h| = 1 and (E) = E hd|| for
all E .
Exercise 4.4.19 For each Borel set in R define
(E) =
Z
E(0,)
Z
sen3 t
sen3 t
dt
dt.
t3
t3
E(,0)
(i) Show that is a real measure and compute (R).

(ii) Find the Hahn decomposition of R relative to .
(iii) Study the Radon-Nikodym derivative of || with respect to m and
compute it if possible.
Exercise 4.4.20 Show that, although (E) = 0 implies (E) = 0, in general
we do not have the - condition of absolute continuity :
P
1
(i) (N, P(N)), the counting measure and =
n=1 2n n .
(ii) ([0, 1], B), d(t) = 1t dt and the Lebesgue measure .
P
(iii) (R, B), (E) = nZ |n|m([n, n + 1) E) and the Lebesgue measure
.
Exercise 4.4.21 Let

f
(x)
=
1 x for x 1 and f (x) = 0 for x > 1
R
and define (E) = E f (x)dx.
Let g(x) = x2 for x 0 and g(x) = 0 for
R
x < 0 and define (E) = E g(x)dx. Get the Lebesgue decomposition of
with respect to .
Exercise 4.4.22 Find the Lebesgue decomposition of the Lebesgue-Stieltjes
measure given by the distribution function F (x) = (E[x])2 (x E[x])2 with
respect to the Lebesgue measure.

Measure Theory

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Measure Theory

Transféré par

Droits d'auteur :

Formats disponibles

MEASURE THEORY

2 Measurable and Integrable functions

product measure and Fubinis

Basic notions on sets

Definition 1.1.1 A non empty family of subsets of X, say A P(X), is

Chapter 1. Abstract measure

1.1. Basic notions on sets

Definition 1.1.8 A pair (X, ) given by a non-empty set X and a -algebra

Chapter 1. Abstract measure

1.2. Basic notions on set functions.

Now let us consider A M(A) \ A. To see that A A observe that

Basic notions on set functions.

Definition 1.2.1 Let A be an algebra over X, a set function : A [0, ]

(6) If is a measure and (An ) A is decreasing, (A1 ) < and

(5) Put B1 = A1 and Bn = An \ An1 for n 2. Note that (Bn ) is a

Chapter 1. Abstract measure

(6) Put Bn = A1 \ An . Then (Bn ) is an increasing sequence in A, n Bn =

Definition 1.2.3 A triplet (X, , ) given by a non-empty set X, a -algebra

(A) = (A) for all A .

1.3. Outer measures.

is -algebra observe first that N. Now if

(A N ) = (A) = 0 then H N . This shows the completeness and

We would like to be able to extend measures defined in an algebra A

(iii) is subadditive, i.e. (n An )

Chapter 1. Abstract measure

Proof: (i) is immediate.

Motivated by these properties we give the following definition.

passing to the limit we get the result.

Then is an outer measure which extends .

1.3. Outer measures.

To see that extends note first that if A A then (A) (A). On

This gives that (A) (A) and the proof is finished.

Chapter 1. Abstract measure

Assume A, B . Let us see that A B . For each E P(X) we

If (An ) is a sequence of pairwise disjoint -measurable sets we have for

Hence passing to the limit as n we get the result.

Taking limits as N we get

1.4. Extension of measures.

To finish the proof we need to show that is complete. Note that

Theorem 1.4.1 (Hahns extension theorem) Let A be an algebra on X and

We now apply Caratheodorys method to get the -algebra of . If we

We show now that

Chapter 1. Abstract measure

for all A (A).

1.5. Borel-Stieltjes measures on R.

Proof: Recall that from Proposition 1.3.1 we have that

Since restricted to coincides with we get also that

Conversely, assume first that A and (A) < . Take E such

(B) = 0. So E \ B A E and this gives that A .

the previous case, that A .

Chapter 1. Abstract measure

Proof: We shall see that if x1 , x2 I with x1 < x2 then F (x2 ) F (x1 ) =

F (bk ) F (ak ) : A k (ak , bk ], ak , bk int(I)}.

1.5. Borel-Stieltjes measures on R.

We shall prove it using induction. The case n = 1 is obvious.

Theorem 1.5.5 If F is a distribution function over I then

F (bnk ) F (ank ) < F (An ) +

Chapter 1. Abstract measure

Hence F (b) F (a) N

F ((a, b] E) + F ((I \ (a, b]) E)

F ((a, b] (ak , bk ]) + F ((I \ (a, b]) (ak , bk ]) =

1.5. Borel-Stieltjes measures on R.

Definition 1.5.6 Given a distribution function F over I we denote by MF

Chapter 1. Abstract measure