Contents

1. σ-algebras 2
1.1. The Borel σ-algebra over R 8
1.2. Product σ-algebras 11
2. Measures 13
3. Outer measures and the Caratheodory Extension Theorem 20
4. Construction of Lebesgue measure 27
4.1. Regularity of Lebesgue measure 31
4.2. Examples 36
5. Premeasures and the Hahn-Kolmogorov Theorem 39
6. Lebesgue-Stieltjes measures on R 44
7. Problems 51

T
8. Measurable functions 60
9. Integration of simple functions 73
AF
10. Integration of unsigned functions 77
11. Integration of signed and complex functions 85
11.1. Basic properties of the absolutely convergent integral 86
12. Modes of convergence 91
12.1. The five modes of convergence 92
DR

12.2. Finite measure spaces 101

12.3. Uniform integrability 103
13. Problems 109
13.1. Measurable functions 109
13.2. The unsigned integral 111
13.3. The signed integral 113
13.4. Modes of convergence 114
14. The Riesz-Markov Representation Theorem 117
14.1. Urysohn’s Lemma and partitions of unity 118
14.2. Proof of Theorem 14.2 119
15. Product measures 127
16. Integration in Rn 142
17. Differentiation theorems 150
18. Signed measures and the Lebesgue-Radon-Nikodym Theorem 157
18.1. Signed measures; the Hahn and Jordan decomposition theorems 158

1
 18.2. The Lebesgue-Radon-Nikodym theorem 165
18.3. Lebesgue differentiation revisited 174
19. Problems 175
19.1. Product measures 175
19.2. Integration on Rn 176
19.3. Differentiation theorems 178
19.4. Signed measures and the Lebesgue-Radon-Nikodym theorem 179
19.5. The Riesz-Markov Theorem 180

MAA6616 COURSE NOTES

FALL 2016

T
AF
1. σ-algebras

Let X be a set, and let 2X denote the set of all subsets of X. Let E c denote the

complement of E in X, and for E, F ⊂ X, write E \ F = E ∩ F c .

DR

Definition 1.1. Let X be a set. A Boolean algebra is a nonempty collection A ⊂ 2X

that is closed under finite unions and complements. A σ-algebra is a Boolean algebra

that is also closed under countable unions.

Remark 1.2. If E ⊂ 2X is any collection of sets in X, then

!c
[ \
Ec = E. (1)
E∈E E∈E

Date: August 17, 2018.

2
 MAA6616 COURSE NOTES FALL 2016 3

Hence a Boolean algebra (resp. σ-algebra) is automatically closed under finite (resp.

countable) intersections. It follows that a Boolean algebra (and a σ-algebra) on X

always contains ∅ and X. (Proof: X = E ∪ E c and ∅ = E ∩ E c .) 

Definition 1.3. A measurable space is a pair (X, M ) where M ⊂ 2X is a σ-algebra. A

function f : X → Y from one measurable space (X, M ) to another (Y, N ) is measurable

if f −1 (E) ∈ M whenever E ∈ N . /

T
Definition 1.4. A topological space X = (X, τ ) consists of a set X and a subset τ of
AF
2X such that

(i) ∅, X ∈ τ ;

(ii) τ is closed under finite intersections;

DR

(iii) τ is closed under arbitrary unions.

The set τ is a topology on X.

(a) Elements of τ are open sets;

(b) A subset S of X is closed if X \ S is open;

(c) S is a Gδ if S = ∩∞
j=1 Oj for open sets Oj ;

(d) S is an Fσ if it is an (at most) countable union of closed sets;

(e) A subset C of X is compact, if for any collection F ⊂ τ such that C ⊂ ∪{T : T ∈ F}

there exist a finite subset G ⊂ F such that C ⊂ ∪{T : T ∈ G}; and

4 MAA6616 COURSE NOTES FALL 2016

(f) If (X, τ ) and (Y, σ) are topological spaces, a function f : X → Y is continuous if

S ∈ σ implies f −1 (S) ∈ τ .

Example 1.5. If (X, d) is a metric space, then the collection τ of open sets (in the

metric space sense) is a topology on X. There are important topologies in analysis that

are not metrizable (do not come from a metric). 4

T
Remark 1.6. There is a superficial resemblance between measurable spaces and topo-
AF
logical spaces and between measurable functions and continuous functions. In particular,

a topology on X is a collection of subsets of X closed under arbitrary unions and finite

intersections, whereas for a σ-algebra we insist only on countable unions, but require
DR

complements also. For functions, recall that a function between topological spaces is

continuous if and only if pre-images of open sets are open. The definition of measur-

able function is plainly similar. The two categories are related by the Borel algebra

construction appearing later in these notes. 

The disjointification trick in the next Proposition is often useful.

Proposition 1.7 (Disjointification). Suppose ∅ 6= M ⊂ 2X is closed with respect to

complements, finite intersections and countable disjoint unions. If (Gj )∞

j=1 is a sequence

of sets from M , then there exists a sequence (Fj )∞

j=1 of pairwise disjoint sets from M
 MAA6616 COURSE NOTES FALL 2016 5

such that
n
[ n
[
Fj = Gj
j=1 j=1

for n either a positive integer or ∞.

Hence, M is a σ-algebra if and only if M is closed under complement, finite inter-

sections and countable disjoint unions. †

Proof. The proof amounts to the observation that if (Gn ) is a sequence of subsets of X,

T
then the sets
n−1
!
[
n−1 c
Fn = Gn \ Gk = Gn ∩ (∩k=1 Gk ) (2)
AF
k=1
S∞ S∞
are disjoint, in M and
Sn Sn
j=1 Fj = j=1 Gj for all n ∈ N+ (and thus j=1 Fj = j=1 Gj ).

To prove the second part of the Proposition, given a sequence (Gn ) from M use
DR

the disjointification trick to obtain a sequence of disjoint sets Fn ∈ M such that ∪Gn =

∪Fn . 

Example 1.8. Let X be a nonempty set.

(a) The power set 2X is the largest σ-algebra on X.

(b) At the other extreme, the set {∅, X} is the smallest σ-algebra on X.

(c) Let X be an uncountable set. The collection

M = {E ⊂ X : E is at most countable or X \ E is at most countable } (3)

is a σ-algebra (the proof is left as an exercise).

6 MAA6616 COURSE NOTES FALL 2016

(d) If M ⊂ 2X a σ-algebra, and E is any nonempty subset of X, then

ME := {A ∩ E : A ∈ M } ⊂ 2E

is a σ-algebra on E (exercise).

(e) If {Mα : α ∈ A} is a collection of σ-algebras on X, then their intersection ∩α∈A Mα

is also a σ-algebra (checking this statement is a simple exercise). Hence given any

set E ⊂ 2X , we can define the σ-algebra

T
\
AFM (E ) = {M : M is a σ-algebra and E ⊂ M }. (4)

Note that the intersection is over a nonempty collection since E is a subset of the

σ-algebra 2X . The set M (E ) is the σ-algebra generated by E . It is the smallest

σ-algebra on X containing E .
DR

(f) An important instance of the construction in item (e) is when X is a topological

space and E is the collection of open sets of X. In this case the σ-algebra generated

by E is the Borel σ-algebra and is denoted BX . The Borel σ-algebra over R is

studied more closely in Subsection 1.1.

(g) If (Y, N ) is a measurable space and f : X → Y, then the collection

f −1 (N ) = {f −1 (E) : E ∈ N } ⊂ 2X (5)

is a σ-algebra on X (check this) called the pull-back σ-algebra. The pull-back σ-

algebra is the smallest σ-algebra on X such that the function f : X → Y is measur-

able.
 MAA6616 COURSE NOTES FALL 2016 7

(h) More generally given a family of measurable spaces (Yα , Nα ), where α ranges over

some index set A, and functions fα : X → Yα , let

E = {fα−1 (Eα ) : α ∈ A, Eα ∈ Nα } ⊂ 2X

and let M = M (E ). The σ-algebra M is the smallest σ-algebra on X such that

each of the functions fα is measurable. Unlike the case of a single f , the collection

E need not be σ-algebra in general. An important special case of this construction

is the product σ-algebra (see Subsection 1.2).

T
(i) If (X, M) is a measurable space and f : X → Y , then
AF Ωf = {E ⊂ Y : f −1 (E) ∈ M} ⊂ 2Y

is a σ-algebra.

4
DR

The following proposition is trivial but useful.

Proposition 1.9. If M ⊂ 2X is a σ-algebra and E ⊂ M , then M (E ) ⊂ M . †

The proposition is used in the following way. To prove a particular statement say

P is true for every set in some σ-algebra M ⊂ 2X (say, the Borel σ-algebra BX ): (1)

check to see if the collection of sets P ⊂ 2X satisfying property P is itself a σ-algebra

(otherwise it is time to look for a different proof strategy); and (2) find a collection of

sets E (say, the open sets of X) such that each E ∈ E has property P and such that
8 MAA6616 COURSE NOTES FALL 2016

M (E ) = M . It then follows that M = M (E ) ⊂ P. (The monotone class lemma,

which we will study later, is typically used in a similar way.)

A function f : X → Y between topological spaces is said to be Borel measurable if

it is measurable when X and Y are equipped with their respective Borel σ-algebras.

Proposition 1.10. If X and Y are topological spaces and if f : X → Y is continuous,

then f is Borel measurable. †

T
Proof. Problem 7.7. (Hint: follow the strategy described after Proposition 1.9.) 
AF
1.1. The Borel σ-algebra over R. Before going further, we take a closer look at the

Borel σ-algebra over R, beginning with the following useful lemma on the structure of
DR

open subsets of R, which may be familiar to you from advanced calculus.

Lemma 1.11. Every nonempty open subset U ⊂ R is an (at most countable) disjoint

union of open intervals. †

Here the “degenerate” intervals (−∞, a), (a, +∞), (−∞, +∞) are allowed.

Proof outline. First verify that if I and J are intervals and I ∩ J 6= ∅, then I ∪ J is an

interval. Given x ∈ U , let

αx = sup{a : [x, a) ⊂ U }

βx = inf{b : (b, x] ⊂ U }
 MAA6616 COURSE NOTES FALL 2016 9

and let Ix = (αx , βx ). Verify that, for x, y ∈ U either Ix = Iy or Ix ∩ Iy = ∅. Indeed,

x ∼ y if Ix = Iy is an equivalence relation on U . Hence, U = ∪x∈U Ix expresses U as

a disjoint union of nonempty intervals, say U = ∪p∈P Ip where P is an index set and

the Ip are nonempty intervals. For each q ∈ Q ∩ U there exists a unique pq such that

q ∈ Ipq . On the other hand, for each p ∈ P there is a q ∈ Q ∩ U such that q ∈ Ip . Thus,

the mapping from Q ∩ U to P defined by q 7→ pq is onto. It follows that P is at most

countable. 

T
Proposition 1.12 (Generators of BR ). Each of the following collections of sets E ⊂ 2R

generates the Borel σ-algebra BR :

AF
(i) the open intervals E1 = {(a, b) : a, b ∈ R};

(ii) the closed intervals E2 = {[a, b] : a, b ∈ R};

DR

(iii) the (left or right) half-open intervals E3 = {[a, b) : a, b ∈ R} or E4 = {(a, b] : a, b ∈

R};

(iv) the (left or right) open rays E5 = {(−∞, a) : a ∈ R} or E6 = {(a, +∞) : a ∈ R};

(v) the (left or right) closed rays E7 = {(−∞, a] : a ∈ R} or E8 = {[a, +∞) : a ∈ R}.

Proof. Only the open and closed interval cases are proved, the rest are similar and left

as exercises. The proof makes repeated use of Proposition 1.9. Let O denote the open

subsets of R. Thus, by definition, BR = M (O). To prove M (E1 ) = BR , first note that

since each interval (a, b) is open and thus in O, M (E1 ) ⊂ M (O) by Proposition 1.9.
10 MAA6616 COURSE NOTES FALL 2016

Conversely, each open set U ⊂ R is a countable union of open intervals, so M (E1 )

contains O and hence (after another application of Proposition 1.9) M (O) ⊂ M (E1 ).

For the closed intervals E2 , first note that each closed set is a Borel set, since it is

the complement of an open set. Thus E2 ⊂ BR so M (E2 ) ⊂ BR by Proposition 1.9.

Conversely, each open interval (a, b) is a countable union of closed intervals [a + n1 , b − n1 ].

Indeed, for −∞ < a < b < ∞,

∞
[ 1 1
(a, b) = [a + ,b − ]

T
n=N
n n

and a similar construction handles the cases that either a = −∞ or b = ∞. It follows

AF
that E1 ⊂ M (E2 ), so by Proposition 1.9 and the first part of the proof,

BR = M (E1 ) ⊂ M (E2 ).
DR

Sometimes it is convenient to use a more refined version of the above Proposition,

where we consider only dyadic intervals.

Definition 1.13. A dyadic interval is an interval of the form

 
k k+1
I= , (6)
2n 2n

where k, n are integers. /

 MAA6616 COURSE NOTES FALL 2016 11

(Draw a picture of a few of these to get the idea). A key property of dyadic intervals

is the nesting property: if I, J are dyadic intervals, then either they are disjoint, or one is

contained in the other. Dyadic intervals are often used to “discretize” analysis problems.

Proposition 1.14. Every open subset of R is a countable disjoint union of dyadic in-

tervals. †

Proof. Problem 7.5. 

T
It follows (using the same idea as in the proof of Proposition 1.12) that the dyadic
AF
intervals generate BR . The use of half-open intervals here is only a technical convenience,

to allow us to say “disjoint” in the above proposition instead of “almost disjoint.”

DR

1.2. Product σ-algebras. Suppose n ∈ N+ and (Xj , Mj ) are σ-algebras for j =

Qn
1, 2, . . . , n. Let X = j=1 Xj , the product space. Thus X = {(x1 , . . . , xn ) : xj ∈

Xj , j = 1, . . . , n}. Let πj : X → Xj denote the j-th coordinate projection, π(x) = xj .

The product σ-algebra, defined below, is the smallest σ-algebra on X such that each πj

is measurable.

Definition 1.15. Given measurable spaces (Xj , Nj ), j = 1, . . . n, the product σ-algebra

⊗nj=1 Nj is the σ algebra on X =

Qn
j=1 Xj generated by

{πj−1 (Ej ) : Ej ∈ Nj , j = 1, . . . n}.

12 MAA6616 COURSE NOTES FALL 2016

Given Ej ∈ Nj for j = 1, . . . , n, the set ×nj=1 Ej ∈ ⊗nj=1 Nj is a measurable rectangle.

Proposition 1.16. The collection R of measurable rectangles in ⊗nj=1 Nj generates the

product σ-algebra. †

Proof. Each measurable rectangle is a finite intersection of elements of

E = {πj−1 (Ej ) : Ej ∈ Nj , j = 1, . . . n}.

T
Hence R ⊂ M (E ). On the other hand E ⊂ R and hence the reverse inclusion holds. 
AF
There are now two canonical ways of constructing σ-algebras on Rn . The Borel

σ-algebra BRn and the product σ-algebra obtained by giving each copy of R the Borel

σ-algebra BR and forming the product σ-algebra ⊗n1 BR . It is reasonable to suspect that
DR

these two σ-algebras are the same, and indeed they are.

Proposition 1.17. BRn = ⊗nj=1 BR . †

Proof. We use Proposition 1.9 to prove inclusions in both directions. By definition, the

product σ-algebra ⊗nk=1 BR is generated by the collection of sets

E = {πj−1 (Ej ) : Ej ∈ BR , j = 1, . . . n},

where πj (x1 , . . . xn ) = xj is the projection map, π : Rn → R. Summarizing, M (E) =

⊗nj=1 BR .
 MAA6616 COURSE NOTES FALL 2016 13

For each j, the projection πj is continuous and hence, by Proposition 1.10, Borel

measurable. Consequently, if Ej ∈ BR , then

πj−1 (Ej ) = R × · · · × R × Ej × R × · · · × R ∈ BRn .

where Ej is the j th factor. Hence E ⊂ BRn and, by Proposition 1.9, ⊗n1 BR = M (E) ⊂

BRn .

Let On are the open sets in Rn . To prove the reverse inclusion, it suffices to identify

a subset Ro of the product σ-algebra ⊗nj=1 BR such that M (Ro ) ⊃ On , since then

T
⊗nj=1 BR ⊃ M (Ro ) ⊃ M (On ) = BRn .
AF
Qn
Let Ro denote the collection of open rectangles, R = (a1 , b1 )×· · ·×(an , bn ) = j=1 (aj , bj ) ∈

⊗nj=1 BR . Each U ∈ On is a countable union of open rectangles (just take all the open

boxes contained in U having rational vertices). Hence On ⊂ M (Ro ). (Equality holds,

DR

of course. But we only need this inclusion.) 

2. Measures

Definition 2.1. Let X be a set and M a σ-algebra on X. A measure on M is a function

µ : M → [0, +∞] such that

(i) µ(∅) = 0; and

j=1 is a sequence of disjoints sets in M , then

(ii) if (Ej )∞
∞
! ∞
[ X
µ Ej = µ(Ej ).
j=1 j=1
14 MAA6616 COURSE NOTES FALL 2016

If µ(X) < ∞, then µ is finite. If there exists a sequence (Xj ) from M such that

X = ∪∞
j=1 Xj and µ(Xj ) < ∞ for each j, then µ is σ-finite.

A triple (X, M , µ) where X is a set, M is a σ-algebra and µ a measure on M , is a

measure space. /

Almost all of the measures of importance in analysis are σ-finite.

Here are some simple measures and some procedures for producing new measures

T
from old. Non-trivial examples of measures will have to wait for the Caratheodory and

Hahn-Kolmogorov theorems in the following sections.

AF
Example 2.2. (a) Let X be any set and, for E ⊂ X, let |E| denote the cardinality of

E, in the sense of a finite number or ∞. The function µ : 2X → [0, +∞] defined by

DR

µ(E) = |E| is a measure on (X, 2X ), called counting measure. It is finite if and only

if X is finite, and σ-finite if and only if X is at most countable.

(b) Let X be an uncountable set and M the σ-algebra of (at most) countable and co-

countable sets (Example 1.8(b)). The function µ : M → [0, ∞] defined by µ(E) = 0

if E is countable and µ(E) = +∞ if E is co-countable is a measure.

(c) Let (X, M , µ) be a measure space and E ∈ M . Recall ME from Example 1.8(c).

The function µE (A) := µ(A ∩ E) is a measure on (E, ME ). (Why is the assumption

E ∈ M necessary?)
 MAA6616 COURSE NOTES FALL 2016 15

(d) (Linear combinations) If µ is a measure on M and c > 0, then (cµ)(E) =: c µ(E) is

a measure, and if µ1 , . . . µn are measures on the same M , then

(µ1 + · · · µn )(E) := µ1 (E) + · · · µn (E)

P∞
is a measure. Likewise a countably infinite sum of measures n=1 µn is a measure.

(The proof of this last fact requires a small amount of care. See Problem 7.9.)

T
One can also define products and pull-backs of measures, compatible with the con-

structions of product and pull-back σ-algebras. These examples will be postponed until
AF
we have built up some more machinery of measurable functions.

Theorem 2.3 (Basic properties of measures). Let (X, M , µ) be a measure space.

DR

(a) (Monotonicity) If E, F ∈ M and F ⊂ E, then µ(E) = µ(E \ F ) + µ(F ). In

particular, µ(F ) ≤ µ(E) and if µ(E) < ∞, then µ(F \ E) = µ(F ) − µ(E).
S∞ P∞
j=1 ⊂ M , then µ(
(b) (Subadditivity) If (Ej )∞ j=1 Ej ) ≤ j=1 µ(Ej ).

j=1 ⊂ M and Ej ⊂ Ej+1 ∀j, then lim µ(Ej )

(c) (Monotone convergence for sets) If (Ej )∞

exists and moreover µ(∪Ej ) = lim µ(Ej ).

(d) (Dominated convergence for sets) If (Ej )∞

j=1 is a decreasing (Ej ⊃ Ej+1 for all j)

from M and µ(E1 ) < ∞, then lim µ(Ej ) exists and moreover µ(∩Ej ) = lim µ(Ej ).

Proof. (a) Since E = (E \ F ) ∪ F is a disjoint union of measurable sets, by additivity,

µ(E) = µ(E \ F ) + µ(F ) ≥ µ(F ).

16 MAA6616 COURSE NOTES FALL 2016

(b) For 1 ≤ j ≤ g, let

j−1
!
[
Fj = Ej \ Ek .
k=1

By proposition 1.7, the Fj are pairwise disjoint, Fj ⊂ Ej for all j and ∪∞ ∞

j=1 Fj = ∪j=1 Ej .

Thus by countable additivity and (a),

∞
! ∞
! ∞ ∞
[ [ X X
µ Ej =µ Fj = µ(Fj ) ≤ µ(Ej ).

T
AF j=1 j=1 j=1 j=1

(c) With the added assumption that the sequence (Ej )∞

j=1 is nested increasing,

Sj
k=1 Fk = Ej for each j. Thus, by countable additivity,

∞
! ∞
!
DR

[ [
µ Ej =µ Fj
j=1 j=1
∞
X
= µ(Fk )
k=1
j
X
= lim µ(Fk )
j→∞
k=1
j
!
[
= lim µ Fk
j→∞
k=1

= lim µ(Ej ).
j→∞

(d) The sequence µ(Ej ) is decreasing (by (a)) and bounded below, so lim µ(Ej )
S∞ T∞
exists. Let Fj = E1 \ Ej . Then Fj ⊂ Fj+1 for all j, and j=1 Fj = E1 \ j=1 Ej . So by
 MAA6616 COURSE NOTES FALL 2016 17

(a) and (c) applied to the Fj , and since µ(E1 ) < ∞,

∞
\ ∞
\
µ(E1 ) − µ( Ej ) = µ(E1 \ Ej )
j=1 j=1

= lim µ(Fj )

= lim(µ(E1 ) − µ(Ej ))

= µ(E1 ) − lim µ(Ej ).

Again since µ(E1 ) < ∞, it can be subtracted from both sides. 

T
Example 2.4. Note that in item (d) of Theorem 2.3, the hypothesis “µ(E1 ) < ∞” can
AF
be replaced by “µ(Ej ) < ∞ for some j”. However the finiteness hypothesis cannot be

removed entirely. For instance, consider (N, 2N ) equipped with counting measure, and

let Ej = {k : k ≥ j}. Then µ(Ej ) = ∞ for all j but µ( ∞

T
j=1 Ej ) = µ(∅) = 0. 4
DR

For any set X and subset E ⊂ X, there is a function 1E : X → {0, 1} defined by

(
1 if x ∈ E
1E (x) = ,
0 if x 6∈ E

called the characteristic function or indicator function of E. It is easily verified, if (X, M )

is a measure space and E ⊂ X, then E ∈ M if and only if 1E is (M , BR ) measurable.

For a sequence of subsets (En ) of X, by definition (En ) converges to E pointwise if

1En → 1E pointwise1. This notion allows the formulation of a more refined version of

the dominated convergence theorem for sets, which foreshadows (and is a special case

1What would happen if we asked for uniform convergence?

18 MAA6616 COURSE NOTES FALL 2016

of) the dominated convergence theorem for the Lebesgue integral. See Problems 7.12

and 7.13.

Definition 2.5. Let (X, M , µ) be a measure space. A null set (or µ-null set) is a set

E ∈ M with µ(E) = 0. /

It follows immediately from countable subadditivity that a countable union of null

sets is null. The contrapositive of this statement is a measure-theoretic version of the

T
pigeonhole principle:
AF
Proposition 2.6 (Pigeonhole principle for measures). If (En )∞
n=1 is a sequence of sets

in M and µ(∪En ) > 0, then µ(En ) > 0 for some n. †

DR

It will often be tempting to assert that if µ(E) = 0 and F ⊂ E, then µ(F ) = 0, but

one must be careful: F need not be a measurable set. This caveat is not a big deal in

practice, however, because we can always enlarge the σ-algebra on which a measure is

defined so as to contain all subsets of null sets, and it will usually be convenient to do

so.

Definition 2.7. A measure space (X, M , µ) is complete if it contains every subset of

measure 0; i.e., if F ⊂ X and there exists E such that

(i) F ⊂ E;

(ii) E ∈ M ; and
 MAA6616 COURSE NOTES FALL 2016 19

(iii) µ(E) = 0,

then F ∈ M . /

Theorem 2.8. Suppose (X, M , µ) be a measure space and let N := {N ∈ M |µ(N ) =

0}. The collection

M := {E ∪ F |E ∈ M , F ⊂ N for some N ∈ N }

is a σ-algebra, and µ : M → [0, ∞] given by

T
µ(E ∪ F ) := µ(E)
AF
is a well-defined function and a complete measure on M such that µ|M = µ.

The measure space (X, M , µ) is the completion of (X, M , µ). It is evident that if

M ⊂ N ⊂ 2X is a σ-algebra, ν is a measure on N such that ν|M = µ and (X, N , ν)

DR

is complete, then M ⊂ N and ν|M = µ. Thus (X, M , µ) is the smallest complete

measure space extending (X, M , µ).

Some of the proof. First note that M and N are both closed under countable unions,

so M is as well. To see that M is closed under complements, consider E ∪ F ∈ M with

E ∈ M , F ⊂ N ∈ N . Using, F c = N c ∪ (N \ F ),

(F ∪ E)c = F c ∩ E c = (N c ∩ E c ) ∪ (N ∩ F c ∩ E c ).

The first set on the right hand side is in M and the second is a subset of N . Thus the

union is in M as desired. Hence M is a σ-algebra.

20 MAA6616 COURSE NOTES FALL 2016

To prove that µ is well defined, suppose G = E ∪ F = E 0 ∪ F 0 for E, E 0 ∈ M and

F, F 0 ∈ N . In particular, there exists µ-null sets N, N 0 ∈ M with F ⊂ N and F 0 ⊂ N 0 .

Observe that

M 3 E \ E 0 ⊂ G \ E 0 ⊂ F 0 ⊂ N 0.

Thus µ(E \ E 0 ) = 0. On the other hand,

E = (E ∩ E 0 ) ∪ (E \ E 0 ).

T
Thus, µ(E) = µ(E ∩ E 0 ). By symmetry, µ(E 0 ) = µ(E 0 ∩ E).
AF
The proof that µ is a complete measure on M that extends µ, is left as an exercise

(Problem 7.14). 
DR

3. Outer measures and the Caratheodory Extension Theorem

The point of the construction of Lebesgue measure on the real line is to extend the

naive notion of length for intervals to a suitably large family of subsets of R. Indeed, this

family should be a σ-algebra containing all open intervals and hence the Borel σ-algebra.

Definition 3.1. Let X be a nonempty set. A function µ∗ : 2X → [0, +∞] is an outer

measure if

(i) µ∗ (∅) = 0;

(ii) (Monotonicity) if A ⊂ B, then µ∗ (A) ≤ µ∗ (B);

 MAA6616 COURSE NOTES FALL 2016 21

(iii) (Subadditivity) if (Aj )∞ X

j=1 ⊂ 2 , then

∞
! ∞
[ X
µ∗ Aj ≤ µ∗ (Aj ).
j=1 j=1

Definition 3.2. If µ∗ is an outer measure on X, then a set E ⊂ X is outer measurable

(or µ∗ -measurable, or measurable with respect to µ∗ , or just measurable) if

µ∗ (A) = µ∗ (A ∩ E) + µ∗ (A ∩ E c ) (7)

T
for every A ⊂ X. /
AF
The significance of outer measures and (outer) measurable sets stems from the following

theorem.
DR

Theorem 3.3 (Caratheodory Extension Theorem). If µ∗ is an outer measure on X,

then the collection M of outer measurable sets is a σ-algebra and the restriction of µ∗

to M is a complete measure.

The outer measures encountered in these notes arise from the following construction.

Proposition 3.4. Suppose E ⊂ 2X and ∅, X ∈ E . If µ0 : E → [0, +∞] and µ0 (∅) = 0,

then the function µ∗ : 2X → [0, ∞] defined by

(∞ ∞
)
X [
µ∗ (A) = inf µ0 (En ) : En ∈ E and A ⊂ En (8)
n=1 n=1

is an outer measure. †
22 MAA6616 COURSE NOTES FALL 2016

Remark 3.5. A few remarks are in order. Given a (nonempty) index set I and noneg-

ative real numbers aα for α ∈ I, define

X X X
aα := { aα : F ⊂ I , F finite}.
α∈I α∈F

In the case I is countable, if φ : N+ → I is a bijection, then

X ∞
X
aα = aφ(j) .
α∈I j=1

In particular, the sum does not depend on the bijection φ. Hence in the definition of

outer measure, the sums can be interpreted as the sum over countable collections of sets

T
from E that cover A. For instance, in the case I = N+ × N+
AF X ∞
X X
am,n = am,n .
(m,n)∈N+ ×N+ s=1 m+n=s

It is also true that

∞ X
X ∞ X ∞ X
X ∞
am,n = am,n = am,n .
DR

n=1 m=1 (m,n)∈N+ ×N+ m=1 n=1

The proofs of these assertions are left as an exercise. See Problem 7.9.

Note that we have assumed ∅, X ∈ E , so there is at least one covering of A by sets

in E (take E1 = X and all other Ej empty), so the definition (8) makes sense. On the

other hand, Proposition 3.4 is mute on whether E ∈ E is µ∗ -outer measurable or, in the

case E is outer measurable, whether µ∗ (E) and µ0 (E) agree. 

Proof of Proposition 3.4. It is immediate from the definition that µ∗ (∅) = 0 (cover the

empty set by empty sets) and that µ∗ (A) ≤ µ∗ (B) whenever A ⊂ B (any covering of B

is also a covering of A). To prove countable subadditivity, we make our first use of the
 MAA6616 COURSE NOTES FALL 2016 23

“/2n ” trick. Let (An ) be a sequence in 2X . If µ∗ (Aj ) = ∞ for some j, then the desired

subadditive inequality holds by monotonicity. Otherwise let  > 0 be given. For each n
S∞
k=1 in E such that An ⊂
there exists a countable collection of sets (En,k )∞ k=1 En,k and

∞
X
µ0 (En,k ) − 2−n < µ∗ (An ).
k=1

But now the countable collection (En,k )∞ ∞

n,k=1 covers ∪n=1 An , and, using Remark 3.5,

∞
X ∞
X ∞
X
∗ ∗ −n
µ (∪An ) ≤ µ0 (En,k ) < (µ (An ) + 2 )=+ µ∗ (An ).
k,n=1 n=1 n=1

T
P∞
Since  > 0 was arbitrary, µ∗ (∪Aj ) ≤ n=1 µ∗ (An ). 
AF
Example 3.6. [Lebesgue outer measure] Let E ⊂ 2R be the collection of all open

intervals (a, b) ⊂ R, with −∞ ≤ a < b ≤ +∞, together with ∅. Define m0 ((a, b)) = b−a
DR

and m0 (∅) = 0. The corresponding outer measure is Lebesgue outer measure and it is

the mapping m∗ : 2R → [0, ∞] defined, for A ∈ 2X , by

(∞ ∞
)
X [
m∗ (A) = inf (bn − an ) : A ⊂ (an , bn ) (9)
n=1 n=1

where we allow the degenerate intervals R = (−∞, +∞) and ∅. The value m∗ (A) is the

Lebesgue outer measure of A. In the next section we will construct Lebesgue measure

from m∗ via the Caratheodory Extension Theorem. The main issues will be to show

that the outer measure of an interval is equal to its length, and that every Borel subset

of R is outer measurable. The other desirable properties of Lebesgue measure (such as

translation invariance) will follow from this construction. 4

24 MAA6616 COURSE NOTES FALL 2016

Before proving Theorem 3.3 will make repeated use of the following observation.

Namely, if µ∗ is an outer measure on a set X, to prove that a subset E ⊂ X is outer

measurable, it suffices to prove that

µ∗ (A) ≥ µ∗ (A ∩ E) + µ∗ (A \ E)

for all A ⊂ X, since the opposite inequality for all A is immediate from the subadditivity

of µ∗ .

The following lemma will be used to show the measure constructed in the proof of

T
Theorem 3.3 is complete. A set E ⊂ X is called µ∗ -null if µ∗ (E) = 0.
AF
Lemma 3.7. Every µ∗ -null set is µ∗ -measurable. †

Proof. Let E be µ∗ -null and A ⊂ X. By monotonicity, A ∩ E is also µ∗ -null, so by

monotonicity again,
DR

µ∗ (A) ≥ µ∗ (A \ E) = µ∗ (A ∩ E) + µ∗ (A \ E).

Thus the lemma follows from the observation immediately preceding the lemma. 

Proof of Theorem 3.3. We first show that M is a σ-algebra. It is immediate from Defi-

nition 3.2 that M contains ∅ and X, and since (7) is symmetric with respect to E and

E c , M is also closed under complementation. Next we check that M is closed under

finite unions (which will prove that M is a Boolean algebra). So, let E, F ∈ M and fix

an arbitrary A ⊂ X. Since F is outer measurable,

µ∗ (A ∩ E c ) = µ∗ ((A ∩ E c ) ∩ F ) + µ∗ ((A ∩ E c ) ∩ F c ). (10)

 MAA6616 COURSE NOTES FALL 2016 25

By subadditivity and the set equality A ∩ (E ∪ F ) = (A ∩ E) ∪ (A ∩ (F ∩ E c )),

µ∗ (A ∩ (E ∪ F )) ≤ µ∗ (A ∩ E) + µ∗ (A ∩ (F ∩ E c )). (11)

Using equations (11) and (10) and the outer measurability of E in that order,
µ∗ (A ∩ (E ∪ F ))+µ∗ (A ∩ ((E ∪ F )c ))

≤µ∗ (A ∩ E) + µ∗ (A ∩ (F ∩ E c )) + µ∗ (A ∩ (F c ∩ E c ))

=µ∗ (A ∩ E) + µ∗ (A ∩ E c )

=µ∗ (A).

T
Hence E ∪ F is outer measurable.
AF
Now we show that M is closed under countable disjoint unions. (It then follows

from Proposition 1.7 that M is a σ-algebra.) Let (En ) be a sequence of disjoint outer

measurable sets, and let A ⊂ X be given. It is enough to show

DR

∞
[ ∞
[
∗ ∗ ∗
µ (A) ≥ µ (A ∩ En ) + µ (A \ En ).
n=1 n=1

SN
For each N ≥ 1, we have already proved that GN = n=1 En is outer measurable, and

therefore
N
[ N
[
∗ ∗ ∗
µ (A) ≥ µ (A ∩ En ) + µ (A \ En ).
n=1 n=1
SN S∞
By monotonicity, µ∗ (A \ n=1 En ) ≥ µ∗ (A \ n=1 En ). Thus it suffices to prove

N
[ ∞
[
∗ ∗
lim µ (A ∩ En ) ≥ µ (A ∩ En ). (12)
N →∞
n=1 n=1

(The limit exists as an extended real number since the sequence is increasing by mono-
SN
tonicity of the outer measure.) By the outer measurability of GN = n=1 En and
26 MAA6616 COURSE NOTES FALL 2016

disjointness of the En ,
N
[ +1
µ∗ (A ∩ En ) =µ∗ (A ∩ GN +1 )
n=1

=µ∗ ((A ∩ GN +1 ) ∩ GN ) + µ∗ ((A ∩ GN +1 ) ∩ GcN ) (13)

N
[
∗
=µ (A ∩ En ) + µ∗ (A ∩ EN +1 ).
n=1

Iterating the identity of equation (13) gives.

N
[ +1 N
X +1
µ∗ (A ∩ En ) = µ∗ (A ∩ Ek ). (14)
n=1 k=0

T
(In particular, choosing A = X, it follows that µ∗ is finitely additive on M .) Letting N

tend to infinity in equation (14) gives

AF N
[ ∞
X
lim µ∗ (A ∩ En ) = µ∗ (A ∩ EN +1 ).
N →∞
n=1 N =0

From countable subadditivity,

N ∞ ∞ ∞
DR

[ X [ [
∗ ∗ ∗ ∗
lim µ (A ∩ En ) = µ (A ∩ EN +1 ) ≥ µ ( (A ∩ Ej )) = µ (A ∩ Ej ),
N →∞
n=1 N =0 j=1 j=1

proving the inequality of equation (12) and thus that M is a σ-algebra. Further, from

monotonicity and equation (13),

∞
[ N
[ N
X
∗ ∗
µ (A ∩ En ) ≥ µ (A ∩ En ) = µ∗ (A ∩ En )
n=1 n=1 n=0

and hence
∞
[ ∞
X
µ∗ (A ∩ En ) ≥ µ∗ (A ∩ En ).
n=1 n=1

Since the reverse inequality holds by subadditivity,

∞
[ ∞
X
∗
µ (A ∩ En ) = µ∗ (A ∩ En ).
n=1 n=1
 MAA6616 COURSE NOTES FALL 2016 27

In particular, choosing A = X proves that µ∗ is countably additive on the σ-algebra of

µ∗ -outer measurable sets and hence µ∗ |M is a measure.

Finally, that µ∗ is a complete measure on M is an immediate consequence of

Lemma 3.7. 

4. Construction of Lebesgue measure

In this section, by an interval we mean any set I ⊂ R of the from (a, b), [a, b], (a, b], [a, b),

T
including ∅, open and closed half-lines and R itself. Let |I| = b − a, the length of the

interval I, interpreted as +∞ in the line and half-line cases and 0 for ∅. Recall the
AF
definition of Lebesgue outer measure of a set A ⊂ R from Example 9:
(∞ ∞
)
X [
m∗ (A) = inf |In | : A ⊂ In
n=1 n=1
DR

where the In are open intervals, or empty.

Theorem 4.1. If I ⊂ R is an interval, then m∗ (I) = |I|.

Proof. We first consider the case where I is a finite, closed interval [a, b]. For any  > 0,

the single open interval (a − , b + ) covers I, so m∗ (I) ≤ (b − a) + 2 = |I| + 2, and

thus m∗ (I) ≤ |I|. For the reverse inequality, again choose  > 0, and let (In ) be a cover
P∞
of I by open intervals such that n=1 |In | < m∗ (I) + . Since I is compact, there is a

finite subcollection (Ink )N

k=1 of the In that covers I. We claim

N
X
|Ink | > b − a = |I|. (15)
k=1
28 MAA6616 COURSE NOTES FALL 2016

To verify this statement, observe that by passing to a further subcollection, we can

assume that none of the intervals Ink is contained in another one and so that each has

non-trivial intersection with [a, b]. Re-index the intervals as I1 , . . . IN so that the left

endpoints a1 , . . . aN are listed in increasing order. Since these intervals cover I, there

are no containments, and each intersects [a, b] it follows that a1 < a, a2 < b1 , a3 < b2 ,

...aN < bN −1 and b < bN . (Draw a picture.) Therefore

N
X N
X N
X −1
|Ik | = (bk − ak ) = bN − a1 + (bk − ak+1 ) ≥ bN − a1 > b − a = |I|.

T
k=1 k=1 k=1

From the inequality (15) it follows m∗ (I) = |I|.

AF
Now we consider the cases of bounded, but not closed, intervals (a, b), (a, b], [a, b).

If I is such an interval, then I = [a, b] its closure Since m∗ is an outer measure, by

monotonicity m∗ (I) ≤ m∗ (I) = |I|. On the other hand, if  > 0, then I := [a+, b−] ⊂
DR

I and thus, by monotonicity again, m∗ (I) ≥ m∗ (I ) = |I| − 2. Hence m∗ (I) ≥ |I|.

Finally, the result is immediate in the case of unbounded intervals, since any un-

bounded interval contains arbitrarily large bounded intervals and m∗ is monotonic. 

Theorem 4.2. Every Borel set E ∈ BR is m∗ -measurable.

Proof. By the Caratheodory extension theorem, the collection of m∗ -measurable sets is a

σ-algebra, so by Propositions 1.12 and 1.9, it suffices to show that the open rays (a, +∞)

are m∗ -measurable. Fix a ∈ R and an arbitrary set A ⊂ R. We must prove

m∗ (A) ≥ m∗ (A ∩ (a, +∞)) + m∗ (A ∩ (−∞, a]).

 MAA6616 COURSE NOTES FALL 2016 29

To simplify the notation put A1 = A ∩ (a, +∞), A2 = A ∩ (−∞, a]. Let (In ) be a cover

of A by open intervals. For each n let In0 = In ∩ (a, +∞) and In00 = In ∩ (−∞, a]. The

families (In0 ), (In00 ) are intervals (not necessarily open) that cover A1 , A2 respectively. Now
∞
X ∞
X ∞
X
|In | = |In0 | + |In00 |
n=1 n=1 n=1
X∞ ∞
X
∗
= m (In0 ) + m∗ (In00 )
n=1 n=1
∞
[ ∞
[
≥ m∗ ( In0 ) + m∗ ( In00 )
n=1 n=1

T
∗ ∗
≥ m (A1 ) + m (A2 ),

where the second equality follows from Theorem 4.1, the first inequality from sub-
AF
additivity and the last inequality by monotonicity. Since this inequality holds for

all coverings of A by open intervals, taking the infimum on the left hand side gives

m∗ (A) ≥ m∗ (A1 ) + m∗ (A2 ). 

DR

Definition 4.3. A set E ⊂ R is called Lebesgue measurable if

m∗ (A) = m∗ (A ∩ E) + m∗ (A ∩ E c ) (16)

for all A ⊂ R. The restriction of m∗ to the Lebesgue measurable sets is called Lebesgue

measure, denoted m. /

By Theorem 3.3, m is a measure. By Theorem 4.2, every Borel set is Lebesgue

measurable, and by Theorem 4.1 the Lebesgue measure of an interval is its length. It

should also be evident by now that m is σ-finite. So, we have arrived at the promised
30 MAA6616 COURSE NOTES FALL 2016

extension of the length function on intervals to a measure. (A proof of uniqueness of m

will have to wait until for the Hahn Uniqueness Theorem. See Corollary 5.5.)

Next we prove that m has the desired invariance properties. Given E ⊂ R, x ∈ R,

and t > 0, let

E +x = {y ∈ R : y−x ∈ E}, −E = {y ∈ R : −y ∈ E}, and tE = {y ∈ R : y/t ∈ E}.

It is evident that m∗ (E + x) = µ∗ (E), m∗ (−E) = m∗ (E) and, m∗ (tE) = tm∗ (E) since,

T
if I is an interval, then |I + x| = |I|, | − I| = |I| and |tI| = t|I|. Thus m∗ has the desired
AF
invariance properties. In particular, if both E and E + x are Lebesgue measurable, then

m(E + x) = m(E). What remains to be shown is that if E is Lebesgue measurable, then

so are E + x, −E and tE. This fact follows easily from the corresponding invariance
DR

property of m∗ .

Theorem 4.4. If E ⊂ R is Lebesgue measurable, x ∈ R, and t > 0, then the sets E + x,

−E, and tE are Lebesgue measurable. Moreover m(E + x) = m(E), m(−E) = m(E),

and m(tE) = tm(E).

Proof. We give the proof for E + x. Proof of the others are similar and left as exercises.

Accordingly, suppose E is measurable. To prove E +x is measurable, let A ⊂ R be given

and observe that A ∩ (E + x) = ((A − x) ∩ E) + x and A ∩ (E + x)c = ((A − x) ∩ E c ) + x.

 MAA6616 COURSE NOTES FALL 2016 31

Thus,

m∗ (A) = m∗ (A − x)

= m∗ ((A − x) ∩ E) + m∗ ((A − x) ∩ E c )

= m∗ (((A − x) ∩ E) + x) + m∗ (((A − x) ∩ E c ) + x)

= m∗ (A ∩ (E + x)) + m∗ (A ∩ (E + x)c ),

where measurability of E is used in the second equality. Hence E + x is Lebesgue

T
measurable and m(E + x) = m(E). 
AF
4.1. Regularity of Lebesgue measure. The condition (16) does not make clear which

subsets of R are Lebesgue measurable. Theorems 4.5 and 4.6 are fundamental approxi-

mation results. They say 1) up to sets of measure zero, every Lebesgue measurable set

is a Gδ or an Fσ , and 2) if we are willing to ignore sets of measure , then every set of

DR

finite Lebesgue measure is a union of intervals. (Recall that a set in a topological space

is called a Gδ -set if it is a countable intersection of open sets, and an Fσ -set if it is a

countable union of closed sets.)

Theorem 4.5. Let E ⊂ R. The following are equivalent.

(a) E is Lebesgue measurable.

(b) For every  > 0, there is an open set U ⊃ E such that m∗ (U \ E) < .

(c) For every  > 0, there is a closed set F ⊂ E such that m∗ (E \ F ) < .

(d) There is a Gδ set G such that E ⊂ G and m∗ (G \ E) = 0.

32 MAA6616 COURSE NOTES FALL 2016

(e) There is an Fσ set F such that E ⊃ F and m∗ (E \ F ) = 0.

Proof. To prove (a) implies (b) let E a (Lebesgue) measurable set and  > 0 be given.

Further, suppose for the moment that m(E) < ∞. Because E is measurable, m(E) =

m∗ (E). From the definition of m∗ , there is a covering of E by open intervals In such

P∞ S∞
that n=1 |In | < m∗ (E) + . Put U = n=1 In . By subadditivity of m,

∞
X ∞
X
m(U ) ≤ m(In ) = |In | < m(E) + .

T
n=1 n=1

Since U ⊃ E and m(E) < ∞ (and both U and E are Lebesgue measurable), Theorem
AF
2.3 implies m∗ (U \ E) = m(U \ E) = m(U ) − m(E) < .

To remove the finiteness assumption on E, we apply the /2n trick: for each n ∈ Z

let En = E ∩ (n, n + 1]. The En are disjoint measurable sets whose union is E, and
DR

m(En ) < ∞ for all n. For each n, by the first part of the proof we can pick an open

set Un so that m(Un \ En ) < /2|n| . Let U be the union of the Un . Thus U is open
S∞
and U \ E ⊂ n=1 (Un \ En ) since E c ⊂ Enc . The subadditivity of m gives m(U \ E) <

2−|n| = 3.
P
n∈Z

To prove that (b) implies (d), let E ⊂ R be given and for each n ≥ 1 choose (using
T∞
(b)) an open set Un ⊃ E such that m∗ (Un \ E) < 1
n
. Put G = n=1 Un . Thus G is

a Gδ containing E, and G \ E ⊂ Un \ E for every n. By monotonicity of m∗ we see

m∗ (G \ E) < 1
n
for every n and thus m∗ (G \ E) = 0. (Note that in this portion of the

proof we cannot (and do not!) assume E is measurable.)

 MAA6616 COURSE NOTES FALL 2016 33

To prove (d) implies (a), suppose G is a Gδ set such that E ⊂ G and µ∗ (G \ E) = 0.

Since G is a Gδ , it is a Borel set and hence Lebesgue measurable by Theorem 4.2. By

Lemma 3.7, every m∗ -null set is Lebesgue measurable, so G \ E and hence Gc ∪ E is

Lebesgue measurable. Finally, E = G ∩ (Gc ∩ E) is Lebesgue measuarable.

To prove that (a) implies (c), suppose E is Lebesgue measurable and let  > 0

be given. Thus E c is Lebesgue measurable and, by the already established implication

(a) implies (b), there is an open set U such that E c ⊂ U and m(U \ E c ) < . Since

T
U \ E c = U ∩ E = E \ U c , it follows that µ(E \ U c ) < . Observing that U c is closed

completes the proof.

AF
Now suppose E ⊂ R and (c) holds. Choose a sequence of closed sets (Fn ) such that

Fn ⊂ E and m∗ (E \ Fn ) < n1 . The set F = ∪∞

j=1 Fj is an Fσ and, by monotonicity, for

each n we have m∗ (E \ F ) ≤ m∗ (E \ Fn ) < n1 . Hence m∗ (E \ F ) = 0. Thus (c) implies

DR

(e).

Finally, if (e) holds, then E = F ∪(E\F ) for some closed set F ⊂ E with µ∗ (E\F ) =

0. Thus, E is the union of a closed (and hence Lebesgue) set and a set of outer measure

zero (which is thus Lebesgue). Since the Lebesgue sets are closed under union, E is

Lebesgue and the proof is complete. 

Recall the symmetric difference of sets A, B ⊂ X is A∆B = (A \ B) ∪ (B \ A) =

(A ∪ B) \ (A ∩ B).
34 MAA6616 COURSE NOTES FALL 2016

Theorem 4.6. If E is Lebesgue measurable and m(E) < ∞, then for each  > 0 there

exists a set A that is a finite union of open intervals such that m(E∆A) < .

Proof. Since E is measurable, m(E) < ∞ and m(E) = m∗ (E), there exists a sequence

(In ) of open intervals that covers E such that

∞
X
|In | < m(E) + /2. (17)
n=1

Since the sum is finite there exists an integer N so that

T
∞
X
|In | < /2. (18)
n=N +1

Let U =
S∞
n=1 In
AF and A =
SN
n=1 In . Then A \ E ⊂ U \ E, so m(A \ E) ≤ m(U ) − m(E) <
S∞
/2 by (17). Similarly E \A ⊂ U \A ⊂ n=N +1 In , so m(E \A) < /2 by (18). Therefore

m(E∆A) < . 
DR

Thus, while the “typical” measurable set can be quite complicated in the set-

theoretic sense (i.e. in terms of the Borel hierarchy), for most questions in analysis

this complexity is irrelevant. In fact, Theorem 4.6 is the precise expression of a useful

heuristic:

Littlewood’s First Principle of Analysis: Every measurable set E ⊂ R with m(E) <

∞ is almost a finite union of intervals.

Definition 4.7. Let X be a topological space. A neighborhood U of a point x ∈ X is

an open set such that x ∈ U .

 MAA6616 COURSE NOTES FALL 2016 35

A topological space X is locally compact if for each x ∈ X there is a neighborhood

Ux of x and a compact set Cx such that x ∈ Ux ⊂ Cx .

A topological space is Hausdorff if given x, y ∈ X with x 6= y, there exists neigh-

borhoods U and V of x and y respectively such that U ∩ V = ∅. (Distinct points can

be separated by open sets.)

A Borel measure is a measure on the Borel σ-algebra BX of a locally compact

Hausdorff space X.

T
A Borel measure µ is outer regular if, for all E ∈ BX ,
AF µ(E) = inf{µ(U ) : U ⊃ E and U is open}

and is inner regular if

µ(E) = sup{µ(K) : K ⊂ E and K is compact}.

DR

Finally µ is regular if it is both inner and outer regular. /

Theorem 4.8. If E ⊂ R is Lebesgue measurable, then

m(E) = inf{m(U ) : U ⊃ E and U is open}

= sup{m(K) : K ⊂ E and K is compact}

That is, m is a regular Borel measure.

Proof. Fix E. Let ρ(E) denote the infimum in the first equality. By monotonicity,

ρ(E) ≥ m(E). If m(E) = ∞, then equality is evident. The case m(E) < ∞ follows

from Theorem 4.5(b) (together with the additivity of m).

36 MAA6616 COURSE NOTES FALL 2016

For the second equality, let ν(E) be the value of the supremum on the right-hand

side. By monotonicity m(E) ≥ ν(E). For the reverse inequality, first assume m(E) < ∞

and let  > 0 be given. By Theorem 4.5(c), there is a closed subset F ⊂ E with

m(E \ F ) < /2. Since m(E) < ∞, by additivity m(E) < m(F ) + /2. Thus m(F ) >

m(E) − /2. However this F need not be compact. To fix this potential shortcoming, for

each n ≥ 1 let Kn = F ∩[−n, n]. Then the Kn are an increasing sequence of compact sets

whose union is F . By monotone convergence for sets (Theorem 2.3(c)), there is an n so

T
that m(Kn ) > m(F ) − /2. It follows that m(Kn ) > m(E) − , and thus ν(E) ≥ m(E).

The case m(E) = +∞ is left as an exercise. 

AF
4.2. Examples.

Example 4.9. [The Cantor set] Recall the usual construction of the “middle thirds”
DR

Cantor set. Let E0 denote the unit interval [0, 1]. Obtain E1 from E0 by deleting the

middle third (open) subinterval of E0 , so E1 = [0, 31 ] ∪ [ 32 , 1]. Continue inductively as

follows. At the nth step delete the middle thirds of all the intervals present at that step.

So, (En ) is nested decreasing and En is a union of 2n closed intervals of length 3−n . The
T∞
Cantor set is defined as the intersection C = n=0 En . It is well-known (though not

obvious and not proven here) that C is uncountable. It is clear that C is a closed set

(hence Borel) that contains no (non-trivial) interval, since if J is an interval of length `

and n is chosen so that 3−n < `, then J 6⊂ En and thus J 6⊂ C. The Lebesgue measure

of En is (2/3)n , which goes to 0 as n → ∞, and thus by monotonicity (or dominated

 MAA6616 COURSE NOTES FALL 2016 37

convergence for sets) m(C) = 0. So, C is an example of an uncountable, closed set of

measure 0. Another way to see that C has measure zero is to note that at the nth stage

(n ≥ 1) we have deleted a collection of 2n−1 disjoint open intervals, each of length 3−n .

Thus the Lebesgue measure of [0, 1] \ C is

∞
X 1 32
2n−1 3−n = 2 = 1.
n=1
21− 3

Thus m(C) = 0. 4

Example 4.10. [Fat Cantor sets] The standard construction of the Cantor set can be

T
modified in the following way. Fix a number 0 < c < 1 and imitate the construction
AF
of the Cantor set, except at the nth stage delete, from each interval I present at that

stage, an open interval centered at the midpoint of I of length 3−n c. (In the previous

construction c = 1.) Again at each stage we have a set En that is a union of 2n closed
DR

Pn
intervals each of which has length at most ( 3−c
6
)n and m([0, 1] \ En ) = j=1 2j−1 3cj .
T∞
Let F = n=0 En . One can prove (in much the same way as for C) that 1) F is an

uncountable, closed set; 2) F contains no intervals; and 3) m(F ) = 1 − c > 0. Thus, F

is a closed set of positive measure that contains no (non-trivial) interval. 4

Example 4.11. [Construction of Vitali sets] The Vitali sets are perhaps the most ele-

mentary examples of subsets of R that are not Lebesgue measurable. The construction

depends on the axiom of choice. Since Q ⊂ R is a subgroup of the abelian group R

under addition, declaring x ∼ y if and only if x − y ∈ Q defines an equivalence relation

on R. This relation partitions R into disjoint equivalence classes whose union is R. In

38 MAA6616 COURSE NOTES FALL 2016

particular, for each x ∈ R its equivalence class is the set {x + q : q ∈ Q}. Each equiv-

alence class C contains an element of the closed interval [0, 1]. By the axiom of choice,

there is a set E ⊂ [0, 1] that contains exactly one member xC from each class C. Each

such E is a Vitali set. We claim any such E is not Lebesgue measurable.

To prove the claim, let y ∈ [0, 1] be given. Let C denote the equivalence class of y.

Thus y differs from xC by some rational number in the interval [−1, 1]. Hence

[
[0, 1] ⊂ (E + q).

T
q∈[−1,1]∩Q

On the other hand, since E ⊂ [0, 1] and |q| ≤ 1,

AF [
(E + q) ⊂ [−1, 2].
q∈[−1,1]∩Q

Finally, by the construction of E the sets E + p and E + q are disjoint if p, q are distinct

rationals. Arguing by contradiction, suppose E is measurable. In this case the sets E +q

DR

are measurable by Theorem 4.4, and, by the countable additivity and monotonicity of

m,
X
1≤ m(E + q) ≤ 3. (19)
q∈[−1,1]∩Q

But by translation invariance (Theorem 4.4 again), all of the m(E + q) must be equal

yielding the contradiction that the sum in equation (19) is either 0 or ∞. 4

Remark: The construction of Example 4.11 can be modified to show if F is any

Lebesgue set with m(F ) > 0, then F contains a nonmeasurable (i.e., a non-Lebesgue)

subset. See Problem 7.29.

 MAA6616 COURSE NOTES FALL 2016 39

5. Premeasures and the Hahn-Kolmogorov Theorem

Definition 5.1. Let A ⊂ 2X be a Boolean algebra. A premeasure on A is a function

µ0 : A → [0, +∞] satisfying

(i) µ0 (∅) = 0; and

j=1 is a sequence of disjoint sets in A and ∪1 Aj ∈ A , then

(ii) if (Aj )∞ ∞

∞
! ∞
[ X
µ0 Aj = µ0 (Aj )
j=1 j=1

T
AF /

Finiteness and σ-finiteness are defined for premeasures in the same way as for mea-

sures. Note that a premeasure is automatically finitely additive and hence monotone.

Example 5.2. By an h-interval we mean a (finite or infinite) interval of the form (a, b].
DR

(By convention (a, +∞) is an h-interval.) The collection A ⊂ 2R of finite unions of

h-intervals is a Boolean algebra. The function µ0 : A → [0, ∞] defined by

n
X
µ0 (I) = b j − aj
j=1

for I ∈ A written as the disjoint union ∪n1 (aj , bj ] is a premeasure on A . (Warning:

that µ0 is well defined and a premeasure is immediate if we have already constructed

Lebesgue measure, but it is not as obvious as it seems to prove from scratch - there are

many different ways to decompose a given I as a finite or countable disjoint union of

h-interval. Thus verifying that µ0 is well defined and countable additivity is somewhat

delicate. See Section 6.) 4

40 MAA6616 COURSE NOTES FALL 2016

Theorem 5.3 (Hahn-Kolmogorov Theorem). If µ0 is a premeasure on a Boolean algebra

A ⊂ 2X and µ∗ is the outer measure on X defined by (8), then every set A ∈ A is outer

measurable and µ∗ |A = µ0 . Thus, letting M denote the σ-algebra of µ∗ -outer measurable

sets and µ the (complete) measure µ = µ∗ |M , we have µ0 is the restriction of µ to A .

In particular, if µ0 : A → [0, +∞] is a premeasure on a Boolean algebra A , then

there exists a σ-algebra B ⊃ A and a measure µ : B → [0, +∞] such that µ|A = µ0 .

Proof. If A ⊂ 2X is a Boolean algebra and µ0 : A → [0, +∞] is a premeasure, then

T
µ0 determines an outer measure µ∗ by Proposition 3.4. We will prove that (1) µ∗ |A =
AF
µ0 , and (2) every set in A is µ∗ -outer measurable. The theorem then follows from

Theorem 3.3.

To prove (1), let E ∈ A . It is immediate that µ∗ (E) ≤ µ0 (E), since for a covering
DR

of E from A we can take A1 = E and Aj = ∅ for all other j. For the reverse inequality,

let (Aj ) be any covering of E by sets Aj ∈ A and let (A0n ) denote the disjointification

of (An )
n−1
[
A0n = An \ Aj .
j=1

By Proposition 1.7, the A0n ∈ A , are pairwise disjoint and ∪n1 Aj = ∪n1 A0j for n =

1, 2, . . . , ∞. Let Bn = E ∩ A0n . Hence the Bn are disjoint sets in A whose union is E,

and Bn ⊂ An for all n. Thus by the countable additivity and monotonicity of µ0 ,

∞
X ∞
X
µ0 (E) = µ0 (Bn ) ≤ µ0 (An )
n=1 n=1

Since the covering was arbitrary, µ0 (E) ≤ µ∗ (E).

 MAA6616 COURSE NOTES FALL 2016 41

For (2), let E ∈ A , A ⊂ X, and  > 0 be given. There exists a sequence of sets
S∞ P∞
(Bj ) ⊂ A such that A ⊂ j=1 Bj and j=1 µ0 (Bj ) < µ∗ (A) + . By additivity and

monotonicity of µ0 ,

∞
X
µ∗ (A) +  > µ0 (Bj )
j=1
X∞ ∞
X
= µ0 (Bj ∩ E) + µ0 (Bj ∩ E c ).
j=1 j=1

Since (Bj ∩ E) is a sequence from A and A ∩ E ⊂ ∪∞

j=1 (Bj ∩ E), it follows that

T
∞
X
∗
µ (A ∩ E) ≤ µ0 (Bj ∩ E)
AF j=1

and similarly for E c . Thus µ∗ (A) +  ≥ µ∗ (A ∩ E) + µ∗ (A ∩ E c ) and since  > 0 as well

as A ⊂ R and E ∈ A are arbitrary, the proof is complete. 

DR

The measure µ constructed in Theorem 5.3 is the Hahn-Kolmogorov extension of the

premeasure µ0 . The relationship between premeasures, outer measures, and measures

in this construction is summarized in the following table:

domain additivity condition

premeasure Boolean alge- countably additive, when
bra A possible
X
outer measure all of 2 monotone, countably
subadditive
measure σ-algebra countably additive
containing A

The premeasure µ0 has the right additivity properties, but is defined on too few subsets

of X to be useful. The corresponding outer measure µ∗ constructed in Proposition 3.4

42 MAA6616 COURSE NOTES FALL 2016

is defined on all of 2X , but we cannot guarantee countable additivity. By Theorem 5.3,

restricting µ∗ to the σ-algebra of outer measurable sets is “just right.”

We have established that every premeasure µ0 on an algebra A can be extended to

a measure on the σ-algebra generated by A . The next theorem addresses the uniqueness

of this extension.

Theorem 5.4 (Hahn uniqueness theorem). Suppose A is a Boolean algebra and let N

denote the σ-algebra it generates. If µ0 is premeasure on A and µ∗ is the outer measure

T
it determines, then every extension of µ0 to a measure on N agrees on sets E ∈ N of
AF
finite outer measure.

Further, if µ0 is σ-finite, then µ0 has a unique extension to a measure ν on N .

DR

Uniqueness can fail in the non-σ-finite case. An example is outlined in Problem 7.20.

Proof. Let N be the σ-algebra generated by A , let ν denote the Hahn-Kolmogorov

extension of µ0 , but restricted to N . Thus, letting µ denote the outer measure µ∗

determined by µ0 restricted to the µ∗ -outer measurable sets M , we have ν = µ|N . Let

νe be any other extension of µ0 to N . We first show, if E ∈ N , then νe(E) ≤ ν(E). Let

S∞
E ∈ N and let (An ) be a sequence in A such that E ⊂ n=1 An . Then
∞
X ∞
X
νe(E) ≤ νe(An ) = ν0 (An )
n=1 n=1

Taking the infimum over all such coverings of E, it follows that νe(E) ≤ ν(E). (Recall

the definition of µ.)

 MAA6616 COURSE NOTES FALL 2016 43

Next we show, if E ∈ N and ν(E) < ∞ (the finite outer measure assumption),

then ν(E) ≤ νe(E). As a first observation, note that given a sequence (An ) from A and
S∞
letting A = n=1 An ∈ N , monotone convergence for sets (twice) implies

N
[ N
[
νe(A) = lim νe( An ) = lim ν( An ) = ν(A). (20)
N →∞ N →∞
n=1 n=1

Now let  > 0 be given and choose a covering (An ) of E by sets in A such that, letting
S∞
A = n=1 An , we have ν(A) < ν(E) + . Consequently ν(A \ E) < . In particular,

νe(A \ E) ≤ ν(A \ E) < , since A \ E ∈ N . Thus, using equation (20),

T
ν(E) ≤ ν(A)
AF = νe(A)

= νe(E) + νe(A \ E)

< νe(E) + .
DR

Since  was arbitrary, we conclude ν(E) ≤ νe(E). At this point the first part of the

Theorem is proved.

Now suppose µ0 is σ-finite. Thus there exists a sequence of sets (Xn ) such that

Xn ∈ A , µ0 (Xn ) < ∞ and X = ∪Xn . By Proposition 1.7, we may assume the (Xn ) are

pairwise disjoint. If E ∈ N , then E ∩ Xn ∈ N and ν(E ∩ Xn ) ≤ ν(Xn ) = µ0 (Xn ) < ∞.

Therefore, from what has already been proved,

∞
X ∞
X
ν(E) = ν(E ∩ Xn ) = νe(E ∩ Xn ) = νe(E).
n=1 n=1


44 MAA6616 COURSE NOTES FALL 2016

Corollary 5.5 (Uniqueness of Lebesgue measure). If µ is a Borel measure on R such

that µ(I) = |I| for every interval I, then µ(E) = m(E) for every Borel set E ⊂ R. †

6. Lebesgue-Stieltjes measures on R

Let µ be a Borel measure on R. (Thus the domain of µ contains all Borel sets,

though we allow that the domain of µ may be larger.) The measure µ is locally finite

if µ(I) < ∞ for every compact interval I. (Equivalently, µ(I) is finite for every finite

T
interval.) Given a locally finite Borel measure, define a function F : R → R by

0
 if x = 0,
AF F (x) = µ((0, x]) if x > 0, (21)

−µ((x, 0]) if x < 0.

It is not hard to show, using dominated and monotone convergence for sets, that F

is nondecreasing and continuous from the right; that is, F (a) = limx→a+ F (x) for all
DR

a ∈ R (see Problems 7.22 and 7.23). In this section we prove the converse: given any

increasing, right-continuous function F : R → R, there is a unique locally finite Borel

measure µ such that (21) holds. The proof will use the Hahn-Kolmogorov extension

theorem.

Let A ⊂ 2R denote the Boolean algebra generated by the half-open intervals (a, b].

(We insist that the interval be open on the left and closed on the right, a convention

compatible with the definition of F .) More precisely, A consists of all finite unions of

intervals of the form (a, b] (with (−∞, b] and (a, +∞) allowed). Fix a nondecreasing,

right-continuous function F : R → R. Since F is monotone, the limits F (+∞) :=

 MAA6616 COURSE NOTES FALL 2016 45

limx→+∞ F (x) and F (−∞) := limx→−∞ F (x) exist (possibly +∞ or −∞ respectively).

For each interval I = (a, b] in A , we define its F -length by

|I|F := F (b) − F (a).

Given a set A ∈ A , we can write it as a disjoint union of intervals A =

SN
n=1 In with

In = (an , bn ]. Define

N
X N
X
µ0 (A) = |In |F = F (bn ) − F (an ). (22)
n=1 n=1

T
Proposition 6.1. The expression (22) is a well-defined premeasure on A . †
AF
Proof. That µ0 is well-defined and finitely additive on A is left as an exercise.

To prove that µ0 is a premeasure, let (In ) be a disjoint sequence of intervals in A

DR

S∞
and suppose J = n=1 In ∈ A . For now assume J is an h-interval. By finite additivity

(and monotonicity),
N
! N
[ X
µ0 (J) ≥ µ0 In = µ0 (In ).
n=1 n=1

Taking limits, we conclude µ( ∞

S P∞
n=1 In ) ≥ n=1 µ0 (In ).

For the reverse inequality, we employ a compactness argument similar to the one

used in the proof of Theorem 4.1. However, the situation is more complicated since

we are dealing with half-open intervals. The strategy will be to shrink J to a slightly

smaller compact interval, and enlarge the In to open intervals, using the right-continuity

of F and the /2n trick to control their F -lengths.

46 MAA6616 COURSE NOTES FALL 2016

We’ll prove the reverse inequality assuming J = (a, b] is a finite interval, leaving the

cases of the infinite intervals as an exercise. Accordingly, fix  > 0. By right continuity

of F , there is a δ > 0 such that F (a+δ)−F (a) < . Likewise, writing In = (an , bn ], there

exist δn > 0 such that F (bn +δn )−F (bn ) < 2−n . Let J˜ = [a+δ, b] and I˜n = (an , bn +δn ).

It follows that J˜ ⊂ J = ∪In ⊂ ∪I˜n . Hence, by compactness, finitely many of the I˜n

˜ and these may be chosen so that none is contained in another, each has non-
cover J,

trivial intersection with J˜ and we may reindex so that these n intervals are relabeled as

T
I˜1 , . . . I˜N so their left endpoints are listed in increasing order. (This rearrangement does

not change the sum

AF
DR

∞
X
[F (bj + δj ) − F (aj )]
j=1

though it does of course change the partial sums.)

 MAA6616 COURSE NOTES FALL 2016 47

As in the proof of Theorem 4.1, it follows that a1 < a + δ, a2 < b1 + δ1 , a3 < b2 ,

...aN < bN −1 + δN −1 and b < bN + δN . Thus,

µ0 (J) ≤ F (b) − [F (a + δ) − ]

≤ F (bN + δN ) − F (a1 ) + 
N
X −1
= F (bN + δN ) − F (aN ) + (F (aj+1 ) − F (aj )) + 
j=1
N
X −1
≤ F (bN + δN ) − F (aN ) + (F (bj + δj ) − F (aj )) + 
j=1

T
∞
X
≤ (F (bj + δj ) − F (aj )) + 
j=1
AF X∞
≤ µ0 (Ij ) + 2.
j=1

P∞
Since  was arbitrary, µ0 (J) ≤ j=1 µ0 (Ij ) as claimed.

Now suppose J ∈ A is not necessarily an h-interval. In any case, J = ∪m

k=1 Jk is a
DR

finite disjoint union of h-intervals. Each Jk is the a countable disjoint union

Jk = ∪∞
j=1 Jk ∩ Ij

P∞
of h-intervals From what has already been proved, µ0 (Jk ) = j=1 µ0 (Jk ∩ Ij ). Since µ0

is finitely additive,

m
X X ∞
m X ∞ X
X m ∞
X
µ0 (Jk ) = µ0 (Jk ∩ Ij ) = µ0 (Jk ∩ Ij ) = µ0 (Ij ).
k=1 k=1 j=1 j=1 k=1 j=1

n=1 is a disjoint sequence from A and J = ∪An ∈ A . For each

Finally, suppose (An )∞

n, there exists disjoint h-intervals (In,` )N

`=1 such that An = ∪` In,` . Thus J = ∪In,` is a
n
48 MAA6616 COURSE NOTES FALL 2016

disjoint union of h-intervals and, by what has already been proved and finite additivity,

XX X
µ0 (J) = µ0 (In,` ) = µ0 (An ).
n ` n

By the Hahn-Kolmogorov Theorem, µ0 extends to a Borel measure µF , called the

Lebesgue-Stieltjes measure associated to F . It is immediate from the definition that µ0

is σ-finite (each interval (n, n + 1] has finite F -length), so the restriction of µF to the

T
Borel σ-algebra is uniquely determined by F by Theorem 5.4. In particular we conclude
AF
that the case F (x) = x recovers Lebesgue measure.

Example 6.2. (a) (Dirac measure) Define the Heaviside function

(
DR

1 if x ≥ 0
H(x) =
0 if x < 0

Then for any interval I = (a, b], µH (I) = 1 if 0 ∈ I and 0 otherwise. Since Dirac

measure δ0 is a Borel measure and also has this property, and the intervals (a, b]

generate the Borel σ-algebra, it follows from the Hahn Uniqueness Theorem (The-

orem 5.4) that µH (E) = δ0 (E) for all Borel sets E ⊂ R. There is nothing special

about 0 here. Given p ∈ R, let δp denote the Borel measure defined by δ(E) = 1 if

p ∈ E and 0 if p ∈
/ E. For a finite set x1 , . . . xn in R and positive numbers c1 , . . . cn ,
Pn Pn
let F (x) = j=1 cj H(x − xj ). Then µF = j=1 cj δxj . Not that F is continuous

except at the points xj where F (xj −) = cj .

 MAA6616 COURSE NOTES FALL 2016 49

(b) (Infinite sums of point masses) Even more generally, if (xn )∞

n=1 is an infinite sequence

P∞
in R and (cn ) is a sequence of positive numbers with n=1 cn < ∞, define F (x) =
P P
cn H(x − xn ) = n:xn ≤x cn . It follows, using Theorem 5.4 and Problem 7.9, that
P P∞
µF (E) = n:xn ∈E cn ; i.e., µF = n=1 cn δxn . A particularly interesting case is when

the xn enumerate the rationals; the resulting function F is continuous precisely on

the irrationals. We will return to this example after the Radon-Nikodym theorem.

(c) (Cantor measure) Recall the construction of the Cantor set C from Example 4.9.

T
P∞
Each number x ∈ [0, 1] has a base 3 expansion, of the form x = n=1 an 3−n , where

an ∈ {0, 1, 2} for all n. The expansion is unique if we insist that every terminating
AF
expansion (an = 0 for all n sufficiently large) is replaced with an expansion ending

with an infinite string of 2’s (that is, an = 2 for all n sufficiently large). With these

conventions, it is well-known that C consists of all points x ∈ [0, 1] such that the
DR

base 3 expansion of x contains only 0’s and 2’s. (Referring again to the construction

of C, x belongs to E1 if and only if a1 is 0 or 2, x belongs to E2 if and only if both

a1 , a2 belong to {0, 2}, etc.) Using this fact, we can define a function F : C → [0, 1]
P∞
by taking the base 3 expansion x = n=1 an 3−n , setting bn = an /2, and putting
P∞ −n
F (x) = n=1 bn 2 . (The ternary string of 0’s and 2’s is sent to the binary string of

0’s and 1’s.) If x, y ∈ C and x < y, then F (x) < F (y) unless x, y are the endpoints

of a deleted interval, in which case F (x) = p2−k for some integers p and k, and

F (x) and F (y) are the two base 2 expansions of this number. We can then extend

F to have this constant value on the deleted interval (x, y). The resulting F is
50 MAA6616 COURSE NOTES FALL 2016

monotone and maps [0, 1] onto [0, 1]. Since F is onto and monotone, it has no jump

discontinuities, and again by monotonicity, F is continuous. This function is called

the Cantor-Lebesgue function, or in some books the Devil’s Staircase. Finally, if we

extend F to be 0 for x < 0 and 1 for x > 1, we can form a Lebesgue-Stieltjes measure

µF supported on C (that is, µF (E) = 0 if E ∩ C = ∅ equivalent µ(C c ) = 0). This

measure is called the Cantor measure. It is said to be singular because it is supported

on a set of Lebesgue measure 0 (see Problem 7.30). It will be an important example

T
of what is called a singular continuous measure on R.

4
AF
One can prove that the Lebesgue-Stieltjes measures µF have similar regularity prop-

erties as Lebesgue measure; since the proofs involve no new ideas they are left as exer-
DR

cises.

Lemma 6.3. Let µF be a Lebesgue-Stieltjes measure. If E ⊂ R is a Borel set, then

∞
X ∞
[
µF (E) = inf{ µF (an , bn ) : E ⊂ (an , bn )}
n=1 n=1

Proof. Problem 7.25. 

 MAA6616 COURSE NOTES FALL 2016 51

Theorem 6.4. Let µF be a Lebesgue-Stieltjes measure. If E ⊂ R is a Borel set, then

µF (E) = inf{µF (U ) : E ⊂ U, U open}

= sup{µF (K) : K ⊂ E, K compact}

Proof. Problem 7.26. 

7. Problems

T
Problem 7.1. Let X = {0, 1, 2, 3} and let

N = ∅, X, {0, 1}, {0, 2}, {0, 3}, {2, 3}, {1, 3}, {1, 2} .

AF
Verify that N is closed under complements and countable disjoint unions, but is not a

σ-algebra.
DR

Problem 7.2. Prove the “exercise” claims in Example 1.8.

Problem 7.3. (a) Let X be a set and let A = (An )∞

n=1 be a sequence of disjoint,

nonempty subsets whose union is X. Prove that the set of all finite or countable

unions of members of A (together with ∅) is a σ-algebra. (A σ-algebra of this type

is called atomic.)

(b) Prove that the Borel σ-algebra BR is not atomic. (Hint: there exists an uncountable

family of mutually disjoint Borel subsets of R.)

Problem 7.4. Can a σ-algebra be, as a set, countably infinite?

52 MAA6616 COURSE NOTES FALL 2016

Problem 7.5. a) Prove Proposition 1.14. (First prove that every open set U is a union

of dyadic intervals. To get disjointness, show that for each point x ∈ U there is a unique

largest dyadic interval I such that x ∈ I ⊂ U .) b) Prove that the dyadic intervals

generate the Borel σ-algebra BR .

Problem 7.6. Fix an integer n ≥ 1. Prove that the set of finite unions of dyadic

subintervals of (0, 1] of length at most 2−n (together with ∅) is a Boolean algebra (of

subsets of (0, 1]).

T
Problem 7.7. Prove that if X, Y are topological spaces and f : X → Y is continuous,
AF
then f is Borel measurable.

Problem 7.8. Let (X, M ) be a measurable space and suppose µ : M → [0, +∞] is
DR

a finitely additive measure that satisfies item (c) of Theorem 2.3. Prove that µ is a

measure.

Problem 7.9. Prove that a countably infinite sum of measures is a measure (Exam-

ple 2.2(d)). You will need the following fact from elementary analysis: if (amn )∞
m,n=1 is a

P∞ P∞ P∞ P∞
doubly indexed sequence of nonnegative reals, then n=1 m=1 amn = m=1 n=1 amn .

Indeed, prove the claims about sums over countable sets appearing after Proposition 3.4.

Problem 7.10. Let A be an atomic σ-algebra generated by a partition (An )∞

n=1 of a

set X (see Problem 7.3).

 MAA6616 COURSE NOTES FALL 2016 53

(a) Fix n ≥ 1. Prove that the function δn : A → [0, 1] defined by

(
1 if An ⊂ A
δn (A) =
0 if An 6⊂ A

is a measure on A .

(b) Prove that if µ is any measure on (X, A ), then there exists a unique sequence (cn )

with each cn ∈ [0, +∞] such that

∞
X
µ(A) = cn δn (A)
n=1

for all A ∈ A .

T
Problem 7.11. Let E∆F denote the symmetric difference of subsets E and F of a set
AF
X,

E∆F := (E \ F ) ∪ (F \ E) = (E ∪ F ) \ (E ∩ F ).

Let (X, M , µ) be a measure space. Prove the following:

DR

(a) If E, F ∈ M and µ(E∆F ) = 0 then µ(E) = µ(E ∩ F ) = µ(F ).

(b) Define E ∼ F if and only if µ(E∆F ) = 0. Show ∼ is an equivalence relation on M .

(c) Assume now that µ is a finite measure. For E, F ∈ M define d(E, F ) = µ(E∆F ).

Show d defines (determines) a metric on the set of equivalence classes M / ∼.

Problem 7.12. Let X be a set. For a sequence of subsets (En ) of X, define

∞ [
\ ∞ ∞ \
[ ∞
lim sup En = En , lim inf En = En .
N =1 n=N N =1 n=N

a) Prove that lim sup 1En = 1lim sup En and lim inf 1En = 1lim inf En (thus justifying the

names). Conclude that En → E pointwise if and only if lim sup En = lim inf En = E.
54 MAA6616 COURSE NOTES FALL 2016

(Hint: for the first part, observe that x ∈ lim sup En if and only if x lies in infinitely

many of the En , and x ∈ lim inf En if and only if x lies in all but finitely many En .)

b) Prove that if the En are measurable, then so are lim sup En and lim inf En . De-

duce that if (En ) converges to E pointwise and all the En are measurable, then E is

measurable.

Problem 7.13. [Fatou theorem for sets] Let (X, M , µ) be a measure space, and let

T
(En ) be a sequence of measurable sets.

a) Prove that
AF
µ(lim inf En ) ≤ lim inf µ(En ). (23)

S∞
b) Assume in addition that µ( n=1 En ) < ∞. Prove that
DR

µ(lim sup En ) ≥ lim sup µ(En ). (24)

c) Prove the following stronger form of the dominated convergence theorem for sets:

suppose (En ) is a sequence of measurable sets, and there is a measurable set F ⊂ X

such that En ⊂ F for all n and µ(F ) < ∞. Prove that if (En ) converges to E pointwise,

then (µ(En )) converges to µ(E). Give an example to show the finiteness hypothesis on

F cannot be dropped.

(For parts (a) and (b), use Theorem 2.3.)

Problem 7.14. Complete the proof of Theorem 2.8.

 MAA6616 COURSE NOTES FALL 2016 55

Problem 7.15. Complete the proof of Theorem 4.4.

Problem 7.16. Given an example of a measurable function f : X → Y between measure

spaces and a subset E ⊂ X such that f (E) is not measurable.

Problem 7.17. Prove the following dyadic version of Theorem 4.6: If m(E) < ∞ and

 > 0, there exists an integer n ≥ 1 and a set A, that is a finite union of dyadic intervals

of length 2−n , such that m(E∆A) < . (This result says, loosely, that measurable sets

T
look “pixelated” at sufficiently fine scales.)
AF
Problem 7.18. a) Prove the following strengthening of Theorem 4.6: if E ⊂ R and

m∗ (E) < ∞, then E is Lebesgue measurable if and only if for every  > 0, there exists
SN
a set A = n=1 In (a finite union of open intervals) such that m∗ (E∆A) < .
DR

b) State and prove a dyadic version of the theorem in part (a).

Problem 7.19. Prove the claims made about the Fat Cantor set in Example 4.10.

Problem 7.20. Let A ⊂ 2R be the Boolean algebra generated by the half-open intervals

(a, b]. For A ∈ A , let µ0 (A) = +∞ if A is nonempty and µ0 (∅) = 0.

(a) Prove that µ0 is a premeasure. If µ is the Hahn-Kolmogorov extension of µ0 and

E ⊂ R is a nonempty Borel set, prove that µ(E) = +∞.

(b) Prove that if µ0 is counting measure on (R, BR ), then µ0 is an extension of µ0 different

from µ.
56 MAA6616 COURSE NOTES FALL 2016

Here is a variant of this example. Let A ⊂ 2Q denote the Boolean algebra generated

by the half-open intervals (a, b] (intersect with Q of course). Note that the σ-algebra

generated by A is 2Q . For A ∈ A, let µ0 (A) = +∞ if A is nonempty and µ0 (∅) = 0.

Show µ0 is a premeasure and its Hahn-Kolmogorov extension µ to 2Q is given by µ(E) = 0

if E = ∅ and µ(E) = ∞ otherwise. Show counting measure c is another extension of µ0

to 2Q . In particular, counting measure c is a σ-finite measure on 2Q , but the premeasure

obtained by restricting c to A is not σ-finite.

T
Problem 7.21. Suppose (X, M , µ) is a measure space and A ⊂ 2X is a Boolean algebra

that generates M and that there is a sequence (An ) from A such that µ(An ) < ∞ and
AF
∪An = X. Prove that if E ∈ M and µ(E) < ∞, then for every  > 0 there exists a set

A ∈ A such that µ(E∆A) < . (Hint: let µ0 be the premeasure obtained by restricting

µ to A . One may then assume that µ is equal the Hahn-Kolmogorov extension of µ0 .

DR

(Why?))

Problem 7.22. Prove that if µ is a locally finite Borel measure and F is defined by

(21), then F is nondecreasing and right-continuous. (Note, once it has been shown that

F is nondecreasing, all one sided limits of F exist. The only issue that remains is the

value of these limits.)

Problem 7.23. Let µF be a Lebesgue-Stieltjes measure. Write F (a− ) := limx→a− F (x).

Prove that

(a) µF ({a}) = F (a) − F (a− ),

 MAA6616 COURSE NOTES FALL 2016 57

(b) µF ([a, b)) = F (b− ) − F (a− ),

(c) µF ([a, b]) = F (b) − F (a− ), and

(d) µF ((a, b)) = F (b− ) − F (a).

Problem 7.24. Complete the proof of Proposition 6.1

Problem 7.25. Prove Lemma 6.3.

T
Problem 7.26. Prove Theorem 6.4. (Use Lemma 6.3.)
AF
Problem 7.27. Let E ⊂ R measurable and m(E) > 0.

(a) Prove that for each 0 < α < 1, there is an open interval I such that m(E ∩ I) >

α m(I).
DR

(b) Show that the set E − E = {x − y : x, y ∈ E} contains an open interval centered at

0. (Choose I as in part (a) with α > 3/4; then E − E contains (−m(I)/2, m(I)/2).)

Problem 7.28. This problem gives another construction of a set E ⊂ R that is not

Lebesgue measurable.

(a) Prove that there is a subset E ⊂ Qc such that for each x ∈ Qc exactly one of x or

−x is in E and, for all rational numbers q, E + q = E. Suggestion: Well order the

irrationals by say ≺ and let E denote the set of those irrational numbers x such that

min(x + Q) ≺ min(−x + Q).

58 MAA6616 COURSE NOTES FALL 2016

(b) Prove that any set E with the properties above (for x ∈ Qc exactly one of x or −x

is in E and E + q = E for all q ∈ Q) is not Lebesgue measurable. (Hint: suppose it

is. Prove, for every interval I with rational endpoints, m(E ∩ I) = 12 |I| and apply

part (a) of Problem 7.27.)

Problem 7.29. Let E be the nonmeasurable set described in Example 4.11.

(a) Show if F ⊂ R is (Lebesgue) measurable, bounded and (F + q) ∩ (F + r) = ∅ for

T
distinct rationals q, r, then m(F ) = 0.

(b) Show that if q ∈ Q, F ⊂ E + q and F is Lebesgue measurable, then m(F ) = 0.

AF
(c) Prove that if G ⊂ R has positive measure, then G contains a nonmeasurable subset.

(Observe G = ∪q∈Q G ∩ (E + q).)

DR

Problem 7.30. Suppose µ is a regular Borel measure on a compact Hausdorff space and

µ(X) = 1. Let O denote the collection of µ-null open subsets of X and let U = ∪O∈O O.

Prove U is also µ-null. Hence U is the largest µ-null subset of X. Prove there exists a

smallest compact subset K of X such that µ(K) = 1. The set K is the support of µ.

Problem 7.31. Given a set X and a subset ρ ⊂ 2X , there is a smallest topology τ on

X containing ρ, called the topology generated by ρ. (Proof idea: 2X is a topology on X

and the intersection of topologies is also a topology.) Let N be a positive integer and
Q
(Xj , τj ) for 1 ≤ j ≤ N be topological spaces. The product topology π on X = Xj is
 MAA6616 COURSE NOTES FALL 2016 59

the topology generated by the open rectangles; i.e., by the set

ρ = {×N X
1 Uj = U1 × · · · × UN : Uj ∈ τj } ⊂ 2 .

Observe that each of the projection maps πj : X → Xj is continuous. Prove, if every

each open set W in the product topology on X is an at most countable union from ρ,

then ⊗BXj = BX ; i.e., the product of the Borel σ algebras on the Xj is the same as

the Borel sigma algebra on X given the product topology.

T
Problem 7.32. Give a proof of Theorem 4.6 based upon Theorem 4.8.
AF
Problem 7.33. Prove, if X is a compact metric space, then every compact (closed)

set in X is a Gδ and likewise every open set an Fσ . Prove, a finite Borel measure on a

compact metric space is regular.

DR
60 MAA6616 COURSE NOTES FALL 2016

8. Measurable functions

We will state and prove a few “categorical” properties of measurable functions be-

tween general measurable spaces, however in these notes we will mostly be interested in

functions from a measurable space taking values in the extended positive axis [0, +∞],

the real line R, or the complex numbers C.

Definition 8.1. Let (X, M ) and (Y, N ) be measurable spaces. A function f : X → Y

is called measurable (or (M , N ) measurable) if f −1 (E) ∈ M for all E ∈ N . A

T
function f : X → R is measurable if it is (M , BR ) measurable unless indicated otherwise.
AF
Likewise, a function f : X → C is measurable if it is (M , BC ) measurable (where C is

identified with R2 topologically). /

It is immediate from the definition that if (X, M ), (Y, N ), (Z, O) are measurable
DR

spaces and f : X → Y, g : Y → Z are measurable functions, then the composition

g ◦ f : X → Z is measurable. The following is a routine application of Proposition 1.9.

The proof is left as an exercise.

Proposition 8.2. Suppose (X, M ) and (Y, N ) are measurable spaces and the collection

of sets E ⊂ 2Y generates N as a σ-algebra. Then f : X → Y is measurable if and only

if f −1 (E) ∈ M for all E ∈ E . †

Proof. Suppose f −1 (E) ∈ M for all E ∈ E . Let Ωf = {E ⊂ Y : f −1 (E) ∈ M }. Thus

Ωf contains E by assumption. Moreover, Ωf is a σ-algebra (the pushforward σ-algebra).

 MAA6616 COURSE NOTES FALL 2016 61

Since Ωf is a σ-algebra containing E , it follows that N = M (E ) ⊂ Ωf . Hence f is

measurable. 

Corollary 8.3. Let X, Y be topological spaces equipped with their Borel σ-algebras

BX , BY respectively. Every continuous function f : X → Y is (BX , BY )-measurable

(or Borel measurable for short). In particular, if f : X → F is continuous and X is

given its Borel σ-algebra, then f is measurable, where F is either R or C, †

Proof. Since the open sets U ⊂ Y generate BY and f −1 (U ) is open (hence in BX ) by

T
hypothesis, this corollary is an immediate consequence of Proposition 8.2. 
AF
Definition 8.4. Let F = R or C. A function f : R → F is called Lebesgue measurable

(resp. Borel measurable) if it is (L , BF ) (resp. (BR , BF )) measurable. Here L is the

Lebesgue σ-algebra. /
DR

Remark 8.5. Note that since BR ⊂ L , being Lebesgue measurable is a weaker con-

dition than being Borel measurable. If f is Borel measurable, then f ◦ g is Borel or

Lebesgue measurable if g is. However if f is only Lebesgue measurable, then f ◦ g need

not be Lebesgue measurable, even if g is continuous. (The difficulty is that we have

no control over g −1 (E) when E is a Lebesgue set.) A counterexample is described in

Problem 13.7. 

It will sometimes be convenient to consider functions that are allowed to take the

values ±∞.
62 MAA6616 COURSE NOTES FALL 2016

Definition 8.6. [The extended real line] Let R denote the set of real numbers together

with the symbols ±∞. The arithmetic operations + and · can be (partially) extended

to R by declaring

±∞ + x = x + ±∞ = ±∞

for all x ∈ R,

+∞ · x = x · +∞ = +∞

for all nonzero x ∈ (0, +∞) (and similar rules for the other choices of signs),

T
0 · ±∞ = ±∞ · 0 = 0,
AF
The order < is extended to R by declaring

−∞ < x < +∞
DR

for all x ∈ R. /

The symbol +∞ + (−∞) is not defined, so some care must be taken in working

out the rules of arithmetic in R. Typically we will be performing addition only when

all values are finite, or when all values are nonnegative (that is for x ∈ [0, +∞]). In

these cases most of the familiar rules of arithmetic hold (for example the commutative,

associative, and distributive laws), and the inequality ≤ is preserved by multiplying

both sides by the same quantity. However cancellation laws are not in general valid when

infinite quantities are permitted; in particular from x·+∞ = y·+∞ or x++∞ = y++∞

one cannot conclude that x = y.

 MAA6616 COURSE NOTES FALL 2016 63

The order property allows us to extend the concepts of supremum and infimum, by

defining the supremum of a set that is unbounded from above, or set containing +∞, to
P
be +∞; similarly for inf and −∞. This also means every sum n xn with xn ∈ [0, +∞]

can be meaningfully assigned a value in [0, +∞], namely the supremum of the finite
P
partial sums n∈F xn .

The collection of sets U ⊂ R such that either U is an open subset of R or U is the

union of an open set in R with a interval of the form (a, ∞] and/or of the form [−∞, b)

T
is a topology on R and, of course, we refer to these sets as open (in R). Similarly, the

collection of open sets in R together with open sets in R union an interval of the form
AF
(a, ∞] is a topology on (−∞, ∞].

Definition 8.7. [Extended Borel σ-algebra] The extended Borel σ-algebra over R is the
DR

σ-algebra over generated by open sets of R and is denoted BR . Similarly B(−∞,∞] is the

Borel σ-algebra on (−∞, ∞]. /

Proposition 8.8. The collection E = {(a, ∞] : a ∈ R} generates BR . Similarly each of

the collections Ej from Proposition 1.12 generates B(−∞,∞] . †

Proof sketch. Since (b, ∞]c = [−∞, b] (complement in R), it follows that M (E ) contains

the (finite) intervals of the form (a, b]. Hence, from Proposition 1.12, M (E ) contains

all open interval in R and hence all open sets in R. Similarly E contains the intervals

[−∞, b). The first part of the Proposition now follows.

64 MAA6616 COURSE NOTES FALL 2016

For the second part of the Proposition, since (−∞, a)c = [a, ∞] and since (a, ∞] =

∪∞ 1
n=1 [a − n , ∞] it follows that σ-algebra of subsets of (−∞, ∞] generated by the open

intervals in R contains the intervals of the form (a, ∞]. It now follows that easily that

open the open intervals (in R) generate B(−∞,∞] . 

Definition 8.9. [Measurable function] Let (X, M ) be a measurable space. A function

f : X → R is called measurable if it is (M , BR ) measurable; that is, if f −1 (U ) ∈ M for

T
every open set U ⊂ R. The notion of measurable functions f : X → (−∞, ∞] is defined

similarly. /
AF
In particular, the following criteria for measurability will be used repeatedly.
DR

Corollary 8.10 (Equivalent criteria for measurability). Let (X, M ) be a measurable

space.

(a) A function f : X → R is measurable if and only if the sets

f −1 ((t, +∞]) = {x : f (x) > t}

are measurable for all t ∈ R; and

(b) A function f : X → R or f : X → (−∞, ∞] is measurable if and only if f −1 (E) ∈ M

for all E ∈ E , where E is any of the collections of sets Ej appearing in Proposi-

tion 1.12.
 MAA6616 COURSE NOTES FALL 2016 65

(c) A function f : X → C is measurable if and only if f −1 ((a, b) × (c, d)) is measurable

for every a, b, c, d ∈ R. (Here (a, b) × (c, d) is identified with the open box {z ∈ C :

a < Re(z) < b, c < Im(z) < d}.)

In item (a), t ∈ R can be replaced by t ∈ Q.

Proof sketch. Combine Propositions 8.8 and 8.2. 

T
Example 8.11. [Examples of measurable functions]
AF
(a) An indicator function 1E is measurable if and only if E is measurable. Indeed, the

set {x : 1E (x) > t} is either empty, E, or all of X, in the cases t ≥ 1, 0 ≤ t < 1, or

t < 0, respectively.
DR

(b) The next series of propositions will show that measurability is preserved by most of

the familiar operations of analysis, including sums, products, sups, infs, and limits

(provided one is careful about arithmetic of infinities).

(c) Corollary 8.21 below will show that examples (a) and (b) above in fact generate

all the examples in the case of R or C valued functions. That is, every measurable

function is a pointwise limit of linear combinations of measurable indicator functions.

Proposition 8.12. Let (X, M ) be a measurable space. A function f : X → C is

measurable if and only if Ref and Imf are measurable. †

66 MAA6616 COURSE NOTES FALL 2016

Proof. As a topological space, C2 is R2 and the Borel σ-algebra of R2 is generated by

open rectangles (a, b) × (c, d). Suppose f : X → C is measurable. The real part u of

f is measurable since it is the composition u = π1 ◦ f , of the continuous (hence Borel

measurable) projection π1 of R2 onto the first coordinate with the measurable function

f . Likewise the imaginary part v of f is measuarable.

Conversely, suppose u, v are measurable. Fix an open rectangle R = (a, b) × (c, d)

and note that

T
f −1 (R) = u−1 ((a, b)) ∩ v −1 ((c, d)),

which lies in M by hypothesis. So f is measurable by Corollary 8.10. 

AF
Proposition 8.13. Suppose (X, M ) is a measurable space and let F denote any of R,

(−∞, ∞] and R. If f : X → F is measurable, then so is −f . †

DR

Proof. In any case it suffices to prove Et = {−f > t} is measurable for each t ∈ R. We

have Et = {f ≤ −t} = {f > −t}c is measurable. 

Proposition 8.14. Let (fn ) be a sequence of R-valued measurable functions.

(a) The functions

sup fn , inf fn , lim sup fn , lim inf fn

n→∞ n→∞

are measurable;

(b) The set on which (fn ) converges is a measurable set; and

(c) If (fn ) converges to f pointwise, then f is measurable.

 MAA6616 COURSE NOTES FALL 2016 67

Proof. Let f (x) = supn fn (x). Given t ∈ R, we have f (x) > t if and only if fn (x) > t for

some n. Thus
∞
[
{x : f (x) > t} = {x : fn (x) > t}.
n=1

It follows that f is measurable. Likewise inf fn is measurable, since inf fn = − sup(−fn )

and two applications of Proposition 8.13. Consequently, gN = supn≥N fn is measurable

T
for each positive integer N and hence lim sup fn = inf gN is also measurable.
AF
If (fn ) converges pointwise to f , then f = lim sup fn = lim inf fn is measurable.

Part (b) is left as an exercise. 

In the Proposition 8.14 it is of course essential that the supremum is taken only over
DR

a countable set of measurable functions; the supremum of an uncountable collection of

measurable functions need not be measurable. Problem 13.6 asks for a counterexample.

Theorem 8.15. Let (X, M ) be a measurable space.

(a) If f, g : X → C are measurable functions, and c ∈ C. Then cf , f + g, and f g are

measurable.

(b) If f, g : X → [−∞, ∞] are measurable and, for each x, {f (x), g(x)} =

6 {±∞}, then

f + g is measurable.

(c) If f, g : X → [−∞, ∞] are measurable, then so is f g.

68 MAA6616 COURSE NOTES FALL 2016

Proof. To prove (b), supppose f, g : X → [ − ∞, ∞] are measurable and f + g is defined.

Using Corollary 8.10 (a), it suffices to prove, for a t ∈ R, that

[
{x ∈ X : f (x) + g(x) > t} = {x : f (x) > q} ∩ {x : g(x) > t − q},
q∈Q

since all the sets in the last line are measurable, the intersection is finite and the union

countable. The inclusion of the set on the right into the set on the left is evident.

Suppose f (x) + g(x) > t. In particular, g(x) 6= −∞ and thus f (x) > t − g(x). There is

T
a rational q ∈ Q such that f (x) > q > t − g(x) and the reverse inclusion follows.
Assuming f, g : X → [0, ∞] are measurable, a proof that f g is measurable can be
AF
modeled after the proof for f +g. The details are left as an exercise (Problem 13.8). From
Proposition 8.14, if f : X → [−∞, ∞] is measurable, then so are f + (x) = max{f (x), 0}
and f − (x) = − min{f (x), 0}. Of course f = f + − f − . Now suppose f, g : X → R. Let
F = f + g + +f − g − and G = −f − g + −f + g − and note that f g = F +G. Since F and G are
measurable and f ± , g ± take values in [0, ∞) and are measurable all the products f ± g ±
are measurable. Hence, using (b) several times and Proposition 8.13, F + G, and thus
f g is measurable. Finally, suppose now that f, g : X → R. Let Ω±∞ = (f g)−1 (±∞)
DR

and Ω = (f g)−1 (R). In particular,

Ω∞ = ({f = ∞} ∩ {g > 0}) ∪ ({f = −∞} ∪ {g < 0}) ∪ ({g = ∞} ∩ {f > 0}) ∪ ({g = −∞} ∩ {f < 0}) ∈ M .

Likewise Ω−∞ is measurable and therefore Ω is measurable too. Let Ωf = f −1 (R) and

Ωg = g −1 (R). Both are measurable. Let fˆ = f 1Ωf and ĝ = g1Ωg . It is easily checked that

both are measurable. Given x ∈ Ω either f (x), g(x) ∈ R or f (x) = ±∞ and g(x) = 0 or

g(x) = ±∞ and f (x) = 0. In each case it is readily verified that f (x)g(x) = fˆ(x)ĝ(x).

Hence,

f g1Ω = fˆ ĝ.
 MAA6616 COURSE NOTES FALL 2016 69

Since fˆ and ĝ are measurable and real-valued their product is measurable and thus f g1Ω

is measurable. Finally, since

f g = ∞1Ω∞ − ∞1Ω−∞ + (f g)1Ω ,

an application of item (b) completes the proof.

The proof of (a) is straightforward using parts (b) and (c) and Proposition 8.12.

T
Given a measure space (X, M , µ) a property P = P (x) is said to hold almost

everywhere with respect to µ, abbreviated a.e. µ, or just a.e. when the measure µ is
AF
understood from context, if the set of points x where P (x) does not hold is measurable

and has measure zero. In the case the measure space is complete, a property holds a.e.

if and only if the set where it doesn’t hold is a subset of a set of measure zero.
DR

Proposition 8.16. Suppose (X, M , µ) is a complete measure space and (Y, N ) is a

measurable space.

(a) Suppose f, g : X → Y. If f is measurable and g = f a.e., then g is measurable.

(b) If fn : X → R are measurable functions and fn → f a.e., then f is measurable. The

same conclusion holds if R is replaced by C.

Proposition 8.17. Let (X, M , µ) be a measure space and (X, M , µ) its completion.

Let F denote either R, R or C.

70 MAA6616 COURSE NOTES FALL 2016

(i) If f : X → F is a M -measurable function, then there is an M -measurable function

g such that µ({x : f (x) 6= g(x)}) = 0.

(ii) If (fn ) is a sequence of M measurable functions, fn : X → F, that converges a.e. µ

to a function f , then there is a M measurable function g such that (fn ) converges

a.e. µ to g.

T
The proofs of Propositions 8.16 and 8.17 are left to the reader as Problem 13.9.
AF
Definition 8.18. [Unsigned simple function] Recall, a function f on a set X is unsigned

if its codomain is a subset of [0, ∞]. An unsigned function s : X → [0, +∞] is called

simple if its range is a finite set. /

DR

Many statements about general measurable functions can be reduced to the unsigned

case. For instance, one simple but important application of Proposition 8.14 is that if

f, g are R-valued measurable functions, then f ∧ g := min(f, g) and f ∨ g := max(f, g)

are measurable; in particular f + := max(f, 0) and f − := − min(f, 0) are measurable

if f is. It also follows that |f | := f + + f − is measurable when f is. Together with

Proposition 8.12, these observations show every C valued measurable function f is a

linear combination of four unsigned measurable functions (the positive and negative

parts of the real and imaginary parts of f ).

 MAA6616 COURSE NOTES FALL 2016 71

A partition P of the set X is, for some n ∈ N, a set P = {E0 , . . . , En } of pairwise

disjoint subsets of X whose union is X. If each Ej is measurable, then P is a measurable

partition.

Proposition 8.19. Suppose s is an unsigned function on X. The following are equiva-

lent.

(i) s is a (measurable) simple function;

T
(ii) there exists an n, scalars c1 , . . . , cn ∈ [0, ∞] and (measurable) subsets Fj ⊂ X such

that
AF n
X
s= cj 1Fj ;
j=1

(iii) there exists a (measurable) partition P = {E1 , . . . , Em }, and c1 , . . . , cm in [0, ∞]

such that
DR

m
X
s= ck 1Ek .
k=1

The proof of this proposition is an easy exercise. Letting {c1 , c2 , . . . , cm } denote the

range of s,
n
X
s= cj Ej ,
j=1

where Ej = s−1 ({cj }). Evidently {E1 , . . . , En } is a partition of X that is measurable if

s is measurable. This is the standard representation of s.

72 MAA6616 COURSE NOTES FALL 2016

Theorem 8.20 (The Ziggurat approximation). Let (X, M ) be a measurable space. If

f : X → [0, +∞] is an unsigned measurable function, then there exists a increasing

sequence of unsigned, measurable simple functions sn : X → [0, +∞) such that sn → f

pointwise increasing on X. If f is bounded, the sequence can be chosen to converge

uniformly.

k
Proof. For positive integers n and integers 0 ≤ k < n2n , let En,k = {x : 2n
< f (x) ≤

k+1
}, let En,n2n = {x : n < f (x)} and define

T
2n

n2n
X k
sn (x) = 1E . (25)
2n n,k
AF k=0

Verify that (sn ) is pointwise increasing with limit f and if f is bounded, then the

convergence is uniform. 
DR

It will be helpful to record for future use the round-off procedure used in this proof.

Let f : X → [0, +∞] be an unsigned function. For any  > 0, if 0 < f (x) < +∞ there

is a unique integer k such that

k < f (x) ≤ (k + 1).

Define the “rounded down” function f (x) to be k when f (x) ∈ (0, +∞) and equal to

0 or +∞ when f (x) = 0 or +∞ respectively. Similarly we can defined the “rounded

up” function f  to be (k + 1), 0, or +∞ as appropriate. (So, in the previous proof, the

function gn was f1/n .) In particular, for  > 0

f ≤ f ≤ f  ,
 MAA6616 COURSE NOTES FALL 2016 73

and f , f  are measurable if f is. Moreover the same argument used in the above proof

shows that f , f  → f pointwise as  → 0.

Finally, by the remarks following Proposition 8.14, the following corollary is imme-

diate (since its proof reduces to the unsigned case):

Corollary 8.21. Every R- or C-valued measurable function is a pointwise limit of mea-

surable simple functions. †

T
9. Integration of simple functions
AF
We will build up the integration theory for measurable functions in three stages. We

first define the integral for unsigned simple functions, then extend it to general unsigned

functions, and finally to general (R or C-valued) functions. Throughout this section and
DR

the next, we fix a measure space (X, M , µ); all functions are defined on this measure

space.

Suppose P = {E0 , . . . , En } is a measurable partition of X, c0 , c1 , . . . , cn ≥ 0 and

n
X
s= cj 1Ej . (26)
j=0

If Q = {F0 , . . . , Fm } is another measurable partition, d0 , d1 , . . . , dn ≥ 0 and

m
X
s= dk 1Fk ,
k=0

then it is an exercise (see Problem 13.10) to show

n
X m
X
cn µ(En ) = dm µ(Fm ).
j=0 k=0
74 MAA6616 COURSE NOTES FALL 2016

Indeed, for this exercise it is helpful to consider the common refinement {Ej ∩ Fk : 1 ≤

j ≤ n, 1 ≤ k ≤ m} of the partitions P and Q. It is now possible to make the following

definition.

PN
By convention, when writing a simple measurable function s as s = n=0 cn 1En the

sets En are assumed measurable.

Definition 9.1. Let (X, M , µ) be a measure space and f =

PN
n=0 cn 1En an unsigned

T
measurable simple function. The integral of f (with respect to the measure µ) is defined

to be
AF Z N
X
f dµ := cn µ(En ).
X n=0

/
DR

One thinks of the graph of the function c1E as “rectangle” with height c and “base”

E; since µ tells us how to measure the length of E the quantity c · µ(E) is interpreted

as the “area” of the rectangle. This intuition can be made more precise once we have

proved Fubini’s theorem. Note too that the definition explains the convention 0 · ∞ = 0,

since the set on which s is 0 should not contribute to the integral.

Let L+ denote the set of all unsigned measurable functions on (X, M ). We begin

by collecting some basic properties of the integrals of simple functions. When X and µ
R R
are understood we drop them from the notation and simply write f for X
f dµ.
 MAA6616 COURSE NOTES FALL 2016 75

Theorem 9.2 (Basic properties of simple integrals). Let (X, M , µ) be a measure space

and let f, g ∈ L+ be simple functions.

R R
(a) (Homogeneity) If c ≥ 0, then cf = c f .
R R
(b) (Monotonicity) If f ≤ g, then f≤ g.
R R R
(c) (Finite additivity) f +g = f+ g.

(d) (Almost everywhere equivalence) If f (x) = g(x) for µ-almost every x ∈ X, then
R R
f= g.

T
R
(e) (Finiteness) f < +∞ if and only if is finite almost everywhere and supported on
AF
a set of finite measure.
R
(f ) (Vanishing) f = 0 if and only if f = 0 almost everywhere.

Proof. (a) is trivial; we prove (b) and (c) and leave the rest as (simple!) exercises.
DR

Pn Pm
To prove (b), write f = j=0 cj 1Ej and g = k=0 dk 1Fk for measurable partitions

P = {E0 , . . . , En } and Q = {F0 , . . . , Fm } of X. It follows that R = {Ej ∩ Fk : 0 ≤ j ≤

n, 0 ≤ k ≤ m} is a measurable partition of X too and

X
f= cj 1Ej ∩Fk
j,k

and similarly for g. From the assumption f ≤ g we deduce that cj ≤ dk whenever

Ej ∩ Fk 6= ∅. In particular, either cj ≤ dk or µ(Ej ∩ Fk ) = 0. Thus,

Z X X Z
f= cj µ(Ej ∩ Fk ) ≤ dk µ(Ej ∩ Fk ) = g.
j,k j,k
76 MAA6616 COURSE NOTES FALL 2016
Sm Sn
For item (c), since Ej = k=0 Ej ∩ Fk for each j and Fk = j=0 Fk ∩ Ej for each k,

it follows from the finite additivity of µ that

Z Z X
f+ g= (cj + dk )µ(Ej ∩ Fk ).
j,k
P
Since f +g = j,k (cj +dk )1Ej ∩Fk , and {Ej ∩Fk : 1 ≤ j ≤ n, 1 ≤ k ≤ m} is a measurable
R
partition, the right hand side is (f + g). 

If f : X → [0, +∞] is a measurable simple function, then so is 1E f for any measur-

R R

T
able set E. We write E
f dµ := 1E f dµ.

Proposition 9.3. Let (X, M , µ) be a measure space. If f is an unsigned measurable

AF
simple function, then the function ν : M → [0, ∞] defined by
Z
ν(E) := f dµ
E
DR

is a measure on (X, M ). †

Pm
Sketch of proof. Write f as j=1 cj 1Fj with respect to a measurable partition {F1 , . . . , Fm }

and observe, for E ∈ M ,

Z m
X
ν(E) = 1E f dµ = cj µ(Fj ∩ E).
j=0

For a fixed measurable set F , the mapping νF : M → [0, ∞] defined by τF (E) =

µ(E ∩ F ) is a measure by Example 2.2 item c. Given n ∈ N, numbers c1 , . . . , cm ≥ 0

and measurable sets F1 , . . . , Fm , the mapping τ : M → [0, ∞] defined by

X X
τ (E) = cj τFj (E) = cj µ(E ∩ Fj )
 MAA6616 COURSE NOTES FALL 2016 77

is a measure by Example 2.2 item d. The resutl follows. 

10. Integration of unsigned functions

We now extend the definition of the integral to all (not necessarily simple) functions

in L+ . First note that if (X, M , µ) is a measure space and s is a measurable unsigned

simple function, then, by Theorem 9.2(b),

Z Z
s dµ = sup{ t dµ : 0 ≤ s ≤ t, t is a measurable unsigned simple function}.
X X

T
Hence, the following definition is consistent with that of the integral for unsigned simple

functions.
AF
Definition 10.1. Let (X, M , µ) be a measure space. For an unsigned measurable

function f : X → [0, +∞], define the integral of f with respect to µ by

DR

Z Z
f dµ := sup{ s dµ : 0 ≤ s ≤ f ; s simple and measurable} (27)
X X
R R
Often we write f instead of X
f dµ when µ and X are understood.

For E ∈ M let
Z Z
f dµ = f 1E dµ.
E X

Theorem 10.2 (Basic properties of unsigned integrals). Let (X, M , µ) be a measure

space and let f, g be unsigned measurable functions on X.

R R
(a) (Homogeneity) If c ≥ 0 then cf = c f .
78 MAA6616 COURSE NOTES FALL 2016
R R
(b) (Monotonicity) If f ≤ g then f≤ g.

(c) (Almost everywhere equivalence) If f (x) = g(x) for µ-almost every x ∈ X, then
R R
f= g.
R
(d) (Finiteness) If f < +∞, then f (x) < +∞ for µ-a.e. x.
R
(e) (Vanishing) f = 0 if and only if f = 0 almost everywhere.
R
(f ) (Bounded) If f is bounded measurable function and µ(X) < ∞, then f dµ < ∞.

T
The integral is also additive; however the proof is surprisingly subtle and will have

to wait until we have established the Monotone Convergence Theorem.

AF
Proof of Theorem 10.2. As in the simple case homogeneity is trivial. Monotonicity is

also evident, since any simple function less than f is also less than g.
DR

Turning to item (c) let E be a measurable set with µ(E c ) = 0. If s is a simple

function, then 1E s and s are simple functions that agree almost everywhere. Thus
R R
1E s = s, by Theorem 9.2(d). Further, if 0 ≤ s ≤ f, then 1E s ≤ 1E f. Hence, using

monotonicity (item (b)) and taking suprema over simple functions,

Z Z Z Z Z Z
1E f ≤ f = sup s = sup 1E s ≤ sup t= 1E f.
0≤s≤f 0≤s≤f 0≤t≤1E f

Now suppose f = g a.e. Thus, letting E = {f = g}, the set E c has measure zero and

1E f = 1E g. Hence, from what has already been proved (twice),

Z Z Z Z
f dµ = 1E f dµ = 1E g dµ = g dµ.
 MAA6616 COURSE NOTES FALL 2016 79

To prove item (d) observe if f = +∞ on a measurable set E and µ(E) > 0, then
R R R
f ≥ n1E = nµ(E) for all n, so f = +∞. (A direct proof can be obtained from

Markov’s inequality below.)

If f = 0 a.e. and 0 ≤ s ≤ f is a simple function, then s = 0 a.e. and hence, by

R R
Theorem 9.2 item (f) s = 0. Hence f = 0. Conversely, suppose there is a set E of

1
positive measure such that f (x) > 0 for all x ∈ E. Let En = {x ∈ E : f (x) > n
}.
S∞
Then E = n=1 En , so by the pigeonhole principle µ(EN ) > 0 for some N . But then

T
1 1
R
0≤ 1
N EN
≤ f , so f≥ N
µ(EN ) > 0 and item (e) is proved.

Finally, for item (f), by hypothesis there is a positive real number C so that 0 ≤
AF
f (x) ≤ C for x ∈ X. With g denoting the simple function C1X , we have 0 ≤ f ≤ g.

Hence item (f) follows from monotonicity (item (b)). 

DR

Theorem 10.3 (Monotone Convergence Theorem). Let (X, M , µ) be a measure space

and suppose (fn ) is a sequence of unsigned measurable functions fn : X → [0, ∞]. If

R R
(fn ) increases to f pointwise, then fn → f, where f is the pointwise limit of (fn ).

Proof. Since (fn ) converges pointwise to f and each fn is measurable, f is measurable by

Proposition 8.14 item (c). By monotonicity of the integral, Theorem 10.2, the sequence
R R R R
( fn ) is increasing and fn ≤ f for all n. Thus the sequence ( fn ) converges (perhaps
R R
to ∞) and lim fn ≤ f . For the reverse inequality, fix a measurable simple function

with 0 ≤ s ≤ f . Let  > 0 be given. Consider the sets

En = {x : fn (x) ≥ (1 − )s(x)}.
80 MAA6616 COURSE NOTES FALL 2016

Since (fn ) is pointwise increasing, (En ) is an increasing sequence of measurable sets

whose union is X. For all n,

Z Z Z
fn ≥ fn ≥ (1 − ) s.
En En

R
By Monotone convergence for sets (Theorem 2.3(c)) applied to the measure ν(E) = E
s

(Proposition 9.3), we see that

Z Z
lim s= s.
En X
R R R R

T
Thus lim fn ≥ (1 − ) s. Since 1 >  > 0 is arbitrary, lim fn ≥ s. Since 0 ≤ s ≤ f
R
was an arbitrary simple function, lim fn is an upper bound for the set whose supremum
AF
R R R
is, by definition, f . Thus lim fn ≥ f. 

Before going on we mention two frequently used applications of the Monotone Con-

vergence Theorem:
DR

Corollary 10.4. (i) (Vertical truncation) If f is an unsigned measurable function,

R R
then the sequence ( min(f, n)) converges to f .

(ii) (Horizontal truncation) If f is an unsigned measurable function and (En )∞

n=1 is an

R R
increasing sequence of measurable sets whose union is X, then En
f→ f.

Proof. Since min(f, n) and 1En f are measurable for all n and increase pointwise to f ,

these follow from the Monotone Convergence Theorem. 

 MAA6616 COURSE NOTES FALL 2016 81

Theorem 10.5 (Additivity of the unisgned integral). If f, g are unsigned measurable

R R R
functions, then f +g = f+ g.

Proof. By Theorem 8.20, there exist sequences of unsigned, measurable simple functions

fn , gn that increase pointwise to f, g respectively. Thus fn + gn increases to f + g, so by

Theorem 9.2(c) and the Monotone Convergence Theorem,

Z Z Z Z  Z Z
f + g = lim [fn + gn ] = lim fn + gn = f+ g.

T

AF
Corollary 10.6 (Tonelli’s theorem for sums and integrals). If (fn ) is a sequence of
R P∞ P∞ R
unsigned measurable functions, then n=1 fn = n=1 fn . †
DR

PN
Proof. Let gN = n=1 fn . Thus (gN ) is an increasing sequence with pointwise limit
P∞
g= n=1 fn . In particular, g is measurable and by the Monotone convergence theorem
R R
( gN ) converges to g. By induction on Theorem 10.5,

Z N Z
X
gN = fn
n=1

and the result follows by taking the limit on N . 

Pointwise convergence is not sufficient to imply convergence of the integrals (see

Examples 10.8 below), however the following weaker result holds.

82 MAA6616 COURSE NOTES FALL 2016

Theorem 10.7 (Fatou’s Theorem). If fn is a sequence of unsigned measurable functions,

then
Z Z
lim inf fn ≤ lim inf fn .

Proof. For n ∈ N, the function gn (x) := inf m≥n fm (x) is unsigned, gn ≤ fn pointwise and,

by Proposition 8.14, measurable. By definition of the lim inf, the sequence (gn ) increases

pointwise to lim inf fn . By the Monotone Convergence Theorem and monotonicity

T
Z Z Z Z Z
lim inf fn = lim gn = lim gn = lim inf gn ≤ lim inf fn .
AF


Example 10.8. [Failure of convergence of integrals] This example highlights three

DR

R R
modes of failure of the convergence ( fn ) to f for sequences of unsigned measur-

able functions fn : R → [0, +∞] and Lebesgue measure. In each case (fn ) converges to
R
the zero function pointwise, but fn = 1 for all n:

(1) (Escape to height infinity) fn = n1(0, 1 )

n

1
(2) (Escape to width infinity) fn = 1
2n (−n,n)

(3) (Escape to support infinity) fn = 1(n,n+1)

Note that in the second example the convergence is even uniform. These examples can be

though of as moving bump functions. In each case we have a rectangle and can vary the

height, width, and position. If we think of fn as describing a density of mass distributed

 MAA6616 COURSE NOTES FALL 2016 83
R
over the real line, then fn gives the total “mass”; Fatou’s theorem says mass cannot

be created in the limit, but these examples show mass can be destroyed. 4

Proposition 10.9 (Markov’s inequality). If f is an unsigned measurable function, then

for all t > 0

Z
1
µ({x : f (x) > t}) ≤ f
t

T
R
Proof. Let Et = {x : f (x) > t}. Then by definition, t1Et ≤ f , so tµ(Et ) = t1Et ≤
R
f.
AF 

We conclude this section with a few frequently-used corollaries of the monotone

convergence theorem, and a converse to it.

DR

Theorem 10.10 (Change of variables). Let (X, M , µ) be a measure space, (Y, N ) a

measurable space, and φ : X → Y a measurable function. The function φ∗ µ : N →

[0, +∞] defined by

φ∗ µ(E) = µ(φ−1 (E)). (28)

is a measure on (Y, N ), and for every unsigned measurable function f : Y → [0, +∞],
Z Z
f d(φ∗ µ) = (f ◦ φ) dµ. (29)
Y X

Proof. Problem 13.16. The measure φ∗ µ is called the push-forward of µ under φ. 

84 MAA6616 COURSE NOTES FALL 2016

Lemma 10.11 (Borel-Cantelli Lemma). Let (X, M , µ) be a measure space and suppose

(En )∞
n=1 is a sequence of measurable sets. If

∞
X
µ(En ) < ∞,
n=1

then for almost every x ∈ X is contained in at most finitely many of the En (that is,

letting Nx := {n : x ∈ En } ⊂ N, the set {x : |Nx | = ∞} has measure 0). †

P∞ R
Sketch of proof. Consider the series S = n=1 1En . By Tonelli (Corollary 10.6), S is

T
finite. Hence S is finite a.e. by Theorem 10.2(d). On the other hand, {x : |Nx | = ∞} =

S −1 ({∞}. 
AF
There is a sense in which the monotone convergence theorem has a converse, namely

that any map from unsigned measurable functions on a measurable space (X, M ) to
DR

[0, +∞], satisfying some reasonable axioms (including MCT) must come from integration

against a measure. The precise statement is the following:

Theorem 10.12. Let (X, M ) be a measurable space and let U (X, M ) denote the set of

all unsigned measurable functions f : X → [0, +∞]. Suppose L : U (X, M ) → [0, +∞]

is a function obeying the following axioms:

(a) (Homogeneity) For every f ∈ U (X, M ) and every real number c ≥ 0, L(cf ) = cL(f ).

(b) (Additivity) For every pair f, g ∈ U (X, M ), L(f + g) = L(f ) + L(g).

(c) (Monotone convergence) If fn is a sequence in U (X, M ) increasing pointwise to f ,

then limn→∞ L(fn ) = L(f ).

 MAA6616 COURSE NOTES FALL 2016 85

Then there is a unique measure µ : M → [0, +∞] such that L(f ) =

R
X
f dµ for all

f ∈ U (X, M ). In fact, µ(E) = L(1E ).

Proof. Problem 13.17. 

11. Integration of signed and complex functions

Again we work on a fixed measure space (X, M , µ). Suppose f : X → R is mea-

T
surable. Split f into its positive and negative parts f = f + − f − . If at least one of

f + , f − is finite f is semi-integrable and the integral of f is defined as

R R
AF
Z Z Z
f= +
f − f −.

If both are finite, we say f is integrable (or sometimes absolutely integrable). Note that
DR

|f | < +∞; this is immediate since |f | = f + + f − and the

R
f is integrable if and only if

integral is additive on unsigned functions. We write

Z
kf k1 := |f | dµ
X

when f is integrable. In the complex case, a measurable f : X → C is integrable (or

absolutely integrable) if |f | is integrable. From the inequalities

max(|Ref |, |Imf |) ≤ |f | ≤ |Ref | + |Imf |

it is clear that f : X → C is (absolutely) integrable if and only if Ref and Imf are.

If f is complex-valued and absolutely integrable (that is, f is measurable and |f | is

86 MAA6616 COURSE NOTES FALL 2016

integrable), we define
Z Z Z
f= Ref + i Imf.

R
We also write kf k1 := X
|f | dµ in the complex case.

If f : X → R is absolutely integrable, then necessarily the set {x : |f (x)| = +∞}

has measure 0 by Theorem 10.2(d). We may therefore redefine f to be 0, say, on this

set, without affecting the integral of f (by Theorem 10.2(c)). Thus when working with

T
absolutely integrable functions, we often can (and often will) always assume that f is

finite-valued everywhere.
AF
11.1. Basic properties of the absolutely convergent integral. The next few propo-

sitions collect some basic properties of the absolutely convergent integral. Let L1 (µ) de-
DR

note the set of all absolutely integrable C-valued functions on X. If the measure space

is understood, as it is in this section, we just write L1 .)

Theorem 11.1 (Basic properties of L1 functions). Let f, g ∈ L1 and c ∈ C. Then:

(a) L1 is a vector space over C;

R
(b) the mapping Λ : L1 → C defined by Λ(f ) = f is linear;
R  R
(c)  f  ≤ |f |.

(d) kcf k1 = |c|kf k1 .

(e) kf + gk1 ≤ kf k1 + kgk1 .

 MAA6616 COURSE NOTES FALL 2016 87

Proof. To prove L1 is a vector space, suppose f, g ∈ L1 and c ∈ C. Since f, g are

measurable, so is f + g, thus |f + g| has an integral. Moreover, since |f + g| ≤ |f | + |g|,

monotonicity and additivity of the unsigned integral (Theorems 10.2 and 10.5) gives
Z Z Z
kf + gk1 = |f + g| ≤ |f | + |g| ≤ kf k1 + kgk1 ,

R R
proving item (e) and that L1 is closed under addition. Next, |cf | = |c| |f | = |c| kf k1

(using homogeneity of the unsigned integral in Theorem 10.2). Thus cf ∈ L1 and item

(d) holds. Further L1 is a vector space.

T
To prove that Λ is linear, first assume f and g are real-valued and c ∈ R. Checking
R R
AF
c f = cf is straightforward. For additivity, let h = f + g and observe

h+ − h− = f + + g + − f − − g − .

Therefore
DR

h+ + f − + g − = h− + f + + g + .

Thus,
Z Z
− −
+
h +f +g = h− + f + + g +

and rearranging, using additivity of the unsigned integral and finiteness of all the in-
R R R
tegrals involved, gives h = f+ g. The complex case now follows essentially by

definition. Hence Λ is linear proving item (b).

If f is real, then, using additivity of the unsigned integral,

Z   Z Z  Z Z Z Z
 f  =  f + − f −  ≤ f + + f − = (f + + f − ) = |f |.
   
   
88 MAA6616 COURSE NOTES FALL 2016
R
Hence (c) holds for real-valued functions. When f is complex, assume f 6= 0 and let
R R R
t = sgn f . Then |t| = 1 and | f | = t f . It follows that, using part (c) for the

real-valued function Retf ,

Z  Z Z Z Z Z Z Z
 
 f  = t f = tf = Re tf = Retf ≤ |Retf | ≤ |tf | = |f |.
 

R
Because of cancellation, it is clear that f = 0 does not imply f = 0 a.e. when f

T
is a signed or complex function. However the conclusion f = 0 a.e. can be recovered if
R
we assume the vanishing of all the integrals f , over all measurable sets E.
AF E

Proposition 11.2. Let f ∈ L1 . The following are equivalent:

(a) f = 0 almost everywhere,

DR

R
(b) |f | = 0,
R
(c) For every measurable set E, E
f = 0.

Proof. Since f = 0 a.e. if and only if |f | = 0 a.e., (a) and (b) are equivalent by

Theorem 10.2(e). Now assuming (b), if E is measurable then by monotonicity and

Theorem 11.1(b)
Z  Z Z
 
 f ≤ |f | ≤ |f | = 0,
 
E E

so item (c) holds.

 MAA6616 COURSE NOTES FALL 2016 89

Now suppose (c) holds and f is real-valued. Let E = {f > 0} and note f + = f 1E .
R R
Hence f + is unsigned and, by assumption f+ = E
f = 0. Thus, by Theorem 10.2,

f + = 0 a.e. Similarly f − = 0 a.e. and thus f is the difference of two functions that are

zero a.e. To complete the proof, write f in terms of its real and imaginary parts. 

R R
Corollary 11.3. If f, g ∈ L1 and f = g µ-a.e., then f= g. †

Proof. Apply Proposition 11.2 to f − g. 

T
A consequence of Corollary 11.3 is that we can introduces an equivalence relation

on L1 (µ) by declaring f ∼ g if and only if f = g a.e. If [f ] denotes the equivalence class

AF
of f under this relation, we may define the integral on equivalence classes by declaring
R R
[f ] := f . Corollary 11.3 shows that this is well-defined. It is straightforward to check

that [cf + g] = [cf ] + [g] for all f, g ∈ L1 and scalars c (so that L1 / ∼ is a vector space),
DR

and that the properties of the integral given in Theorem 11.1 all persist if we work
R
with equivalence classes. The advantage is that now [|f |] = 0 if and only if [|f |] = 0.

This means that the quantity k[f ]k1 is a norm on L1 / ∼. Henceforth will we agree to

impose this relation whenever we talk about L1 , but for simplicity we will drop the [·]

notation, and also write just L1 for L1 / ∼. So, when we refer to an L1 function, it is

now understood that we refer to the equivalence class of functions equal to f a.e., but

in practice this abuse of terminology should cause no confusion.

Just as the Monotone Convergence Theorem is associated to the unsigned integral,

there is a convergence theorem for the absolutely convergent integral.

90 MAA6616 COURSE NOTES FALL 2016

Theorem 11.4 (Dominated Convergence Theorem). Suppose (fn )∞

n=1 is a sequence from

L1 that converges pointwise a.e. to a measurable function f . If there exists a function

g ∈ L1 such that for every n, we have |fn | ≤ g a.e., then f ∈ L1 , and

Z Z
lim fn = f.
n→∞

Proof. First observe that |f | ≤ g and hence f ∈ L1 . By considering the real and imagi-

nary parts separately, we may assume f and all the fn are real valued. By hypothesis,

g ± fn ≥ 0 a.e. Applying Fatou’s theorem and linearity of the integral to these sequences

T
gives
Z Z Z Z Z Z
g+
AF f= (g + f ) ≤ lim inf (g + fn ) = g + lim inf fn

and
Z Z Z Z Z Z
g− f= (g − f ) ≤ lim inf (g − fn ) = g − lim sup fn .
DR

R R R
It follows that lim inf f≥ f ≥ lim sup f . 

R R R R 
The conclusion fn → f (equivalently,  fn − f  → 0) can be strengthened

somewhat:

Corollary 11.5. If fn , f, g satisfy the hypotheses of the Dominated Convergence theo-

R
rem, then limn→∞ kfn − f k1 = 0 (that is, lim |fn − f | = 0). †

Proof. Problem 13.20. 

Theorem 11.6 (Density of simple functions in L1 ). If f ∈ L1 , then there is a sequence

(fn ) of simple functions from L1 such that,

 MAA6616 COURSE NOTES FALL 2016 91

(a) |fn | ≤ |f | for all n,

(b) fn → f pointwise, and

(c) limn→∞ kfn − f k1 = 0.

Item (c) says (fn ) converges to f in L1 .

Proof. Write f = u + iv with u, v real, and u = u+ − u− , v = v + − v − . Each of the

four functions u± , v ± is unsigned and measurable and each is in L1 since f ∈ L1 . By

T
the ziggurat approximation we can choose four sequences of unsigned measurable simple

functions u± ±
AF ± ±
n , vn increasing pointwise to u , v respectively. Now put un = u+ −
n − un ,

vn = vn+ − vn− , and fn = un + ivn . By construction, each fn is simple (and measurable).

Moreover
DR

− −
|un | = u+ +
n + un ≤ u + u = |u|,

and similarly |vn | ≤ |v|, so |fn | ≤ |u| + |v| ≤ 2|f |. Since f ∈ L1 each fn is in L1 , and

fn → f pointwise by construction. Thus the sequence (fn ) satisfies the hypothesis of

the dominated convergence theorem (with g = 2|f |) and hence item (c) follows from

Corollary 11.5. 

12. Modes of convergence

In this section we consider five different ways in which a sequence of functions

on a measure space (X, M , µ) can be said to converge. There is no simple or easily

92 MAA6616 COURSE NOTES FALL 2016

summarized description of the relation between the five modes. At the end of the

section the reader is encouraged to draw a diagram showing the implications.

12.1. The five modes of convergence.

Definition 12.1. Let (X, M , µ) be a measure space and (fn )∞

n=1 , f be measurable

functions from X to C.

(a) The sequence (fn ) converges to f pointwise almost everywhere if µ({lim fn 6= f }) = 0.

T
(b) The sequence (fn ) converges to f essentially uniformly or in the L∞ norm if for

every  > 0, there exists N ∈ N such that µ({|fn − f | ≥ }) = 0 for all n ≥ N.
AF
(c) The sequence (fn ) converges to f almost uniformly if for every  > 0, there exists

an (exceptional set) E ∈ M such that µ(E) <  and (fn ) converges to f uniformly

on the complement of E.
DR

R
(d) The sequence (fn ) converges to f in the L1 norm if (kfn − f k1 := X
|fn − f | dµ)

converges to 0.

(e) The sequence (fn ) converges to f in measure if for every  > 0, the sequence µ({x :


|fn − f | > }) converges to 0.

The first thing to notice is that each of these modes of convergence is unaffected

if f or the fn are modified on sets of measure 0 (this is not true of ordinary pointwise

or uniform convergence). Thus these are modes of convergence appropriate to measure

 MAA6616 COURSE NOTES FALL 2016 93

theory. The L1 and L∞ modes are special cases of Lp convergence, which will be studied

later in the course.

We first treat a few basic properties common to all five modes of convergence.

Proposition 12.2 (Linearity of convergence). Let (fn ), (gn ), f, g be measurable functions

and c a complex number.

(a) For each of the five modes, (fn ) converges to f in the given mode if and only if

(|fn − f |) converges to 0 in the given mode.

T
(b) If (fn ) converges to f and (gn ) converges to g, then (cfn + gn ) converges to cf + g
AF
in the given mode.

Proof. The proof is left as an exercise. (Problem 13.24) 

DR

Proposition 12.3. Let (fn ) be a sequence of measurable functions and suppose f is

measurable.

(a) If (fn ) converges to f essentially uniformly, then (fn ) converges to f almost uni-

formly.

(b) If (fn ) converges to f almost uniformly, then (fn ) converges to f pointwise a.e. and

in measure.

(c) If (fn ) converges to f in the L1 norm, then (fn ) converges to f in measure.

†
94 MAA6616 COURSE NOTES FALL 2016

Proof. (a) is immediate from definitions. For (b), given k ≥ 1 we can find a measurable

1
set Ek with µ(Ek ) < k
such that fn → f uniformly (hence pointwise) on Ekc . It follows
S∞ T∞
that fn → f pointwise on the set k=1 Ekc . The complement of this set, k=1 Ek has

measure zero, since µ( ∞ 1

for all m ∈ N+ . The second part of (b) is
T
k=1 Ek ) ≤ µ(Em ) ≤ m

left as an exercise.

Finally, (c) follows from Markov’s inequality (Proposition 10.9). For  > 0 fixed,
Z
1 1
µ({x : |fn (x) − f (x)| > }) ≤ |fn − f | dµ = kfn − f k1
 

T
X

and the sequence (kfn − f k1 ) converges to 0 by hypothesis. 

AF
Of the twenty possible implications that can hold between the five modes of con-

vergence, only the four implications ((b) is really two implications) given in the last

proposition (and the ones that follow by combining these) are true without further hy-
DR

potheses.

To understand the differing modes of convergence, and the failure of the remaining

possible implications in Proposition 12.3, it is helpful to work out what they say in the

simplest possible case, namely that of step functions. A step function is a function of the

form c1E for a positive constant c and measurable set E. Convergence of a sequence of

step functions to 0 in each of the five modes, turns out to be largely determined by three

objects associated to the sequence (cn 1En )∞

n=1 : the heights cn , the widths µ(En ), and the

S
tail supports Tn := j≥n Ej . The proof of the following theorem involves nothing more

than the definitions, but is an instructive exercise.

 MAA6616 COURSE NOTES FALL 2016 95

Theorem 12.4. Let fn = cn 1En be a sequence of step functions.

(a) Assuming µ(En ) > 0 for each n, the sequence (fn ) converges to 0 in L∞ if and only

if (cn ) converges to 0.

(b) The sequence (fn ) converges to 0 almost uniformly if either (cn ) or (µ(Tn )) converges

to 0.

(c) If (|cn |) is (eventually) bounded away from 0 and (fn ) converges almost uniformly to

0, then (µ(Tn )) converges to 0.

T
(d) The sequence (fn ) converges to 0 in measure if and only if the sequence (min{cn , µ(En )})

converges to 0.
AF
Proof. To prove item (c), suppose, without loss of generality, that there is a C > 0 such

that |cn | ≥ C for all n and (fn ) converges almost uniformly to 0. Given  > 0 there is
DR

a set F such that µ(F c ) <  and (fn ) converges uniformly on F . In particular, for each

k ∈ N+ there is an Nk such that

1
F ⊂ ∩n≥Nk {|fn | < }.
k

Equivalently,
1
F c ⊃ ∪n≥Nk {|fn | ≥ }.
k
1
Choose k such that k
< C. Since |cn | ≥ C > k1 , we see {|fn | ≥ k1 } = En for each n.

Hence,

F c ⊃ TNk .

Thus,  > µ(F c ) ≥ µ(Tn ) for all n ≥ Nk . It follows that (µ(Tn )) converges to 0.
96 MAA6616 COURSE NOTES FALL 2016

The remaining parts of the Theorem are similar (and easier) and are left as Prob-

lem 13.29. 

The moving bump examples

(a) (Escape to height infinity) fn = n1(0, 1 )

n

1
(b) (Escape to width infinity) fn = 1
2n (−n,n)

(c) (Escape to horizontal infinity) fn = 1(n,n+1)

T
(d) (Escape to horizontal infinity alternate) fn = 1(n,n+ 1 ) ,
n

together with the typewriter example below suffice to produce counterexamples to all of
AF
the implications not covered in Proposition 12.3.

Example 12.5. [The Typewriter Sequence] Consider Lebesgue measure on (0, 1]. Let
DR

Ink ⊂ (0, 1] denote the dyadic interval ( 2kn , k+1

2n
] for n ≥ 1, 0 ≤ k < 2n . List these intervals

in dictionary order (first by increasing n, then by increasing k). So the first few intervals

are I10 = (0, 12 ], I11 = ( 21 , 1], I20 = (0, 14 ], I21 = ( 14 , 12 ], etc. (Draw a picture to see what is

going on.) The sequence of indicator functions of these intervals (in this order) converges

in measure to 0, since for any 0 <  < 1 we have m({x : 1Ink > } = m(Ink ) = 2−n .

However since each point in (0, 1] lies in infinitely many Ink and also lies outside infinitely

many Ink , the sequence 1Ink (x) does not converge at any point of (0, 1]. 4

To go further we begin with a closer investigation of convergence in measure.

 MAA6616 COURSE NOTES FALL 2016 97

Definition 12.6. A sequence (fn )∞

n=1 of C-valued measurable functions is Cauchy in

measure if for every , η > 0, there is an N such that for n, m ≥ N ,

µ({x : |fn (x) − fm (x)| > }) < η.

It is readily seen that if (fn ) converges to f in measure, then the sequence (fn ) is

Cauchy in measure. Indeed, by the triangle inequality,

T
{x : |fn (x)−fm (x)| > } ⊂ {x : |fn (x)−f (x)| > /2}∪{x : |fm (x)−f (x)| > /2}, (30)
AF
and thus the result follows by subadditivity of µ.

Proposition 12.7. A sequence (fn )∞

n=1 of measurable functions fn : X → C is Cauchy
DR

in measure if and only if for every  > 0 there exists an integer N ≥ 1 such that

µ({x : |fn (x) − fm (x)| > }) < 

for all n, m ≥ N . Similarly (fn ) converges to f in measure if and only if for every  > 0

there exists N such that

µ({x : |fn (x) − f (x)| > }) < 

for all n ≥ N . †

Proof. Problem 13.27. 

98 MAA6616 COURSE NOTES FALL 2016

We have already seen that convergence in measure does not imply pointwise a.e.

convergence (the typewriter sequence). Note, however, that in that example there is at

least a subsequence converging pointwise a.e. to 0 (give an example).

Proposition 12.8. If (fn )∞

n=1 is a sequence of measurable functions that is Cauchy in

measure, then

a) there is a measurable function f and a subsequence (fnk )∞

k=1 such that (fnk )k

T
converges to f almost uniformly; and

b) with f as in part (a), (fn ) converges to f in measure, and if also (fn ) converges
AF
to g in measure, then f = g a.e.

†
DR

In other words, if the sequence (fn ) is Cauchy in measure, then it converges in

measure to a unique (a.e.) f , and a subsequence of (fn ) converges to f a.e.

Proof. With  = 2−1 , there is an n1 such that if m ≥ n1 , then µ{|fm − fn1 | > 2−1 } <

2−1 . Now with  = 2−2 , there is an n2 such that n2 > n1 and if m ≥ m2 , then

µ{|fm −fn2 | > 2−2 } < 2−2 . In particular, µ{|fn2 −fn1 | > 2−1 } < 2−1 . Taking  = 2−k = η

in the definition of convergence in measure,choose inductively a sequence of integers

n1 < n2 < . . . nk < . . . such that

µ({x : |fnk (x) − fnk+1 (x)| > 2−k }) < 2−k . (31)
 MAA6616 COURSE NOTES FALL 2016 99

Put gk = fnk . Let

Ek = {x : |gk (x) − gk+1 (x)| > 2−k }.

S∞ T∞
By (31) µ(Ek ) < 2−k . Let FN = k=N Ek and F = N =1 FN = lim sup Ek and observe
P∞
µ(FN ) ≤ k=N 2−k = 2−N −1 . For x 6∈ FN and m ≥ n ≥ N , the estimate
m−1
X m−1
X
|gn (x) − gm (x)| ≤ |gk+1 (x) − gk (x)| ≤ 2−k ≤ 2−n−1 ≤ 2N −1 (32)
k=n k=n

shows that (gn ) is uniformly Cauchy on FNc . Hence (gn ) is pointwise Cauchy on F c and

thus converges pointwise a.e. to a measurable function f by Proposition 8.17. Finally

T
(gn ) converges almost uniformly to f .
AF
Part (b) is a version of the fact that if a Cauchy sequence has a convergent subse-

quence, then the sequence converges; and, if a sequence has a limit, then the limit is

unique. Thus, to prove part (b), let (gk ) and f be as in the proof of part (a). Since
DR

(gk ) converges to f almost uniformly, it converges to f in measure by Proposition 12.3.

Using the triangle inequality,

{x : |fn (x) − f (x)| > } ⊂ {x : |fn (x) − gk (x)| > /2} ∪ {x : |gk (x) − f (x)| > /2},

and, using the Cauchy in measure assumption on (fn ) and that (gk ) converges to f in

measure, the measures of the sets on the right can be made less than  by choosing n, k

sufficiently large. The details are left to the gentle reader.

Finally, suppose also (fn ) converges to g in measure. By one more application of

the triangle inequality trick, for a fixed  > 0,

{x : |f (x) − g(x)| > } ⊂ {x : |f (x) − fn (x)| > /2} ∪ {x : |fn (x) − g(x)| > /2},
100 MAA6616 COURSE NOTES FALL 2016

and the measures of the sets on the right-hand side go to 0 by hypothesis. So µ({x :

|f (x) − g(x)| > n1 }) = 0 for all n ∈ N+ . By the pigeon hole principle, Proposition 2.6,

µ({|f − g| =
6 0}) = 0. 

Corollary 12.9. If (fn ) converges to f in L1 , then (fn ) has a subsequence converging

to f a.e. †

Proof. Combine Proposition 12.3(c) and Proposition 12.8. 

T
AF
Proposition 12.10. Let (X, M , µ) be a measure space. The normed vector space L1 =

L1 (µ) is complete. In particular, if (fn ) is an L1 Cauchy sequence, then there is an

f ∈ L1 and a subsequence (gk ) of (fn ) such that (gk ) converges pointwise a.e. to f and
DR

(fn ) converges in L1 to f . †

Proof. Suppose (fn ) is L1 -Cauchy. In this case (fn ) is Cauchy in measure and hence

has a subsequence (hm ) that converges pointwise a.e. to some measurable function f by

Corollary 12.9. Choose a subsequence (gk = hmk ) such that

1
kgk+1 − gk k < .
2k

(Such a subsequence is super Cauchy.) Let

m
X
Gm = |gk+1 − gk |.
k=1
 MAA6616 COURSE NOTES FALL 2016 101

The sequence Gm is an increasing sequence of non-negative functions and hence converges

to some G. By Tonelli (Corollary 10.6),

∞ Z
X 1
1= k
≥ G dµ.
k=1
2 X

In particular G is in L1 . Further,

m
X
gm+1 = [gk+1 − gk ] + g1
k=1

T
is dominated by |g1 | + G and converges pointwise a.e. to f . Hence by Corollary 11.5,

f ∈ L1 and (gm ) converges to f in L1 . Finally, since (fn ) is L1 Cauchy and a subsequence

AF
converges (in L1 ) to f , the full sequence converges in L1 to f . 

12.2. Finite measure spaces. Observe that two of the “moving bump” examples (es-
DR

cape to width infinity and escape to horizontal infinity) exploit the fact that Lebesgue

measure on R is infinite. Moreover, in some cases these are the only counterexamples

(of the four) to particular implications–for example, escape to width infinity is the only

example of convergence in L∞ but not convergence in L1 , and escape to horizontal in-

finity is the only one of pointwise a.e. convergence but not convergence in measure. It

is then plausible that if we work on a finite measure space (µ(X) < ∞), where these

examples are “turned off,” we might recover further convergence results. This is indeed

the case.
102 MAA6616 COURSE NOTES FALL 2016

Proposition 12.11. Suppose (X, M , µ) is a finite measure space, (fn ) is an L1 sequence

and f : X → C is measurable. If (fn ) converges to f essentially uniformly, then f ∈ L1

and (fn ) converges to f in L1 . †

Proof. Problem 13.28. 

Theorem 12.12 (Egorov’s Theorem). Let (X, M , µ) be a measure space with µ(X) <

∞. If fn : X → C is a sequence of measurable functions, f : X → C is measurable and

T
(fn ) converges to f a.e., then (fn ) converges to f almost uniformly.
AF
Proof. There is no loss of generality in assuming (fn ) converges to f everywhere. For

N, k ≥ 1, let
∞
DR

[ 1
EN,k = {x : |fn (x) − f (x)| ≥ }.
n=N
k

1
Fix k. For each x, there is an N such that |fn (x) − f (x)| < k
for all n ≥ N . Hence
T
N ≥1 EN,k = ∅. Since the EN,k are decreasing with N and µ(X) < ∞, by dominated

convergence for sets for each k the sequence (µ(EN,k ))N converges to 0.

Now let  > 0 be given. Choose, for each k, an Nk such that µ(ENk ,k ) < 2−k . Let
S∞
E= k=1 ENk ,k and observe µ(E) < . To prove (fn ) converges to f uniformly on E c ,

1
given η > 0 choose k such that k
< η. Suppose now that x ∈ E c and n ≥ Nk . Since

1
E c ⊂ EN
c
k ,k
, the inequality |fn (x) − f (x)| < k
< η holds and we conclude that (fn )

converges uniformly to f on E c . 
 MAA6616 COURSE NOTES FALL 2016 103

Remark 12.13. Note that in Egorov’s theorem, almost uniform convergence cannot be

improved to essential uniform convergence, as the moving bump example 1[0, 1 ] shows. 
n

12.3. Uniform integrability. In the last section we saw that some convergence im-

plications could be recovered by making an assumption (µ(X) < ∞) that “turns off”

some of the failure modes. In this section we do something similar. In particular note

that the moving bump examples show that of the five modes, L1 convergence is the

T
R R
only one that implies fn → f (assuming the fn and f are integrable). The main

result of this section, a version of the Vitali convergence theorem, says that if we make
AF
certain assumptions on fn (namely “uniform integrability”) that turn off the moving

bump examples, then we can conclude that L1 convergence is equivalent to convergence

in measure. This result is similar in spirit to the classical Vitali Convergence Theorem
DR

(which we will cover later in the context of Lp spaces), though the approach used here

(borrowed from T. Tao, An Introduction to Measure Theory, Section 1.5) is slightly

different.

Definition 12.14. [Uniform integrability] A subset F of L1 is uniformly integrable

provided

(i) [Uniform bound on L1 norm] The set {kf k1 : f ∈ F } is bounded;

|f | dµ : f ∈ F }) → 0 as M → +∞
R
(ii) [No escape to vertical infinity] sup({ {|f |≥M }

|f | dµ : f ∈ F }) → 0 as δ → 0.
R
(iii) [No escape to width infinity] sup({ {|f |≤δ}
104 MAA6616 COURSE NOTES FALL 2016

A sequence fn : X → C of L1 functions is uniformly integrable if the set {fn : n} is

uniformly integrable. /

(To visualize the conditions in items (ii) and (iii), work out what they say for

sequences of step functions.) We warm up by observing that, for a single L1 function f ,

by the Dominated Convergence theorem,

Z
lim |f | dµ = 0
M →+∞ |f |>M

T
and
Z
lim |f | dµ = 0.
AF δ→0 |f |≤δ

Thus F = {f } is uniformly integrable.

R
Uniform integrability for a sequence (fn ) says the quantities |fn |>M
|fn | dµ and
DR

R
|fn |≤δ
|fn | dµ can be made arbitrarily small by choice of large M and small δ, indepen-

dently of n. Proposition 12.17 below says if (fn ) is an L1 Cauchy sequence, then (fn ) is

uniformly integrable. Note too a finite union of uniformly integrable sets is uniformly

integrable. In particular, if (fn ) converges to f ∈ L1 , then F = {fn : n} ∪ {f } is

uniformly integrable.

Before going further we give an equivalent formulation of item (ii) (assuming item

(i)):

Lemma 12.15. If (fn ) is a sequence of L1 functions such that supn kfn k1 < ∞, then

the condition of item (ii) in Definition 12.14 is equivalent to,

 MAA6616 COURSE NOTES FALL 2016 105
R
(iib) for every  > 0, there exists a δ > 0 such that if µ(E) < δ, then E
|fn | <  for

all n.

R
Proof. Suppose (ii) holds and let  > 0 be give. Choose M such that supn |fn |≥M
|fn | <

 
2
, and let δ < 2M
. If E is a measurable set with µ(E) < δ, then for all n,
Z Z Z
|fn | dµ = |fn | dµ + |fn | dµ
E E∩{|fn |≥M } E∩{|fn |<M }

T
Z Z
≤ |fn | dµ + M dµ
{|fn |≥M } E

< + M µ(E)
AF 2

< .

Conversely, suppose item (iib) holds. Fix  > 0, and for δ > 0 satisfying (iib) choose
DR

1
M large enough that M
supn kfn k1 < δ. Then by Markov’s inequality (Proposition 10.9),

kfn k1
for all n we have µ({|fn | ≥ M } ≤ M
< δ. Thus by (iib)
Z
|fn | < 
|fn |≥M

for all n. Hence (i) and (ii) implies (iib).

Remark 12.16. On a finite measure space, escape to width infinity is impossible, and

a sequence is uniformly integrable if and only if supn kf k1 < ∞ and (ii) (equivalently,

(iib)) is satisfied. 


106 MAA6616 COURSE NOTES FALL 2016

Proposition 12.17. If (fn ) is a sequence of L1 functions and (fn ) converges to f in

L1 , then (fn ) is uniformly integrable. †

Proof. Problem 13.35. Since convergent sequence are bounded, there K = sup{kfn k : n}

is finite.

We next turn to verifying item (iib). Accordingly, let  > 0 be given. By Markov’s

K
inequality, µ({|fn | > M }) ≤ M
for M > 0.

Since (fn ) converges to f , there is an N such that kf − fn k <  for all n ≥ N. Since

T
{f } is absolutely integral, by item (iib), there is a η > 0 such that if µ(E) < η, then
R
AF K
E
|f | < . Choose M > η
. If n ≥ N, then
Z Z Z
|fn | ≤ |fn − f | + |f | < 2.
|fn |>M |fn |>M |fn |>M

To prove item (iii), let  > 0 be given. Using the fact that {f } is uniformly integrable,
DR

R
choose δ > 0 such that |f |<2δ
|f | < . There is an N such that if n ≥ N , then

kf − fn k < min{δη, }. It follows from Markov’s inequality that µ({|f − fn | > δ}) < η.

Finally, observe {|fn | < δ} ⊂ {|f | < 2δ} ∪ {|f − fn | > δ} for all n ≥ N. Hence
Z Z Z Z
|fn | ≤ |fn − f | + |f | + |f | < 3.
|fn |<δ |f |<2δ |f −fn |>δ

Theorem 12.18. Suppose fn : X → C is a sequence of L1 functions and f : X → C

is measurable. The sequence (fn ) converges to f in L1 if and only if (fn ) is uniformly

integrable and converges to f in measure.

 MAA6616 COURSE NOTES FALL 2016 107

Proof. The forward implication follows from Propositions 12.10, 12.3 and 12.17.

For the converse, suppose (fn ) is uniformly integrable and converges to f in measure.

To show that f ∈ L1 , first note, by uniform integrability, there is a constant C such

R
that X
|fn | ≤ C for all n, and by Proposition 12.8 there is a subsequence (fnk ) of (fn )

converging to f a.e. Applying Fatou’s theorem to this subsequence, we conclude

Z Z
|f | ≤ lim inf |fnk | ≤ C, (33)
X X

T
so f ∈ L1 .

Since f ∈ L1 and (fn ) is uniformly integrable, the set {fn : n}∪{f } is also uniformly
AF
integrable. Thus, by condition (iii) in the definition of uniformly integrable, given  > 0,

there is a δ > 0 such that for all n

Z
DR

|fn | dµ ≤  (34)
|fn |≤δ

and at the same time

Z
|f | dµ ≤ . (35)
|f |≤δ

From conditions (i) and (ii) and Lemma 12.15, there exists a γ > 0 such that µ(E) ≤ γ

implies
Z
|fn | dµ < 
E
Z
|f | dµ < 
E

δ
for all n. Now choose 0 < η = min{ 2δ , 2C , γ}.
108 MAA6616 COURSE NOTES FALL 2016

From (34) and (35)

Z
|fn | dµ ≤ 
|fn −f |<η, |f |≤δ/2

and
Z
|f | dµ ≤ .
|fn −f |<η, |f |≤δ/2

So by the triangle inequality

Z
|fn − f | dµ ≤ 2. (36)
|fn −f |<η, |f |≤δ/2

We now estimate the integral of |fn − f | over the region |fn − f | < η, |f | > δ/2 via

T
Markov’s inequality. Indeed,
AF µ({x : |f (x)| > δ/2}) ≤
C
.
δ/2

Thus
Z
C
|fn − f | dµ ≤ η ≤ . (37)
δ/2
DR

|fn −f |<η, |f |>δ/2

Combining Equations (37) and (36) gives

Z
|fn − f | dµ ≤ 3. (38)
|fn −f |<η

Finally the convergence in measure hypothesis is used to estimate the integral of

|fn − f | over the set |fn − f | ≥ η. With  = η in the definition of convergence in measure,

there is an N such that for all n ≥ N ,

µ({x : |fn (x) − f (x)| ≥ η} ≤ η.

Hence, by the choice of γ,

Z
|fn | dµ ≤ 
|fn −f |≥η
 MAA6616 COURSE NOTES FALL 2016 109

and
Z
|f | dµ ≤ .
|fn −f |≥η

Thus again by the triangle inequality

Z
|fn − f | dµ ≤ 2
|fn −f |≥η

for all n ≥ N. Putting this last inequality together with (38),

Z
|fn − f | dµ < 5
X

T
for all n ≥ N and the proof is complete.
AF 

Remark 12.19. Say that a sequence of measurable functions fn : X → C is dominated

if there is an L1 function g such that |fn | ≤ |g| for all n. It is not hard to show

(Problem 13.34) that if a sequence (fn ) is dominated, then it is uniformly integrable.

DR

On the other hand, a sequence (fn ) can converge in L1 yet not be dominated. The main

utility of Theorem 12.18 is a criteria for proving L1 convergence for sequences that are

not dominated. 

13. Problems

13.1. Measurable functions.

Problem 13.1. Suppose f : X → C is a measurable function. Prove that the functions

|f | and sgnf are measurable. (Recall that for a complex number z, sgn(z) = z/|z| if

z 6= 0, and sgn(0) = 0.)

110 MAA6616 COURSE NOTES FALL 2016

Problem 13.2. Let f : R → R be a function. For each of the following, prove or give

a counterexample.

a) Suppose f 2 is Lebesgue measurable. Does it follow that f is Lebesgue measur-

able?

b) Suppose f 3 is Lebesgue measurable. Does it follow that f is Lebesgue measur-

able?

Problem 13.3. Recall the definition of an atomic σ-algebra (Problem 7.3). Prove that if

T
(X, M ) is a measurable space and M is an atomic σ-algebra, then a function f : X → R
AF
is measurable if and only if it is constant on each atom An .

Problem 13.4. Prove, if f : R → R is monotone, then f is Borel measurable.

Problem 13.5. Let fn : X → R be a sequence of measurable functions. Prove that the

DR

set {x ∈ X : limn→∞ fn (x) exists} is measurable.

Problem 13.6. Give an example of an uncountable collection F of Lebesgue measurable

functions f : R → [0, +∞] such that the function f (x) = supf ∈F f (x) is not Lebesgue

measurable.

Problem 13.7. Let f : [0, 1] → [0, 1] denote the Cantor-Lebesgue function of Exam-

ple 6.2(c) and define g(x) = f (x) + x.

(i) Prove that g is a homeomorphism of [0, 1] onto [0, 2]. (Hint: it suffices to show

g is a continuous bijection.)
 MAA6616 COURSE NOTES FALL 2016 111

(ii) Let C ⊂ [0, 1] denote the Cantor set. Prove that m(g(C)) = 1. (Here m is

Lebesgue measure.)

(iii) By Problem 7.29, g(C) contains a nonmeasurable set S. Show that g −1 (S) is

Lebesgue measurable, but not Borel.

(iv) Prove that there exists functions F, G on R such that F is Lebesgue measurable,

G is continuous, but F ◦ G is not Lebesgue measurable.

Problem 13.8. Prove that if f, g : X → [0, ∞] are measurable functions, then f g is

T
measurable.
AF
Problem 13.9. Prove Propositions 8.16 and 8.17. For 8.17, you may wish to use the

observation that f : X → R is measurable if and only if {x : f (x) > q} is measurable

for every q ∈ Q. (The following example shows that µ can not be replaced by µ in the
DR

conclusion. Let X = R and M = {∅, X}. Let µ denote the zero measure on M . In this

case the completion of M is 2X and µ is the zero measure on 2X . Let f : X → X denote

the identity function. It is 2X measurable. A function g is M measurable if and only if

it is constant, say with value c. Hence {f 6= g} = X \ {c} and this set is not in M .)

13.2. The unsigned integral.

Problem 13.10. Prove the claim made immediately before Definition 9.1.

Problem 13.11. Complete the proof of Theorem 9.2.

112 MAA6616 COURSE NOTES FALL 2016
R
Problem 13.12. Prove that if f is an unsigned measurable function and f < +∞,

then the set E := {x : f (x) > 0} is σ-finite. (That is, E can be written as a disjoint
S∞
union of measurable sets E = n=1 En with each µ(En ) < +∞.)

R
Problem 13.13. Suppose that f is an unsigned measurable function and f < +∞.

a) Prove that for every  > 0 there is a measurable set E with µ(E) < +∞, such
R R
that f− E
f < .

T
R R
b) Prove that for every  > 0 there is a positive integer n such that f − min(f, n) <

.
AF
Problem 13.14. Prove Fatou’s Theorem (Theorem 10.7) without using the Monotone

Convergence Theorem. Then use Fatou’s theorem to prove the Monotone Convergence
DR

Theorem.

Problem 13.15. Let X be any set and let µ be counting measure on X. Prove that for
R P
every unsigned function f : X → [0, +∞], we have X
f dµ = x∈X f (x).

Problem 13.16. Prove Theorem 10.10. (Hint: to verify the integral formula, use the

Monotone Convergence Theorem.)

Problem 13.17. Prove Theorem 10.12. (Hint: show first that µ(E) := L(1E ) is a
R
measure, then that L(f ) = f dµ. Problem 7.8 may be helpful.)
 MAA6616 COURSE NOTES FALL 2016 113

Problem 13.18. Let f be an unsigned measurable function on the measure space

(X, M , µ). Prove that the function ν : M → [0, ∞] defined by ν(E) :=

R
E
f dµ is
R R
a measure and, if g is an unsigned measurable function on X, then g dν = gf dµ.

Problem 13.19. Prove (using monotone convergence and without using the dominated

convergence theorem) that if fn is a sequence of unsigned measurable functions that

R R R
decreases pointwise to f , and fN < ∞ for some N , then f = lim fn . Give an

example to show that the finiteness hypothesis is necessary.

T
13.3. The signed integral.
AF
Problem 13.20. Prove Corollary 11.5.

Problem 13.21. Prove the following generalization of the dominated convergence the-
DR

orem: suppose (fn ) converges to f a.e. If gn is a sequence of L1 functions converging

R R R R
a.e. to an L1 function g, if |fn | ≤ gn for all n, and gn → g, then fn → f.

R
Problem 13.22. Suppose fn , f ∈ L1 and (fn ) converges to f a.e. Prove that |fn −f | →
R R
0 if and only if |fn | → |f |. (Use the previous problem.)

Problem 13.23. Evaluate each of the following limits, and carefully justify your claims.
Z ∞
sin(x/n)
a) lim dx
n→∞
Z0 ∞ (1 + (x/n))n
1 + nx2
b) lim dx
n→∞
Z0 ∞ (1 + x2 )n
n sin(x/n)
c) lim dx
n→∞ 0 x(1 + x2 )
114 MAA6616 COURSE NOTES FALL 2016
Z ∞
n
d) lim dx
n→∞ 0 1 + n 2 x2

13.4. Modes of convergence.

Problem 13.24. Prove Proposition 12.2

Problem 13.25. Prove that if (fn ) converges to f almost uniformly, then (fn ) converges

to f in measure.

T
Problem 13.26. Show that the implications between modes of convergence not given

in Proposition 12.3 are false.

AF
Problem 13.27. Prove Proposition 12.7.

Problem 13.28. . Prove Proposition 12.11.

DR

Problem 13.29. Prove Theorem 12.4.

Problem 13.30. Let fn = 1(n,n+ 1 ) . Show that (fn ) converges pointwise and in measure,
n

but not almost uniformly, to 0.

1
Let f2n = 1(n,n+ 1
) and f2n+1 = 1
2n (−1,1)
. Show (fn ) converges almost uniformly to
n2

0, but, writing fn = cn 1En , neither (cn ) nor (µ(En )) converges to 0.

Problem 13.31. Let (X, M , µ) be a finite measure space. For measurable functions

f, g : X → C, define
|f − g|
Z
d(f, g) = dµ.
X 1 + |f − g|
 MAA6616 COURSE NOTES FALL 2016 115

Prove that d is a metric on the set of measurable functions (where we identify f and g

when f = g a.e.).

Problem 13.32. Let (X, M , µ) be a finite measure space. Prove that (fn ) converges

to f in measure if and only if d(fn , f ) → 0, where d is the metric in Problem 13.31.

Problem 13.33. [Fast L1 convergence] Suppose (fn ) converges to f in L1 in such a way

P∞
that n=1 kfn − f k1 < ∞. Prove that (fn ) converges to f almost uniformly. (Note that

T
this hypothesis “turns off” the typewriter sequence.) (Hint: first show that, given  > 0

and m ≥ 1, there exists an integer N such that

AF
∞
!
[
µ {x : |fk (x) − f (x)| ≥ 2−m } < .
k=N

To construct your exceptional set E, use the /2n trick.)

DR

Problem 13.34. Prove that if fn is a dominated sequence, then it is uniformly inte-

grable. Give an example of a sequence (fn ) that converges in L1 (and is thus uniformly

integrable), but is not dominated.

Problem 13.35. Prove that if (fn ) converges to f in L1 , then (fn ) is uniformly integrable

(Proposition 12.17).

Problem 13.36. Suppose (fn ) converges to f in measure and fn is dominated. Give a

direct proof that (fn ) converges to f in L1 (without using Theorem 12.18).

116 MAA6616 COURSE NOTES FALL 2016

Problem 13.37. Prove that if (fn ) is a dominated sequence, and (fn ) converges to f

a.e., then (fn ) converges to f almost uniformly. (Hint: imitate the proof of Egorov’s

theorem.) (Thus for dominated sequences, a.e. and a.u. convergence are equivalent.)

Problem 13.38. [Defect version of Fatou’s theorem] Let (fn ) be a sequence of unsigned

L1 functions with supn kfn k1 < ∞. Suppose (fn ) converges to f a.e. Prove that (fn )
R R
converges to f in L1 if and only if fn → f . [Suggestion: Consider the sequence

|f − fn | + (f − fn ).]

T
AF
DR
 MAA6616 COURSE NOTES FALL 2016 117

14. The Riesz-Markov Representation Theorem

Let X be a compact Hausdorff space. Recall the space C(X) is the vector space

{f : X → C : f is continuous} with the norm

kf k∞ = kf k = max{|f (x)| : x ∈ X} (39)

and that, as a metric space (the distance from f to g is kf − gk), C(X) is a complete.

Generally, a complete normed vector space is called a Banach space.

T
For a locally compact Hausdorff space X, a function f : X → C has compact

support if there exists a compact set K such that f (x) = 0 for x ∈

/ K; i.e., the closure of
AF
{x ∈ X : f (x) 6= 0} is compact. Assuming X is a locally compact Hausdorff space, let

Cc (X) denote those f ∈ C(X) with compact support.; The space Cc (X) is also given

the supremum norm as in Equation (39).

DR

Given a vector space V , a linear mapping λ : V → C is called a linear functional.

A linear functional λ : Cc (X) → C is positive, if λ(f ) ≥ 0 whenever f ≥ 0 (meaning

f (x) ≥ 0 is pointwise positive).

Example 14.1. Suppose µ is a regular Borel measure on the locally compact set X and

µ(K) < ∞ for compact subsets of X. (This last condition is automatic if X is compact

and µ(X) < ∞). Thus, by Theorem 10.2(f), µ determines a positive linear functional,

λ, on Cc (X) by
Z
λ(f ) = f dµ.
X
118 MAA6616 COURSE NOTES FALL 2016

As a second example, let X = [0, 1] and note that the mapping I : C([0, 1]) → C

defined by
Z 1
I(f ) = f dx,
0

where the integral is in the Riemann sense, is a positive linear functional on C([0, 1]). 4

Theorem 14.2 (Riesz-Markov Representation Theorem). Let X = (X, τ ) be a locally

compact Hausdorff space. If λ : Cc (X) → C is a positive linear functional, then there

T
exists a unique Borel measure µ on the Borel σ-algebra BX , such that
Z
λ(f ) = f dµ
AF
for f ∈ Cc (X) and such that µ is regular in the sense that

(i) if K is compact, then µ(K) < ∞;

DR

(ii) if E ∈ BX , then µ(E) = inf{µ(U ) : E ⊂ U, U open}; and

(iii) if E ∈ BX and µ(E) < ∞, then µ(E) = sup{µ(K) : K ⊂ E, K compact}.

We will prove the result for the case X is compact. In this case, item (i) implies µ

is a a finite measure, and hence item (iii) applies to all E ∈ BX . In the next subsection,

topological preliminaries are gathered. The proof itself is in Subsection 14.2

14.1. Urysohn’s Lemma and partitions of unity. A topological space X is normal

if for each pair C1 , C2 of disjoint closed subsets of X, there exists disjoint open sets

U1 , U2 such that Cj ⊂ Uj .
 MAA6616 COURSE NOTES FALL 2016 119

Theorem 14.3. A compact Hausdorff space X is normal.

Theorem 14.4 (Urysohn’s lemma). If X is a compact Hausdorff space and A, B are

disjoint closed subsets of X, then there exists a function f : X → [0, 1] such that f is

zero on A and f is 1 on B. In particular, if K is compact, V is open and K ⊂ V , then

there is a continuous f : X → R such that 1K ≤ f ≤ 1V and supp(f ) ⊂ V .

Remark 14.5. Urysohn’s Lemma implies that X is normal by choosing U = f −1 ((−1, 12 ))

T
and V = f −1 (( 21 , 2)).
AF
Note that the lemma does not say A = f −1 ({0}) or B = f −1 ({1}), though this can

be arranged in the case that X is a metric space. 

Theorem 14.6 (Partition of Unity). Suppose V1 , . . . , Vn are open subsets of a compact

DR

Hausdorff space X. If C is closed and C ⊂ ∪Vj , then there exists continuous functions

hj : X → [0, 1] such that

(i) hj ≤ 1Vj ;

(ii) supp(hj ) ⊂ Vj ; and

(iii) for x ∈ C,
n
X
hj (x) = 1.
j=1

14.2. Proof of Theorem 14.2. Suppose X is a compact metric space and λ : C(X) →

C is a positive linear functional. To get an idea how to define µ(V ) for an open set
120 MAA6616 COURSE NOTES FALL 2016

V ∈ τ , note that K = X \ V is compact and the function g : X → R

g(x) = d(x, K) = min{d(x, k) : k ∈ K}

is continuous. The sequence

g 1
fn = ( )n
1+g

is pointwise increasing to the characteristic function (or indicator function) of V , denoted

1V . Thus 1V : X → R is defined by 1V (x) = 0 for x ∈

/ V and 1V (x) = 1 for x ∈ V . If µ

T
exists, then, by the MCT,
Z Z
AF µ(V ) = 1V dµ = lim
n→∞
fn dµ.
X

We are led to make the following definitions. For V open, define

µ0 (V ) = sup{λ(f ) : f ∈ C(X), 0 ≤ f ≤ 1V , supp(f ) ⊂ V }.

DR

Thus, letting τ denote the topology on X, µ0 : τ → [0, ∞).

Define µ∗ : 2X → R by

µ∗ (E) = inf{µ0 (V ) : V is open and E ⊂ V }.

(Note that this definition is forced upon us to achieve outer regularity.)

The proof is now broken down into a series of Lemmas. The functions, unless

otherwise noted, are continuous.

Lemma 14.7. The mapping µ0 is monotone and countably subadditive (on τ ). †

 MAA6616 COURSE NOTES FALL 2016 121

Proof. That µ0 is monotone is evident. To prove that it is countably subadditive, suppose

(Vj ) is a sequence of open sets and let V = ∪Vj . Suppose f is continuous, nonnegative

valued, f ≤ 1V and K = supp(f ) ⊂ V . Since K is compact and K ⊂ ∪Vj , there exists

an N such that K ⊂ ∪N
j=1 Vj . By Theorem 14.6, there exists functions hj ∈ C(X) such

PN
that 0 ≤ hj ≤ 1Vj the support of hj lies in Vj and j=1 hj = 1 on K. It follows that
P
f= f hj and f hj ≤ 1Vj as well as supp(f hj ) ⊂ Vj . Hence,

X N
X ∞
X
λ(f ) = λ(f hj ) ≤ µ0 (Vj ) ≤ µ0 (Vj )

T
j=1 j=1

and inequality that completes the proof. 

AF
Lemma 14.8. If V1 and V2 are disjoint open sets, then µ0 (V1 ∪V2 ) = µ0 (V1 )+µ0 (V2 ). †
DR

Proof. Let W = V1 ∪ V2 . By Lemma 14.7, it suffices to show that µ0 (W ) ≥ µ0 (V1 ) +

µ0 (V2 ). To this end, let  > 0 be given and suppose fj ≤ 1Vj are such that supp(fj ) ⊂ Vj

and µ0 (Vj ) ≤ λ(fj ) + . By disjointness, f1 + f2 ≤ 1W and supp(f1 + f2 ) ⊂ W too.

Hence,

X X
2 + µ0 (W ) ≥ 2 + λ(f ) = 2 + λ(fj ) ≥ µ0 (Vj ).

Since  > 0 is arbitrary, the conclusion follows. 

Lemma 14.9. The mapping µ∗ is an outer measure. Further, if W is open, then

µ∗ (W ) = µ0 (W ). †
122 MAA6616 COURSE NOTES FALL 2016

Proof. Since µ0 (∅) = 0, to prove µ∗ is an outer measure, it suffices to prove that, for

E ⊂ X,
∞
X
∗
µ (E) = inf{ µ0 (Vj ) : (Vj ) is a sequence of open sets and E ⊂ ∪Vj }.
j=1

Hence, it is enough to show that if (Vj ) is a sequence of open sets such that E ⊂ ∪Vj ,
P∞
then µ0 (V ) ≤ j=1 µ0 (Vj ) for some open set E ⊂ V . Choose V = ∪Vj and apply

Lemma 14.7. 

T
Lemma 14.10. If K is compact and 1K ≤ f , then µ∗ (K) ≤ λ(f ).
AF †

Proof. Given 0 < δ < 1, let Vδ = {f > δ}. Note that Vδ contains K and is open.

Moreover, if g ≤ 1Vδ and supp(g) ⊂ Vδ , then δg ≤ f and hence λ(g) ≤ 1δ λ(f ). It follows

that,
1
DR

µ0 (Vδ ) ≤ λ(f ).
δ

Thus, by monotonicity of outer measure, µ∗ (K) ≤ µ0 (Vδ ) ≤ 1δ λ(f ). Letting δ < 1 tend

to 1 gives the result. 

Lemma 14.11. If W is open, K is compact and K ⊂ W , then

µ0 (W ) = µ0 (W \ K) + µ∗ (K).

Proof. One inequality follows from subadditivity of outer measure. To prove the other

inequality note that W \ K is open and let  > 0 be given. Choose 0 ≤ g ≤ 1W \K with
 MAA6616 COURSE NOTES FALL 2016 123

supp(g) ⊂ W \ K and λ(g) +  ≥ µ0 (W \ K). Let C = supp(g). Choose, by Theorem

14.4, 1K ≤ f ≤ 1C c ∩W such that supp(f ) ⊂ C c ∩ W . In particular, 0 ≤ f + g ≤ 1W and

the support of f + g lies in W and, by Lemma 14.10, λ(f ) ≥ µ∗ (K). Thus,

 + µ0 (W ) ≥  + λ(f + g) = λ(f ) + ( + λ(g)) ≥ µ∗ (K) + µ0 (W \ K).

Lemma 14.12. If W is open, K is compact, K ⊂ W and  > 0, then there exists

1K ≤ f ≤ 1W such that supp(f ) ⊂ W and λ(f ) ≤ µ∗ (K) + . †

T
Proof. Choose V an open set such that K ⊂ V and µ0 (V ) ≤ µ∗ (K) + . Replacing W
AF
by V ∩ W if needed, it may be assumed that V ⊂ W . By Theorem 14.4, there exists

1K ≤ f ≤ 1V and supp(f ) ⊂ V . It follows that

λ(f ) ≤ µ0 (V ) ≤ µ∗ (K) + .
DR

Lemma 14.13. If W is open and  > 0, then there is a compact set K such that K ⊂ W

and µ0 (W ) ≤ µ∗ (K) + . †

Proof. There is a 0 ≤ g ≤ 1W such that supp(g) ⊂ W and λ(g) +  > µ0 (W ). Let K

denote the support of g. Hence, K ⊂ W and K is compact. By Lemma 14.12, there

exists an f such that 1K ≤ f ≤ 1W , the support of f lies in W and λ(f ) ≤ µ∗ (K) + .

In particular g ≤ f and hence λ(g) ≤ λ(f ). It follows that

µ0 (W ) ≤ λ(g) +  ≤ λ(f ) +  ≤ µ∗ (K) + 2.

124 MAA6616 COURSE NOTES FALL 2016

Lemma 14.14. If W is open, then W is outer measurable. †

Proof. Let A ⊂ X be given. Given  > 0, choose an open set A ⊂ V such that µ0 (V ) ≤

µ∗ (A)+. Choose, by Lemma 14.13, a compact set K ⊂ W such that µ0 (W ) ≤ µ∗ (K)+.

Now, by monotonicity and Lemma 14.11,

µ∗ (A ∩ W ) ≤ µ0 (V ∩ W )

T
≤ µ∗ (V ∩ K) + µ0 (V ∩ (W \ K))
AF ≤ µ∗ (V ∩ K) + µ0 (W \ K)

≤ µ∗ (V ∩ K) + .

Further, by monotonicity,
DR

µ∗ (A ∩ W c ) ≤ µ∗ (V ∩ W c ).

Now K and W c are disjoint compact sets. Hence, by Theorem 14.3, there exist disjoint

open sets S, T such that K ⊂ S and W c ⊂ T . Consequently, using Lemma 14.8 and

monotonicity,
µ∗ (A ∩ W ) + µ∗ (A ∩ W c ) ≤ µ∗ (V ∩ K) +  + µ∗ (V ∩ W c )

≤ µ0 (V ∩ S) + µ0 (V ∩ T ) + 

= µ0 (V ∩ (S ∪ T )) + 

≤ µ0 (V ) ≤ µ∗ (A) + 2.

It follows that µ∗ (A) ≥ µ∗ (A ∩ W ) + µ∗ (A ∩ W c ) and thus W is outer measurable. 

 MAA6616 COURSE NOTES FALL 2016 125

Let M denote the collection of outer measurable sets. Thus, µ, the restriction of µ∗

to M is a complete measure. Further, M contains all open sets by Lemma 14.14 and

by Lemma 14.9 if W is open then µ(W ) = µ0 (W ). In particular, BX ⊂ M .

Lemma 14.15. The measure µ satisfies the regularity conditions of the theorem. †

Proof. Outer regularity follows immediately from the definition of µ∗ . As for inner

T
regularity, suppose E is measurable. Thus E c is measurable. By outer regularity, there

is an open set V such that E c ⊂ V and µ(V \ E c ) < . Thus K = V c ⊂ E is compact

AF
and µ(E \ K) < . 
DR

Lemma 14.16. If f ∈ C(X), then

Z
λ(f ) = f dµ.
X

Proof. Suppose f ∈ C(X) is real-valued and that [a, b] contains the range of f . Given

 > 0, choose t0 < a < t1 < . . . tn = b such that tj − tj−1 < . Let Ej = f −1 ((tj−1 , tj ])

for j = 1, n. The Ej are Borel sets, hence, by outer regularity, there exists open sets

Vj ⊃ Ej such that µ(Vj ) ≤ µ(Ej ) + n . By Theorem 14.6, there exists hj ∈ C(X) such
126 MAA6616 COURSE NOTES FALL 2016
P
that 0 ≤ hj ≤ 1Vj , the support of hj lies in Vj and hj = 1. Now,
X
λ(f ) = λ(f hj )
X
≤ λ(tj hj )
X
≤ tj µ(Vj )
X 
≤( tj−1 + )(µ(Ej ) + )
n
Z
≤ f dµ + (µ(X) + ).
X

Consequently,

T
Z
λ(f ) ≤ f dµ. (40)
X
AF
The reverse inequality follows by replacing f by −f in Equation (40).

Finally, the case of general continuous f : X → C is reduced to the case f is real

by considering the real and imaginary parts of f separately. 

DR

Lemma 14.17. If µ1 , µ2 are regular Borel measures such that

Z
λ(f ) = f dµj
X

for j = 1, 2 and f ∈ C(X), then µ1 = µ2 . †

Proof. Let K be a given compact set. By outer regularity, given  > 0 there exists an

open set V such that K ⊂ V and µj (V ) ≤ µj (K) + . By Theorem 14.4, there is an

f ∈ C(X) such that 1K ≤ f ≤ 1V . Hence,

Z Z Z
µj (V ) −  ≤ µj (K) = 1K dµj ≤ f dµj = λ(f ) ≤ 1V dµj = µj (V ).
X X X
 MAA6616 COURSE NOTES FALL 2016 127

Hence |µ1 (K) − µ2 (K)| ≤ 2 and therefore µ1 (K) = µ2 (K). By inner regularity, it now

follows that µ1 = µ2 . 

15. Product measures

We now revisit measures and σ-algebras. Recall, given measure spaces (X, M ) and

(Y, N ), the product σ-algebra M ⊗ N ⊂ 2X×Y is the σ-algebra generated by the

measurable rectangles {E × F : E ∈ M , F ∈ N }.

T
Example 15.1. (a) If X, Y are finite sets and X, Y are given the discrete σ-algebras
AF
2X , 2Y , then 2X ⊗ 2Y = 2X×Y .

(b) If we take two copies of R with the Borel σ-algebra BR , then BR ⊗ BR = BR2 . (See

Proposition 1.17.)
DR

Given a pair of measure spaces (X, M , µ), (Y, N , ν), we would like to construct a

“product” measure µ × ν on the Cartesian product measurable space (X × Y, M × N ).

It is natural to insist, if E ∈ M and F ∈ N have finite measure, then µ × ν(E × F ) =

µ(E) ν(F ); i.e., the measure of a measurable rectangle is the product of the measures.

We would also like conditions guaranteeing uniqueness. We will state and prove theorems

for only for two factors, but there is no difficulty in extending to finitely many factors

(Xj , Mj , µj ), j = 1, . . . n. It turns out that the product is associative too. There is also

a construction valid for infinitely many factors when each factor is a probability space
128 MAA6616 COURSE NOTES FALL 2016

(that is, µ(X) = 1) but this requires more care. In these notes we consider only the

finite case.

To express, in what follows, integrals of unsigned functions f : X × Y → [0, +∞]

against product measures as iterated integrals, we introduce the slice functions fx : Y →

[0, +∞] and f y : X → [0, +∞], defined for each x ∈ X (respectively, each y ∈ Y ) by

fx (y) := f (x, y), f y (x) = f (x, y).

In other words, starting with f (x, y) we get functions defined on Y by holding x fixed,

T
and functions defined on X by holding y fixed.
AF
In addition to these, given a set E ⊂ X × Y , we can define for all x ∈ X, y ∈ Y the

slice sets Ex ⊂ Y , E y ⊂ X by

Ex := {y ∈ Y : (x, y) ∈ E}, E y := {x ∈ X : (x, y) ∈ E}

DR

We first show that these constructions preserve measurability.

Lemma 15.2. Let (X, M ), (Y, N ) be measurable spaces.

(i) If E belongs to the product σ-algebra M ⊗ N , then for all x ∈ X and y ∈ Y the

slice sets Ex and E y belong to N and M respectively.

(ii) If (Z, O) is another measurable space and f : X × Y → Z is a measurable function,

then for all x ∈ X and y ∈ Y , the functions fx and f y are measurable on Y and

X respectively.

†
 MAA6616 COURSE NOTES FALL 2016 129

Proof. Let S denote the set of all E ∈ 2X×Y with the property that E y ∈ M and

Ex ∈ N for all x ∈ X, y ∈ Y . It suffices to prove that S is a σ-algebra containing all

measurable rectangles. First observe that S contains all rectangles in M ⊗ N , since if

E = F ×G then Ex is equal to either G or ∅, if x ∈ F or x 6∈ F respectively. In either case

Ex ∈ N . The same proof works for E y . Next, suppose (En ) is a sequence of disjoint sets
S∞ S∞ S∞
in S and E = n=1 En . Then Ex = n=1 (En )x ∈ N ; similarly E y = n=1 Eny ∈ M .

Likewise (E ∩F )x = Ex ∩Fx . Thus if E, F ∈ S , then so is E ∩F . Finally, (E c )x = (Ex )c

T
for all x ∈ X; similarly for E y . Thus, by Proposition 1.7, S is a σ-algebra.

Item (ii) follows from item (i) by observing that for any O ⊂ O and x ∈ X,
AF
(fx )−1 (O) = (f −1 (O))x

and similarly for y. 

DR

Remark 15.3. Even if both (X, M , µ) and (Y, N , ν) are complete measure spaces and

if τ is a measure on M ⊗ N such that τ (E × F ) = µ(E) ν(F ) for all measurable

rectangles E × F with µ(E), ν(F ) < ∞, it need not be the case that the product

measure (X × Y, M ⊗ N , τ ) is complete. Indeed, if there is an set E ⊂ X such that

/ M that is contained in a set G of finite measure and a nonempty F ∈ N of

E ∈

measure zero, then E × F ⊂ G × F and τ (G × F ) = 0, but, for any p ∈ F , the slice set

(E × F )p = {x ∈ X : (x, p) ∈ F } = E is not in M , and hence E × F is not in M ⊗ N

by Lemma 15.2. 
130 MAA6616 COURSE NOTES FALL 2016

The following Lemma is of independent interest.

Definition 15.4. Let X be a set. A monotone class is a collection C ⊂ 2X of subsets

of X such that

S∞
(i) if E1 ⊂ E2 ⊂ · · · belong to C , so does n=1 En ; and
T∞
(ii) if E1 ⊃ E2 ⊃ · · · belong to C , so does n=1 En .

T
Remark 15.5. It is immediate that intersections of monotone classes are monotone
AF
classes. Hence, given a collection A ⊂ 2X , there is a smallest monotone class containing

A. If C is a monotone class, then so is C 0 = {E c : E ∈ C }. Trivially, every σ-algebra is

a monotone class. The next lemma is a partial converse to this statement. 

DR

Lemma 15.6 (Monotone class lemma). If A ⊂ 2X is a Boolean algebra, then the

smallest monotone class containing A is equal to the σ-algebra generated by A . †

Proof. Let M denote the σ-algebra generated by A and C the smallest monotone class

containing A . Since M is a monotone class containing A , we have C ⊂ M and hence

it suffices to prove that M ⊂ C .

Now C 0 is a monotone class and, since E ∈ A implies E c ∈ A , it contains A . Hence

C ⊂ C 0 . Thus, if E ∈ C , then there is an F ∈ C such that E = F c and E c = F ∈ C .

Thus C is closed under complements; that is C 0 = C .

 MAA6616 COURSE NOTES FALL 2016 131

Given E ⊂ X, let CE denote the set of all F ∈ C such that the sets

F \ E, E \ F, F ∩ E, X \ (E ∪ F )

belong to C . A quick check of the definitions shows that CE is a monotone class.

Moreover, if E ∈ A it is immediate that CE contains A and hence CE = E . Let

D = {E ∈ C : CE = C }.

In particular, A ⊂ D. Another definition check shows D is a monotone class. Thus,

C ⊂ D and hence C = D.

T
Now suppose that E, F ∈ C . Since E ∈ D and F ∈ C , it follows that E ∩ F ∈ C .
AF
Hence C is closed under finite intersections. Since it is also closed under complements,

it is closed under finite unions. Since ∅ 6= C is closed under finite unions, complements

and countable increasing unions, it is a σ-algebra. 

DR

The proof strategy employing the monotone class lemma should be clear. To prove

that a statement P holds for a σ-algebra M generated by a Boolean algebra A , it

suffices to prove 1) P is true for all E ∈ A , and 2) the collection of all E ∈ M for which

P is true is a monotone class.

We can now construct the product measure.

Theorem 15.7 (Existence and uniqueness of product measure). Let (X, M , µ), (Y, N , ν)

be σ-finite measure spaces. If P ∈ M ⊗ N , then

R
(i) f : X → [0, ∞] defined by f (x) = ν(Px ) = Y
1Px dν is measurable;
132 MAA6616 COURSE NOTES FALL 2016
R
(ii) g : Y → [0, ∞] defined by g(y) = µ(P y ) = X
1P y dµ is measurable; and

(iii)
Z Z
f dµ = g dν.
X Y

The function µ × ν : M ⊗ N → [0, ∞] defined

Z Z
µ × ν(P ) = f dµ = g dν
X Y

is a σ-finite measure on the product σ-algebra and is uniquely determined by µ × ν(E ×

F ) = µ(E) ν(F ) for E ∈ M and F ∈ N .

T
The measure µ × ν is the product measure.
AF
Proof. We will give the proof assuming at one point both measures are finite, and then

sketch out how this assumption can be relaxed to σ-finiteness. Given sets E ∈ M and

F ∈ N , the set E × F is (measurable) rectangle. The collection of finite disjoint unions

DR

of measurable rectangles, denoted E is a Boolean algebra. Let P denote the collection

of sets P ∈ M ⊗ N satisfying (i), (ii) and (iii). Define µ × ν : P → [0, ∞] as in the

statement of the theorem.

That each measurable rectangle belongs to P is evident. In fact, if P = ∪Ej × Fj

is a finite disjoint union of rectangles, then

X
ν(Px ) = 1Ej (x) ν(Fj ).

Hence, ν(Px ) is measurable and similarly for P y . Moreover,

Z X Z
ν(Px ) dν = µ(Ej )ν(Fj ) = µ(P y ) dµ.
Y X
 MAA6616 COURSE NOTES FALL 2016 133

Now suppose P1 ⊂ P2 ⊂ . . . is an increasing sequence from P and let P = ∪Pj .

Let

fj (x) = ν((Pj )x )

and define gj similarly. Since Px = (∪Pj )x = ∪(Pj )x , it follows from monotone conver-

gence for sets (Theorem 2.3 (iii)) that (fj ) monotone increases to

f (x) = ν(Px ).

Hence f and likewise g are measurable and moreover, by MCT twice,

T
Z Z Z Z
f dµ = lim fj dµ = lim gj dν = g dν.
X X Y Y
AF
Hence P ∈ P and µ × ν(P ) = lim µ × ν(Pj ).

At this point we add the assumption that µ and ν are both finite. Suppose P1 ⊃

P2 ⊃ . . . is a decreasing sequence from P and let P = ∪Pj . Proceeding as above, but

DR

using the DCT instead of the MCT by invoking the finiteness assumptions on µ and ν it

follow that P ∈ P and the proof that P = M ⊗ N is complete under the assumption

that the measures µ and ν are finite. For disjoint sets P1 , . . . , Pn ∈ M ⊗ N , the identity

∪nj=1 (Pj )x = (∪nj=1 Pj )x implies µ × ν is finitely additive. The argument above shows

µ × ν satisfies monotone convergence for sets and hence, by Problem 7.8, µ × ν is indeed

a measure on M ⊗ N under the added finiteness assumption.

In the case the measures are σ-finite, express X = ∪Xn and Y = ∪Yn as increasing

unions of measurable sets of finite measure. Let Zn = Xn × Yn ∈ E and note that each

Zn ∈ P, each µ × ν(Zn ) is finite and X × Y = ∪Zn . In particular, once it is shown that

134 MAA6616 COURSE NOTES FALL 2016

µ × ν is a measure on the product σ-algebra, the σ-finite conclusion will automatically

follow. For positive integers n, let Qn denote those sets P such that

Pn = P ∩ [Xn × Yn ] ∈ P.

In particular, Pn ⊂ Zn and the measures µn : M → [0, ∞] and ν : N → [0, ∞] defined

by µn (E) = µ(E ∩ Xn ) and νn (F ) = ν(F ∩ Yn ) respectively are finite and P ∈ Q if and

only if P ∈ M ⊗ N and
Z Z Z Z
y
νn (Px ) dµn = ν((Pn )x ) dµ = µ((Pn ) ) dν = µn (Px ) dνn .

T
X X Y Y

Hence, by what has already been proved, Qn = M ⊗ N for each n. Given P ∈ P

AF
let fn (x) = µ((Pn )x ) and f (x) = µ(Px ) and likewise for gn . The monotone convergence
R R
argument above shows f is measurable and ( fn dµ) converges to f dµ and likewise

gn dν since P ∈ Qn . Thus P = M ⊗ N . That

R R
for g. On the other hand fn dµ =
DR

µ × ν is measure on M ⊗ N is left as an exercise.

To prove uniqueness suppose ρ is any other measure on M ⊗N such that ρ(E×F ) =

µ(E) ν(F ) for measurable rectangles. Thus ρ agrees with µ × ν on the Boolean algebra

E and µ × ν is σ-finite on E . Hence ρ agrees with µ × ν on all of M ⊗ N by the Hahn

Uniqueness Theorem (Theorem 5.4). 

Corollary 15.8. Let (X, M , µ), (Y, N , ν) be σ-finite measure spaces. If E is a null set

for µ × ν, then ν(Ex ) = 0 for µ-a.e. x ∈ X, and µ(E y ) = 0 for ν-a.e. y ∈ Y . †

Proof. Problem 19.5. 

 MAA6616 COURSE NOTES FALL 2016 135

Example 15.9. a) If X, Y are at most countable, and µX , µY denote counting mea-

sure on X ×Y respectively, then 2X ⊗2Y = 2X×Y and µX ×µY is counting measure

on X × Y .

b) For two copies of R with the Borel σ-algebra and Lebesgue measure m (restricted

to BR ), the product measure is a σ-finite measure on BR2 that has the value

m(E)m(F ) on measurable rectangles. The completion of this measure is Lebesgue

measure on R2 . (By iterating this construction we of course obtain Lebesgue

T
measure on Rn .)

Let L denote the Lebesgue σ-algebra on R and LR2 denote Lebesgue measure
AF
on R2 . If E ∈ L , then E = B ∪ W , where B is Borel and W has Lebesgue

measure zero (and hence is a subset of a Borel set of measure zero) by the

regularity properties of Lebesgue measure. It follows that if E, F ∈ L , then

DR

E × F is the union of a set in BR ⊗ BR with a set contained within a set of

measure zero in BR ⊗ BR and hence E × F ∈ LR2 . Thus, L ⊗ L ⊂ LR2 . On

the other hand, equality does not hold by Remark 15.3.

Theorem 15.10 (Tonelli’s theorem, first version). Suppose (X, M , µ) and (Y, N , ν)

are σ-finite measure spaces. If f : X × Y → [0, +∞] is a M ⊗ N -measurable function,

then
136 MAA6616 COURSE NOTES FALL 2016
R R
(a) the slice integrals g(x) := Y
fx (y) dν(y) and h(y) := X
f y (x) dµ(x) are measurable

on X and Y respectively;
Z Z Z  Z Z 
y
(b) f d(µ × ν) = fx (y) dν(y) dµ(x) = f (x) dµ(x) dν(y); and
X×Y X Y Y X

(c) if f ∈ L1 (µ × ν), then fx and f y are in L1 (ν) and L1 (µ) for a.e. x and a.e y.

Proof. First suppose P ∈ M ⊗ N and let f = 1P . In this case, the result is the

conclusion of Theorem 15.7.

T
To move to general unsigned f , first note that by linearity we conclude immediately

that items (a) and (b) also hold for simple functions. For a general unsigned measurable
AF
f : X ×Y → [0, +∞], use the Ziggurat approximation to choose an increasing a sequence

(fn ) of measurable simple functions converging to f pointwise. Let

DR

Z Z
gn (x) := (fn )x (y) dν(y) and hn (y) := (fn )y (x) dµ(x).
Y X

The monotone convergence theorem implies that the sequences (gn ) and (hn ) increase

and converge pointwise to g and h respectively. Thus g and h are measurable. Two

more applications of monotone convergence gives

Z Z Z Z
g dµ = lim gn dµ = lim fn d(µ × ν) = f d(µ × ν)
X X X×Y X×Y

and similarly for h. Thus, finally, items (a) and (b) hold for all unsigned measurable

functions on X × Y .

Item (c) follows immediately from item (b) and Theorem 10.2 item (d). 
 MAA6616 COURSE NOTES FALL 2016 137

As noted above, the product of complete measures is almost never complete. Typ-

ically we pass to the completion µ × ν of a product measure. To prove the complete

version of Tonelli’s theorem recall a couple of facts about measurability on complete

measure spaces encountered earlier (see Propositions 8.16 and 8.17).

Proposition 15.11. Let (X, M , µ) be a measure space and (X, M , µ) its completion.

a) If f : X → C is M -measurable, then there exists a M -measurable function fe

such that f = fe µ-a.e.

T
b) If f : X → C is M -measurable and g : X → C is a function with g(x) = f (x)
AF
for µ-a.e. x, then g is M -measurable.

Proof. Problem 19.4. 

DR

Theorem 15.12 (Tonelli’s theorem, complete version). Let (X, M , µ), (Y, N , ν) be

complete σ-finite measure spaces. If f : X × Y → [0, +∞] is an M ⊗ N -measurable

function, then

(i) for µ-a.e. x and ν-a.e. y, the functions fx and f y are N and M measurable

respectively;

(ii) there exists M and N measurable functions g and h such that

Z Z
g(x) = fx (y) dν(y), h(y) = f y (x) dµ(x)
Y X

µ-a.e. and ν-a.e. respectively;

138 MAA6616 COURSE NOTES FALL 2016
Z Z Z  Z Z 
y
(iii) f (x, y) dµ × ν = fx (y) dν(y) dµ(x) = f (x) dµ(x) dν(y); and
X×Y X Y Y X

(iv) If f is L1 , then fx and f y are in L1 for µ-a.e. x and ν-a.e. y respectively.

In item (ii), the integrals are defined only almost everywhere since the integrands

are defined only almost everywhere.

Proof. From Proposition 15.11 that there exists an M ⊗ N -measurable function fˆ such

T
that fˆ(x, y) = f (x, y) for µ × ν-a.e. (x, y). Let E be the exceptional set on which

f 6= fˆ. Since µ × ν(E) = 0, there is an M ⊗ N -measurable set Ê containing E such

AF
that (µ × ν)(Ê) = 0 by Theorem 2.8. By Corollary 15.8, ν(Êx ) = 0 for µ-a.e. x, thus

since Ex ⊂ Êx (and since ν is complete!) Ex is in N and ν(Ex ) = 0 as well for almost

every x. Since Ex = {y : fˆx 6= fx }, it follows that fx = fˆx ν-a.e. y for µ-a.e. x. Thus, by
DR

Lemma 15.2 and completeness of ν, the function fx is N -measurable (Proposition 15.11

again) µ-a.e. x. Of course, the analogous proof holds for f y .

By Theorem 15.10 (Tonelli),

Z
ĝ(x) = fˆx dν
Y

is measurable. Hence, as fˆx = fx ν-a.e. y for µ-a.e. x,

Z Z
g(x) = fˆx dν = fx dν µ-a.e. x.
Y Y

Finally (iii) and (iv) follow from (i) and (ii) and Theorem 15.10 applied to fˆ. 
 MAA6616 COURSE NOTES FALL 2016 139

Theorem 15.13 (Fubini’s theorem). Let (X, M , µ), (Y, N , ν) be complete σ-finite mea-

sure spaces. If f : X × Y → C belongs to L1 (µ × ν), then

a) for µ-a.e. x and ν-a.e. y, the functions fx and f y belong to L1 (ν) and L1 (µ)

respectively, and the functions

Z Z
g(x) = fx (y) dν(y), h(y) = f y (x) dµ(x)
Y X

belong to L1 (µ) and L1 (ν) respectively; and

ZZ Z Z  Z Z 
y
b) f (x, y) dµ × ν = fx (y) dν(y) dµ(x) = f (x) dµ(x) dν(y)

T
X×Y X Y Y X

Proof. By taking real and imaginary parts, and then positive and negative parts, it
AF
suffices to consider the case that f is unsigned, but then the theorem follows from

Theorem 15.12. Indeed, when f is unsigned and belongs to L1 (µ × ν), by Tonelli

Z Z  ZZ
DR

fx (y) dν(y) dµ(x) = f (x, y) d(µ × ν) < ∞,

X Y X×Y

R
but then Y
fx (y) dν(y) < ∞ for µ-a.e. x; similarly for f y . 

Corollary 15.14 (Integral as the area under a graph). Let (X, M , µ) be a σ-finite

measure space, and give R the Borel σ-algebra BR and Lebesgue measure m (restricted

to BR ). An unsigned function f : X → [0, +∞) is measurable if and only if the set

Gf := {(x, t) ∈ X × R : 0 ≤ t ≤ f (x)}

is measurable. In this case,

Z
(µ × m)(Gf ) = f dµ.
X
140 MAA6616 COURSE NOTES FALL 2016

Corollary 15.15 (Distribution formula). Let (X, M , µ) be a σ-finite measure space. If

f : X → [0, +∞] an unsigned measurable function, then

Z Z
f (x) dµ(x) = µ({f ≥ t}) dt.
X [0,+∞]

Proof. Let Gf be the region under the graph of f as in Corollary 15.14. Then for fixed

T
t ≥ 0,
Z
1Gf (x, t) dµ(x) = µ({f ≥ t})
AF X

so by Tonelli’s theorem and Corollary 15.14,

Z Z Z  Z
f (x) dµ(x) = (µ × m)(Gf ) = 1Gf (x, t) dµ(x) dt = µ({f ≥ t}) dt.
X [0,+∞] X [0,+∞]
DR

Corollary 15.16 (Compatibility of the Riemann and Lebesgue integrals). If f : [a, b] →

R is continuous, then if we extend f to be 0 off [a, b], the extended f is Lebesgue integrable
R Rb
on R and R
f dm = a
f (x) dx. †

Proof (sketch). We assume f ≥ 0. For a partition P , a = x0 < x1 < · · · < xn = b of

[a, b], define Cj = sup{f (x) : xj ≤ x ≤ xj+1 } and cj = sup{f (x) : xj ≤ x ≤ xj+1 }, and

consider the sums

n
X n
X
U (P, f ) := Cj (xj − xj−1 ) and L(P, f ) := cj (xj − xj−1 ).
j=1 j=1
 MAA6616 COURSE NOTES FALL 2016 141

Let Gf denote the region enclosed by the graph of f, Gf = {(x, y) : 0 ≤ y ≤ f (x)}.

R
Thus Gf is a closed set in R2 (hence Borel measurable), and by 15.14 m2 (Gf ) = R
f dm.

Let R+ and R− denote the finite unions of closed rectangles corresponding to the over-

and under-estimates for the Riemann integral given by U (P, f ) and L(P, f ). Then

R− ⊂ Gf ⊂ R+ , and m2 (R+ ) = U (P, f ), m2 (R− ) = L(P, f ). It then follows that

supP L(P, f ) ≤ m2 (Gf ) ≤ inf P U (P, f ). But by the definition of the Riemann integral,
Rb
the inf and sup are equal to each other, and their common value is a
f (x) dx. 

T
AF
Remark 15.17. The above proof can be modified to drop the continuity hypothesis

(where was it used?), and conclude that every Riemann integrable function on [a, b] is
DR

Lebesgue integrable, and the values of the two integrals agree. With more work it can

be shown that a function f : [a, b] → R is Riemann integrable if and only if the set of

points where f is discontinuous has Lebesgue measure 0. We will not prove this fact in

these notes.

We also note that it is not difficult to extend these facts about the Riemann integral

to “improper” Riemann integrals, defined over [0, +∞) or R. In particular, note that

the distribution function µ({|f | ≥ t}) is a decreasing function of t on [0, +∞), hence

Riemann integrable. Thus Corollary 15.15 says that, in principle, the calculation of any

Lebesgue integral can be reduced to the computation of a Riemann integral. 

142 MAA6616 COURSE NOTES FALL 2016

16. Integration in Rn

In this section we briefly discuss Lebesgue measure and Lebesgue integration on Rn .

We begin with the observation that we can construct Lebesgue measure mn on Rn

in the same way as on R, namely by introducing boxes B = I1 × I2 × · · · × In , where

Qn
each Ij is an interval in R, and declaring |B| = j=1 |Ij |. One can then define Lebesgue

outer measure mn∗ by defining, for all E ⊂ Rn ,

∞ ∞

T
X [
mn∗ (E) = inf{ |Bj | : E ⊂ Bj };
j=1 j=1
AF
the infimum taken over all coverings of E by boxes. By imitating the constructions of

Section 4, we are led to a σ-finite Borel measure on Rn such that the measure of a box B

is its volume |B|. Since the construction proceeds through outer measure, the σ-algebra
DR

LRn of measurable sets is complete and is of course called the Lebesgue σ-algebra. In

particular, the following analog of Theorem 4.5 holds.

Theorem 16.1. Let E ⊂ Rn . The following are equivalent:

a) E is Lebesgue measurable.

b) For every  > 0, there is an open set U ⊃ E such that mn∗ (U \ E) < .

c) For every  > 0, there is a closed set F ⊂ E such that mn∗ (E \ F ) < .

d) There is a Gδ set G such that E ⊂ G and mn∗ (G \ E) = 0.

e) There is an Fσ set F such that E ⊃ F and mn∗ (E \ F ) = 0.

 MAA6616 COURSE NOTES FALL 2016 143

We drop the superscript and just write m for Lebesgue measure on Rn when the

dimension is understood. It follows from Theorem 16.1, if E ⊂ Rn is Lebesgue measur-

able and µ(E) = 0, then there is a Borel set G ⊃ E such that m(G) = 0. Thus m is the

completion of m restricted to BRn as described in Theorem 2.8. Now, let m0 denote the

n-fold product of Lebesgue measure restricted to BR defined on BR ⊗ · · · ⊗ BR = BRn .

The measures m and m0 agree on the Boolean algebra of disjoint union of boxes and

thus, by the Hahn Uniqueness theorem, agree on BRn . Finally, the completion of m0

T
agrees with m and the completion of BRn (with respect to m0 ) is LRn , the Lebesgue

σ-algebra.
AF
Definition 16.2. Lebesgue measure mn on Rn is the completion of the n-fold product

of (R, BR , m) and the completion of BRn is the Lebesgue σ-algebra denoted LR n . /

DR

Rn possesses a larger group of symmetries than R does. In particular we would

like to analyze the behavior of Lebesgue measure under invertible linear transformations

T : Rn → Rn . We have the following analog of Theorem 4.4:

Theorem 16.3. If T : Rn → Rn be an invertible linear transformation, then T is

LRn − LRn measurable; i.e., if E ⊂ Rn is a Lebesgue set, then T −1 (E) ⊂ Rn is a

Lebesgue set too. Moreover,

Z Z
1 n 1
(a) if f ∈ L (R ), (f ◦ T )(x) dx = f (x) dx; and
Rn | det T | Rn

(b) if E ∈ Rn is Lebesgue measurable, then m(T (E)) = | det T |m(E).

144 MAA6616 COURSE NOTES FALL 2016

Problem 13.7 gives an example of a Lebesgue measurable F and continuous G such

that F ◦ G is not measurable. The difficulty is that it is possible E = F −1 (B) is not

Borel for a Borel measurable B and in this case there is no guarantee the inverse image

of E under G will be a Lebesgue set. In the course of the proof it will be shown that

if E ⊂ Rn is a Lebesgue set and T : Rn → Rn is linear and invertible, then T −1 (E) is

Lebesgue.

T
Proof. Let F denote those f ∈ L1 (Rn ) such that the composition f ◦ S is measurable
AF
for all invertible linear transformations S : Rn → Rn .

Note that, if T1 and T2 are both invertible linear transformations and the result of

(a) holds for any f ∈ F and both T1 and T2 , then the result of (a) holds for all f ∈ F
DR

and T = T1 T2 (and T2 T1 ). From linear algebra, every invertible linear transformation of

Rn is a finite product of transformations of one of the following types (we write vectors

in Rn as x = (x1 , . . . xn ), in the standard basis).

i) (Scaling a row) T (x1 , . . . xj , . . . xn ) = (x1 , . . . cxj , . . . xn ), for some j = 1, . . . n

and some c ∈ R

ii) (adding a row to another) T (x1 , . . . xj , . . . xk , . . . xn ) = (x1 , . . . xj , . . . xj +xk , . . . xn ),

some j, k = 1, . . . n

iii) (interchanging rows) T (x1 , . . . xj , . . . xk , . . . xn ) = (x1 , . . . xk , . . . xj , . . . xn ), some

j, k = 1, . . . n.
 MAA6616 COURSE NOTES FALL 2016 145

In the first case, det T = c, in the second, det T = 1, and in the third, det T = −1. By

the multiplicativity of the determinant, it suffices to prove the theorem for T of each

of these types. We may also assume f ≥ 0 (why?) But (a) now follows easily from

Tonelli’s theorem and the invariance properties of one-dimensional Lebesgue measure.

For example, for T of type (i) we integrate with respect to xj first and use the one-

dimensional fact
Z Z
1
g(ct) dt = g(t) dt
R |c|

T
for all c 6= 0. In case (ii) we integrate with respect to xk first and use translation
AF
invariance of one-dimensional Lebesgue measure: for fixed xj ,
Z Z
g(xj + xk ) dxk = g(xk ) dxk ,
R R

while for case (iii) we simply interchange the order of integration with respect to xj and
DR

xk . Thus (a) holds in all three cases. By composition (a) holds for any invertible T and

f ∈ F.

If f ∈ L1 (Rn ) is Borel measurable, then f ∈ F and hence (a) holds. In particular,

if G is a Borel set, then (a) applied to 1T (G) (using T −1 is linear and continuous shows

T (G) is also a Borel set) shows (b) holds for G. In particular, if m(G) = 0, then

m(T (G)) = 0 too. Now suppose E is a Lebesgue set. In this case there exists a

Borel set G with m(G) = 0, a subset N ⊂ G and a Borel set F such that E =

F ∪ N . Hence, as T is one-one, T (E) = T (F ) ∪ T (N ) and T (N ) is a subset of the

Borel set T (G) of measure zero. It follows that T (E) is a Lebesgue measurable set and
146 MAA6616 COURSE NOTES FALL 2016

det(T ) m(T (E)) = det(T ) (m(T (F )) = m(F ) = m(E). We conclude, if T is an invertible

linear transformation, then T maps Lebesgue sets to Lebesgue sets (as does T −1 .

Finally, since T −1 maps Lebesgue sets to Lebesgue sets, if f is measurable, then so

is f ◦ T and hence F = L1 (Rn ) completing the proof. 

Corollary 16.4. Lebesgue measure on Rn is rotation invariant. †

Proof. A rotation of Rn is just a linear transformation satisfying T t = T −1 , which implies

T
that | det T | = 1, so the claim follows from Theorem 16.3. 

One result we will use frequently in the rest of the course is the following fundamental
AF
approximation theorem. We already know that absolutely integrable functions can be

approximated in L1 by simple functions, we now show that in Rn we can approximate

in L1 with continuous functions.

DR

Definition 16.5. We say a function f : X → C is supported on a set E ⊂ X if f = 0 on

the complement of E. When X is a topological space, the closed support of f is equal to

the smallest closed set E such that f is supported on E. Say f is compactly supported

if it is supported on a compact set E. /

Note that since every bounded set in Rn has compact closure, a function f : Rn → C

is compactly supported if and only if it is supported in a bounded set. Since bounded

sets have finite Lebesgue measure, it follows that if f : Rn → C is continuous and

compactly supported, then it belongs to L1 (Rn ).

 MAA6616 COURSE NOTES FALL 2016 147

Theorem 16.6. If f ∈ L1 (Rn ) then there is a sequence of (fn ) of continuous, compactly

supported functions such that (fn ) converges to f in L1 .

Proof. We work in R first, and reduce to the case where f is simple. Let  > 0; since

simple functions are dense in L1 , there is an L1 simple function ψ such kψ − f k1 < 2 .

Since ψ is simple and in L1 , it is supported on a set of finite measure. If we can find

a continuous g ∈ L1 such that kψ − gk1 < /2 we are done. For this, it suffices (by

linearity) to assume ψ = 1E for a set E with m(E) < ∞ and show, given δ > 0 there

T
is a continuous g of compact support such that k1E − gk < δ. By Littlewood’s first
AF
principle Theorem 4.6, we can find a set A, a finite union of disjoint open intervals
Sn
A= j=1 (aj , bj ), such that m(A∆E) < 2δ . It follows that k1A − 1E k1 = k1A∆E k1 < 2δ .

Let η = δ
2n
and let Jj = (aj − η2 , bj + η2 ) and choose a continuous function gj : R → [0, 1]
DR

such that gj = 1 on Ij and gj = 0 on Jjc . Thus, kgj − 1Ij k ≤ η. Thus, with g =

P
gj , it
Pn
follows that kg − 1A k ≤ j=1 kgj − 1Ij k < 2δ .

In higher dimensions, the same approximation scheme works; it suffices (using lin-

earity, Littlewood’s first principle, and the /2n trick as before) to approximate the

indicator function of a box B = I1 × · · · × In (where each Ij has finite measure) again a

piecewise linear function that is 1 on the box and 0 outside a suitably small neighborhood

of the box suffices. The details are left as an exercise. 

Remark 16.7. There is a more sophisticated way to do continuous approximation in

L1 , using convolutions. This will be covered in depth later in the course. Also note
148 MAA6616 COURSE NOTES FALL 2016

that since every L1 convergent sequence has a pointwise a.e. convergent subsequence,

every L1 function f can be approximated by a sequence of continuous functions fn that

converge to f both in the L1 norm and pointwise a.e. 

As an application of the above approximation theorem, we prove a very useful fact

about integration in Rn , namely that translation is continuous in L1 (Rn ). The proof

strategy is to first prove the result from scratch for continuous, compactly supported

f , then use the density of these functions in L1 to get the general result. This density

T
argument is frequently used; we will see it again in the next section when we prove the
AF
Lebesgue Differentiation Theorem.

Proposition 16.8. For h ∈ Rn , and f : Rn → C a function, define fh (x) := f (x − h)

(the translation of f by h). If f ∈ L1 (Rn ), then fh ∈ L1 and fh → f in the L1 norm as

DR

h → 0. †

Proof. First suppose f is continuous and compactly supported. In this case f is uni-

formly continuous, each fh is continuous, and fh → f uniformly on K as h → 0. It

follows from Proposition 12.11 that fh → f in L1 .

Now let f ∈ L1 (Rn ) and  > 0 be given. We can choose a continuous, compactly

supported g such that kg − f k1 < /3. Note, by the translation invariance of Lebesgue

measure, that kgh − fh k1 = kg − f k1 < /3 as well. (Here we have used the readily

verified fact that (|f − g|)h = |fh − gh |). Now, since the result holds for g, there is a
 MAA6616 COURSE NOTES FALL 2016 149

δ > 0 such that for all |h| < δ, kgh − gk1 < /3. Thus

kfh − f k1 ≤ kfh − gh k1 + kgh − gk1 + kg − f k1 < ,

proving the theorem. 

The section concludes with some remarks on integration in polar coordinates. Write

kxk = (x21 + · · · + x2n )1/2 for the Euclidean length of a vector x. Let S n−1 = {x ∈ Rn :

T
kxk = 1} be the unit sphere in Rn . Each nonzero vector x can be expressed uniquely in

x
the form x = kxk kxk (positive scalar times a unit vector), so we may identify Rn \ {0}
AF
x
with (0, +∞) × S n−1 . Precisely, the map Φ(x) = (kxk, kxk ) is a continuous bijection

of Rn \ {0} and (0, +∞) × S n−1 . Using the map Φ we can define the push-forward of

Lebesgue measure to (0, +∞) × S n−1 ; namely m∗ (E) = m(Φ−1 (E)). Let ρ = ρn denote
DR

R
the measure on (0, +∞) defined by ρ(E) = E
rn−1 dr.

Theorem 16.9 (Integration in polar coordinates). There is a unique finite Borel mea-

sure σ = σn−1 on S n−1 such that m∗ = ρ × σ. If f is an unsigned or L1 Borel measurable

function on Rn then

Z Z ∞ Z
f (x) dx = f (rξ)rn−1 dσ(ξ)dr.
Rn 0 S n−1

Proof. 
150 MAA6616 COURSE NOTES FALL 2016

17. Differentiation theorems

One version of the fundamental theorem of calculus says that if f is continuous on

a closed interval [a, b] ⊂ R, if we define the function

Z x
F (x) := f (t) dt
a

then F is differentiable on (a, b) and F 0 (x) = f (x) for all x ∈ (a, b). Using the definition

of derivative, this can be reformulated as

Z x+h
1

T
lim f (t) dt = f (x)
h→0 h x

for all x ∈ (a, b). If we let I(x, h) denote the open interval (x, x + h), then, re-expressing
AF
in terms of the Lebesgue integral, we have
Z
1
lim f dm = f (x).
h→0 m(Ih ) Ih
DR

It is not hard to show that we can replace Ih with the interval B(x, h) centered on x

with radius h; in this case m(B(x, h)) = 1/2h and we still have
Z
1
lim f dm = f (x).
h→0 m(B(x, h)) B(x,h)

This can be interpreted to say that the average values of f over small intervals centered

on x converge to f (x), as one might expect from continuity. Perhaps surprisingly, the

result remains true, at least for (Lebesgue) almost every x, when we drop the continuity

hypothesis.

The goal of this section is to prove the Lebesgue differentiation theorem. To state it

we introduce the notation B(x, r) for the open ball of radius r > 0 centered at a point
 MAA6616 COURSE NOTES FALL 2016 151
R
x ∈ Rn . We also write f (y) dy for integrals against Lebesgue measure. For the rest of

this section L1 refers to Lebesgue measure on Rn unless stated otherwise.

Definition 17.1. [Locally integrable functions] A Lebesgue measurable function f :

R
Rn → C is called locally integrable if K
|f (y)| dy < ∞ for every compact set K ⊂ Rn .

The collection of all locally integrable functions on Rn is denoted L1loc (Rn ). /

Since every compact set in Rn is contained in a closed ball, it suffices in the above

T
R
definition to require only B
|f (y)| dy < ∞ for every ball B.

Theorem 17.2 (Lebesgue Differentiation Theorem). If f ∈ L1loc (Rn ), then, for almost
AF
every x ∈ Rn ,
Z
1
a) lim |f (y) − f (x)| dy = 0
r→0 m(B(x, r)) B(x,r)

and
DR

Z
1
b) lim f (y) dy = f (x).
r→0 m(B(x, r)) B(x,r)

Notice that the second statement follows from the first. One can interpret the

theorem as follows. Given f ∈ L1 , define for each r > 0 the function

Z
1
Af,r (x) := f (y) dy,
m(B(x, r)) B(x,r)

the average value of f over the ball of radius r centered at x. The second statement

says that the functions Ar,f converge to f almost everywhere as r → 0. (It is not hard

to show, using density of continuous functions of compact support in L1 that if f ∈ L1 ,

then Ar,f → f in the L1 norm as r → 0. See Problem 19.15.)

152 MAA6616 COURSE NOTES FALL 2016

To begin with, it is easy to prove Theorem 17.2 in the continuous case:

Lemma 17.3. If f : Rn → C is continuous and compactly supported, then for all

x ∈ Rn ,
Z
1
lim |f (y) − f (x)| dy = 0.
r→0 m(B(x, r)) B(x,r)

Proof of Lemma 17.3. Since f is continuous and compactly supported, it is absolutely

integrable. Fix x ∈ Rn and let  > 0 be given. By uniform continuity there is a δ > 0

T
such that |f (x) − f (y)| <  for all |y − x| < δ. For 0 < r < δ,
AF
1
Z
1
Z
|f (y) − f (x)| dy <  dy
m(B(x, r)) B(x,r) m(B(x, r)) B(x,r)

= .


DR

To move from continuous, compactly supported f to absolutely integrable f we need

the following estimate, which is quite important in its own right. It is an estimate on

the Hardy-Littlewood maximal function, which is defined for f ∈ L1 (Rn ) by

Z
1
Mf (x) = sup |f (y)| dy (41)
r>0 m(B(x, r)) B(x,r)

Theorem 17.4 (Hardy Littlewood Maximal Theorem). If f : Rn → C is in L1 and

t > 0, then
kf k1
m({x ∈ Rn : Mf (x) > t}) ≤ C .
t

for some absolute constant C > 0 that depends only on the dimension n.
 MAA6616 COURSE NOTES FALL 2016 153

Remark 17.5. It turns out that C can be chosen as 3n . If it were the case that

Mf were in L1 (Rn ), then a similar estimate would be an immediate consequence of

Markov’s inequality. However Mf is essentially never in L1 , even in the simplest case of

the indicator function of an interval. 

Before proving Theorem 17.4, we will see how it is used to prove the Lebesgue

Differentiation Theorem.

T
Proof of Therem 17.2. First note that we may assume f ∈ L1 (Rn ) (not just L1loc ); to

see this just replace f by 1B(0,N ) f for N ∈ N. So, let f ∈ L1 (Rn ) and fix , t > 0. We
AF
first prove (b) and then use this to deduce (a). First, by Theorem 16.6 there exists a

continuous, compactly supported g such that

Z
|f (x) − g(x)| dx < .
DR

Rn

Applying the Hardy-Littlewood maximal inequality to |f − g|, we have

Z
n 1
C
m {x ∈ R : sup |f (x) − g(x)| dx > t} ≤ .
r>0 m(B(x, r)) Rn t

In addition, by Markov’s inequality applied to |f − g| we have


m({x ∈ Rn : |f (x) − g(x)| > t}) ≤ .
t
(C+1)
Thus there is a set E ⊂ Rn of measure less than t
such that, outside of E both
Z
1
sup |f (y) − g(y)| dy ≤ t (42)
r>0 m(B(x, r)) B(x,r)

and

|f (x) − g(x)| ≤ t. (43)

154 MAA6616 COURSE NOTES FALL 2016

Now consider x ∈ E c . By the result for continuous, compactly supported functions

(Lemma 17.3), we have for all sufficiently small r > 0

 Z 
 1 

 m(B(x, r)) g(y) dy − g(x) ≤ t.
B(x,r)

In the left-hand side of this inequality, we add and subtract f (x) and the average value

of f over B(x, r). Then by (42), (43), and the triangle inequality, we have
 Z 
 1 

 m(B(x, r)) f (y) dy − f (x) ≤ 3t
B(x,r)

T
for all sufficiently small r > 0. Keeping t fixed, for each n there is a set En with

1
m(En ) < n
such that for each x ∈ Enc there exists an η > 0 such that for 0 < r < η,
AF
|Ar,f (x) − f (x)| ≤ 3t.

Let E = ∩En . Thus m(E) = 0 and for each x ∈ E c there exists an η > 0 such that the

inequality above holds for 0 < r < η. For each m ∈ N+ there exists a set Fm of measure
DR

zero such that for each x ∈ F c there is an η > 0 such that for 0 < r < η

1
|Ar,f (x) − f (x)| ≤ .
m

Finally, let F = ∪Fn and note that m(F ) = 0 and if x ∈ F c then, for every m ∈ N+

there exists an η > 0 such that for all 0 < r < η the inequality above holds completing

the second part of the Lebesgue Differentiation Theorem.

For part (a), note that if f is locally integrable and c ∈ C, then |f (x) − c| is locally

integrable. Thus for each c ∈ C we can apply part (b) to conclude that
Z
1
lim |f (y) − c| dy = |f (x) − c|
r→0 m(B(x, r)) B(x,r)
 MAA6616 COURSE NOTES FALL 2016 155

for all x outside an exceptional set Ec with m(Ec ) = 0. Fix a countable dense subset
S
Q ⊂ C and let E = c∈Q Ec ; then m(E) = 0 and for fixed x ∈
/ E there exists c ∈ Q

with |f (x) − c| < , so |f (y) − f (x)| < |f (y) − c| + , and

Z
1
lim sup |f (y) − f (x)| dy ≤ |f (x) − c| +  < 2.
r→0 m(B(x, r)) B(x,r)

Since  > 0 was arbitrary, this proves (a). 

It remains to prove the Hardy-Littlewood maximal inequality, Theorem 17.4. The

T
proof we give is based on the following lemma, known as the Wiener covering lemma.

Let B denote an open ball in Rn and for a > 0 let aB denote the open ball with the
AF
same center as B, whose radius is a times the radius of B.

Lemma 17.6 (Wiener’s covering lemma). Let B be a collection of open balls in Rn , and
DR

S
let U = B∈B B. If c < m(U ), then there exists finitely many disjoint balls B1 , . . . Bk ∈ B
Sk
such that m( j=1 Bj ) > 3−n c. †

Proof. There exists a compact set K ⊂ U such that m(K) > c. The collection of open

balls B covers K, so there are finitely many balls A1 , . . . Am whose union covers K. From

these we select a disjoint subcollection by a greedy algorithm: from A1 , . . . Am choose a

ball with maximal radius. Call this B1 . Now discard all the balls that intersect B1 . From

the balls that remain, choose one of maximal radius, necessarily disjoint from B1 , call

this B2 . Continue inductively, at each stage choosing a ball of maximal radius disjoint

from the balls that have already been picked. The process halts after a finite number of
156 MAA6616 COURSE NOTES FALL 2016

steps. We claim that the balls B1 , . . . Bk have the desired property. By construction the

Bj are pairwise disjoint. The claimed lower bound on the measure of the union follows

from a geometric observation. If A, A0 are open balls with radii r ≥ r0 respectively and

if A ∩ A0 6= ∅, then A0 ⊂ 3 · A (draw a picture and note the diameter of A0 is at most

twice the radius of A). From this observation, it follows that each ball Aj that was not

picked during the construction is contained in 3 · Bi for some i. In particular, the balls

3·B1 , . . . 3·Bk cover K. From the scaling property of Lebesgue measure (Theorem 16.3),

T
m(3 · B) = 3n m(B). Thus

X X k
[
c < m(K) ≤ m(3 · Bj ) = 3n m(Bj ) = 3n m( Bj ).
AF
j=1

Proof of Theorem 17.4. Let f ∈ L1 (Rn ) and fix λ > 0. Let Eλ = {x ∈ Rn : Mf (x) > λ}.
DR

If x ∈ Eλ , then by definition of Mf there is an rx > 0 such that Arx ,|f | (x) > λ. The open
S
balls B(x, rx ) then cover Eλ . Fix c with m(Eλ ) > c. Then m( x∈Eλ B(x, rx )) > c, so

by the Wiener covering lemma there are finitely many x1 , . . . xk ∈ Eλ so that the balls

Bk := B(xk , rxk ) are disjoint and m( kj=1 Bj ) > 3−n c. From the way the radii rx were
S

chosen, for each 1 ≤ j ≤ k,

Z
1
λ < Arxj ,|f | (xj ) = |f (y)| dy
m(B(xj , rxj )) B(xj ,rxj )

so
Z
1
m(Bj ) < |f (y)| dy.
λ Bj
 MAA6616 COURSE NOTES FALL 2016 157

It follows that

k
3n X 3n 3n
[ X Z Z
n n
c < 3 m( Bj ) = 3 m(Bj ) < |f (y)| dy ≤ |f (y)| dy = kf k1 .
j=1 j
λ j Bj λ Rn λ

This holds for all c < m(Eλ ), so taking the supremum over such c we get finally

kf k1
m(Eλ ) ≤ 3n .
λ

T
18. Signed measures and the Lebesgue-Radon-Nikodym Theorem
AF
A second form of the fundamental theorem of calculus says that if f : [a, b] → R is

differentiable at each point in [a, b] and if f is in L1 ([a, b]), then

Z x
f (x) − f (a) = f 0 (t) dt,
DR

(44)
a

for all a ≤ x ≤ b. Suppose f is increasing on [a, b]. From our construction of Lebesgue-

Stieltjes measures, the formula µ([c, d]) := f (d)−f (c), defined for all subintervals [c, d] ⊂

[a, b], determines a unique Borel measure on [a, b]. On the other hand, from (44), this

measure can equivalently be defined by the formula

Z
µ(E) = f 0 (x) dx. (45)
E

From Problem 13.18, for an unsigned measurable g : [a, b] → R,

Z b Z b
g dµ = gf 0 dm,
a a
158 MAA6616 COURSE NOTES FALL 2016

where m is Lebesgue measure on [a, b]. It is tempting to write dµ = f 0 dt or even more

dµ
suggestively, dm
= f 0 . As will be seen in this section, f 0 is the Radon-Nikodym derivative

of µ with respect to m.

18.1. Signed measures; the Hahn and Jordan decomposition theorems. If µ, ν

are measures on a common measurable space (X, M ), then we have already seen that

we can form new measures cµ (for c ≥ 0) and µ + ν. We would like to extend these

operations to allow negative constants and subtraction. The obvious thing to do is to

T
define the difference of two measures to be
AF
(µ − ν)(E) = µ(E) − ν(E). (46)

The only difficulty is that the right-hand side may take the form ∞ − ∞ and is therefore

undefined. We deal with this problem by avoiding it: the measure µ − ν will be defined
DR

only when at least one of µ, ν is a finite measure, in which case the formula (46) always

makes sense. It is straightforward to check that, under this assumption, the set function

µ − ν is countably additive, and (µ − ν)(∅) = 0. It is of course not monotone. From

these observations we extract the definition of a signed measure:

Definition 18.1. Let (X, M ) be a measurable space. A signed measure is a function

ρ : M → R satisfying:

a) ρ(∅) = 0,

b) ρ takes at most one of the values +∞, −∞,

 MAA6616 COURSE NOTES FALL 2016 159
P∞
c) if (En )∞
n=1 is a disjoint sequence of measurable sets, then n=1 ρ(En ) converges

to ρ( ∞
S S∞
n=1 En ), and the sum is absolutely convergent if ρ( n=1 En ) is finite.

Remark 18.2. Actually, the statement about absolute convergence in (c) is an imme-

diate consequence of the Riemann rearrangement theorem. 

The main result of this section is the Jordan decomposition theorem, which says that

T
every signed measure is canonically the difference of two (unsigned) measures µ and ν.
AF
There is a natural partial order on the set of finite measures on (X, M ), determined by

µ ≥ ν if and only if µ − ν is a positive measure.

Example 18.3. Consider a measure space (X, M , µ) and let f : X → R belong to

DR

L1 (µ). The quantity

Z
µf (E) := f dµ (47)
E

then defines a signed measure on (X, M ). Indeed, decomposing f = f + − f − into its

positive and negative parts f = f + − f − , where f + = max(f, 0) and f − = − min(f, 0),

the signed measure µf can be written as ρ = µf + − µf − where µf ± denotes the measure

Z
µf ± (E) = f ± dµ. (48)
E

In fact this construction will work as long as f is semi-integrable (that is, at least one

of f + , f − is integrable). 4
160 MAA6616 COURSE NOTES FALL 2016

It is not hard to show that monotone and dominated convergence for sets still hold

for signed measures.

Proposition 18.4. Let ρ be a signed measure. If (En )∞

n=1 is an increasing sequence of

S∞
measurable sets, then ρ( n=1 En ) = limn→∞ ρ(En ). If En is a decreasing sequence of

measurable sets and ρ(E1 ) is finite, then ρ( ∞

T
n=1 En ) = limn→∞ ρ(En ). †

Proof. The proof is essentially the same as in the unsigned case (using the disjointifica-

tion trick) and is left as an exercise (Problem 19.18). 

T
Before going further we introduce some notation and a couple of definitions. If ρ is a
AF
signed measure on the measurable space (X, M ) and Y ⊂ X is a measurable set, we let

ρ|Y denote the measure on M defined by ρ|Y (E) := ρ(Y ∩ E). A set E totally positive

for ρ if ρ|E ≥ 0. As is easily verified, E is totally positive for ρ if and only if ρ(F ) ≥ 0
DR

for all F ⊂ E. In the case ρ omits the value +∞, it is totally positive ρ(F ) ≤ ρ(E) for

all measurable F ⊂ E. (Consider E \ F ). The set E totally negative for ρ if ρ|E ≤ 0 and

totally null if ρ|E = 0. It is immediate that E is totally null for ρ if and only if it is both

totally positive and totally negative. Finally, if (En )n is a sequence of totally positive

sets, then ∪En is also totally positive.

Note that when we decompose a real-valued function f into its positive and negative

parts f = f + − f − , the sets X+ := {x : f + (x) > 0} and X− := {x : f − (x) > 0} are

disjoint, and f |X+ ≥ 0, f |X− ≤ 0. (Compare with Example (18.3).) A similar statement

holds for signed measures.

 MAA6616 COURSE NOTES FALL 2016 161

Theorem 18.5 (Hahn Decomposition Theorem). Let ρ be a signed measure. Then there

exists a partition of X into disjoint measurable sets X = X+ ∪ X− such that ρ|X+ ≥ 0

and ρ|X− ≤ 0. Moreover if X+0 , X−0 is another such pair, then X+ ∆X+0 and X− ∆X−0 are

totally null for ρ.

The following lemma will be used in the proof of Theorem 18.5

Lemma 18.6. Let ρ be a signed measure that omits the value +∞. If ρ(G) > 0, then

T
there exists a subset E ⊂ G such that E is totally positive and ρ(E) > 0. †
AF
Proof. For notational convenience, let E1 = G. If E1 is totally positive, then there is

nothing to prove. Accordingly, suppose E1 is not totally positive. Thus E1 contains a

subset F of strictly larger positive measure. In particular, the set

DR

1
J1 = {n ∈ N+ : there is an F ⊂ E1 such that ρ(F ) ≥ ρ(E1 ) + }
n

is nonempty and thus has a smallest element n1 . In particular, if m < n1 and F ⊂ E1 is

measurable, then ρ(F ) < ρ(E1 )+ m1 . Choose any E2 ⊂ E1 such that ρ(E2 ) ≥ ρ(E1 )+1/n1 .

Now, if E2 were totally positive, then the proof is complete. Otherwise, let n2 denote

the smallest element of

1
J2 = {n ∈ N+ : there is an F ⊂ E2 such that ρ(F ) ≥ ρ(E2 ) + }
n

and choose E3 ⊂ E2 such that ρ(E3 ) ≥ ρ(E2 ) + 1/n2 . Continuing by induction produces

a totally positive subset E of G with ρ(E) > 0 or a decreasing sequence of measurable

162 MAA6616 COURSE NOTES FALL 2016

sets Ej+1 ⊂ Ej and a sequence of integers nj such that ρ(Ej ) > 0 for all j and

1
ρ(Ej+1 ) ≥ ρ(Ej ) + ;
nj

and
1
nj = min{n ∈ N+ : there is an F ⊂ Ej such that ρ(F ) ≥ ρ(Ej ) + }. (49)
n
T∞
Assuming this latter case, let E = j=1 Ej . We will show that ρ(E) > 0 and E is totally

positive. By Proposition 18.4, ρ(Ej ) increases to ρ(E) and in particular the set E has

finite positive measure (recall ρ omits the value +∞). Since ρ(E) is finite, the nj go to

T
infinity. To show that E must be totally positive, suppose, by way of contradiction, there
AF
exists an F ⊂ E such that ρ(F ) > ρ(E) and it can be assumed that ρ(F ) > ρ(E) + 1
m

where

1
m = min{n ∈ N+ : there is an F ⊂ E such that ρ(F ) ≥ ρ(E) + }.
n
DR

Thus F ⊂ Ej for every j and ρ(F ) ≥ ρ(E) + 1/m ≥ ρ(Ej ) + 1/m, which, since the nj

go to infinity, contradicts (49) once j is large enough. 

Proof of Theorem 18.5. We may assume ρ avoids the value +∞. The idea of the proof is

to select X+ to be a maximal totally positive set for ρ, and then show that X− := X \X+

is totally negative. The set X+ is obtained by a greedy algorithm. Let M denote

the supremum of ρ(E) over all totally positive sets E. Choose a sequence of totally

positive sets En so that M = lim ρ(En ). Since each En is totally positive, the union
S∞
X+ := n=1 En is also totally positive, and by Proposition 18.4 ρ(X+ ) = M . (In

particular, M is finite.)
 MAA6616 COURSE NOTES FALL 2016 163

The proof is finished if we can show that X− := X \ X+ is totally negative. By way

of contradiction, suppose it is not. In this case there exists a G ⊂ X− with ρ(G) > 0.

By Lemma 18.6, there exists a set E ⊂ G such that E is totally positive and ρ(E) > 0.

Now X+ ∪ E is totally positive and ρ(X+ ∪ E) > ρ(X+ ), contradicting the choice of X+ .

The uniqueness statement in the theorem is left as an exercise (Problem 19.19). 

A set E is a support set for a signed measure ρ if E c is totally null for ρ. Two signed

measures ρ, σ are mutually singular, denoted ρ⊥σ, if they have disjoint support sets; i.e.,

T
there exists disjoint measurable sets E and F such that E c is totally null for ρ and F c is
AF
totally null for σ. In the case ρ, σ are unsigned measures, they are mutually singular if

and only if there exists disjoint (measurable) sets E and F such that ρ(E c ) = 0 = σ(F c )

(in which case it can be assumed that F = E c if desired).

DR

Theorem 18.7 (Jordan Decomposition). If ρ is a signed measure on (X, M ), then there

exist unique positive measures ρ+ , ρ− such that ρ+ ⊥ρ− and ρ = ρ+ − ρ− .

Proof. Let X = X+ ∪X− be a Hahn decomposition for ρ and put ρ+ = ρ|X+ , ρ− = −ρ|X− .

It is immediate from the properties of the Hahn decomposition that ρ+ , ρ− have the

desired properties; uniqueness is left as an exercise (Problem 19.20). 

Example 18.8. Referring to Example 18.3, it is now immediate that the decomposition

mf = mf + − mf − is the Jordan decomposition of mf ; i.e., m+

f = mf + and likewise
164 MAA6616 COURSE NOTES FALL 2016

m−
f = mf − . Thus the Jordan decomposition theorem should be seen as a generalization

of the decomposition of a real-valued function into its positive and negative parts. 4

Let ρ be a signed measure and ρ = ρ+ − ρ− its Jordan decomposition. By analogy

with the identity |f | = f + + f − , we can define a measure |ρ| := ρ+ + ρ− ; this is called

the absolute value or total variation of ρ. The latter name is explained by the following

proposition.

T
Proposition 18.9. Let ρ be a signed measure on the measure space (X, M ). If E ∈
P∞
M , then |ρ|(E) = sup |ρ(En )|, where the supremum is taken over all measurable
AF n=1

S∞
partitions E = n=1 En . †

Proof. The proof is an exercise (Problem 19.22). 

DR

Warning: A moment’s thought shows that in general |ρ(E)| 6= |ρ|(E). As an

exercise, prove that |ρ(E)| ≤ |ρ|(E) always, with equality if and only if E is either

totally positive, totally negative, or totally null for ρ.

We can now define a signed measure ρ to be finite or σ-finite according as |ρ| is finite

or σ-finite. It is not hard to show that ρ is finite if and only if ρ(E) is finite for every E,

if and only if ρ+ , ρ− are finite unsigned measures. It is evident from this that the space

of finite signed measures on (X, M ) is a real vector space, denoted M (X). (We will see

later in the course that the quantity kρk := |ρ|(X) defines a norm on M (X), called the

total variation norm.)

 MAA6616 COURSE NOTES FALL 2016 165

A few remarks about integration against signed measures are in order. If ρ is a

signed measure, then L1 (ρ) is defined to be L1 (|ρ|); note that L1 (|ρ|) = L1 (ρ+ ) ∩ L1 (ρ− ).

For f ∈ L1 (ρ) define

Z Z Z
f dρ := f dρ+ − f dρ− . (50)

Proposition 18.10. Let ρ be a signed measure on (X, M ).

R  R
(a) If f ∈ L1 (ρ), then  f dρ ≤ |f | d|ρ|.

(b) If E ∈ M , then |ρ|(E) = sup{ E f dρ : |f | ≤ 1}.

R 

T
†
AF
Proof. Problem 19.23. 

18.2. The Lebesgue-Radon-Nikodym theorem. Fix for reference a measurable space

DR

(X, M ) and an unsigned measure m on this space. (In this section all measures are de-

fined on the same σ-algebra M .) For an unsigned measurable function f , we have the

measure
Z
mf (E) = f dm. (51)
E

The map f → mf is thus a map from the space of unsigned measurable functions into

the space of nonnegative measures on (X, M ). Likewise the mapping f → mf maps

L1 (µ) into the space of finite signed measures on X. One may ask if every finite measure

µ on X may be expressed as mf for some f , but one can quickly see this is not the case

in general. Indeed, if µ = mf then µ(E) = 0 whenever m(E) = 0, which need not always
166 MAA6616 COURSE NOTES FALL 2016

be the case (e.g., m is Lebesgue measure on R and µ is the point mass at 0.) However,

when the measures involved are σ-finite, it turns out this is the only obstruction.

Given unsigned measures µ and ν on a measure space (X, M ), we say ν is absolutely

continuous with respect to µ if (for measurable sets E) µ(E) = 0 implies ν(E) = 0. The

following proposition explains the reason for the terminology.

Proposition 18.11. If µ and ν are unsigned finite measures on the measurable space

(X, M ), then the following are equivalent.

T
(i) ν << µ;
AF
(ii) for every  > 0, there exists a δ > 0 such that ν(E) <  whenever µ(E) < δ.

†
DR

Proof. Suppose (ii) hold, ν(A) > 0 and ν(A) >  > 0. By assumption there is δ > 0

such that µ(E) < δ implies ν(E) < epsilon. Hence µ(A) > δ > 0 and (i) holds (we did

not need to know the measures were finite here).

Now suppose (ii) fails. There is an  > 0 such that for each n ∈ N+ there exists a

set En such that µ(En ) < 2−n , but ν(En ) ≥ . Let Fn = ∪k≥n En . Thus µ(Fn ) < 2−n+1 .

Hence, with F = ∩Fn , we have, by dominated convergence for sets twice (here is where

finiteness of the measures is used),

µ(F ) = lim µ(Fn ) = 0, ν(F ) = lim µ(Fn ) ≥ .

n n


 MAA6616 COURSE NOTES FALL 2016 167

To see that the σ-finiteness assumption is needed in the implication (ii) implies (i)

of Proposition 18.11, let µ and ν be the measures on (N, 2N ) determined by

µ({n}) = 2−n , ν({n}) = 1.

The only set for either measure is ∅. Hence ν << µ. But with 0 <  < 1, there is no δ

such that µ(E) < δ implies ν(E) < .

Theorem 18.12 (Radon-Nikodym). Suppose m and µ are unsigned σ-finite measures

T
on (X, M ). If µ << m, then there is a unique measurable f : X → [0, ∞) such that

µ = mf . Further, if µ is finite, then f ∈ L1 (m).

AF
Remark 18.13. In the case µ is unsigned, apply Theorem 18.12 to the µ± from its

Jordan decomposition. In this case f is semi-integrable.

DR

To see that the σ-finiteness hypothesis is needed, let m denote Lebesgue measure

on (R, L) and c counting measure. If c(E) = 0, then E = ∅ and m(E) = 0 too. Hence

m << c. On the other hand, if f is unsigned and measurable, then, for each x ∈ R,
Z
cf ({x}) = f dc = f (x).
{x}

Hence, cf 6= m. 

Lemma 18.14. If µ and ν are finite positive measures on (X, M ), then either µ⊥ν, or

else there exist  > 0 and a measurable set E such that µ(E) > 0 and ν(F ) ≥ µ(F ) for

F ⊂ E measurable (that is, E is totally positive for ν − µ).

168 MAA6616 COURSE NOTES FALL 2016

In particular, if ν << µ and ν 6= 0, then there exist  > 0 and a measurable set E

such that µ(E) > 0 and ν(F ) ≥ µ(F ) for F ⊂ E measurable. †

Proof. For each n ≥ 1, let X = X+n ∪ X−n be a Hahn decomposition for ν − n1 µ. Let
S∞ n
T∞
P = n=1 X+ and N = n=1 X−n . In particular N = P c . Since N is totally negative for

ν − n1 µ for all n, it follows that ν(N ) = 0. If µ(P ) = 0, then µ⊥ν. Otherwise, µ(X+n ) > 0

for some n, and by construction X+n is totally positive for ν − n1 µ (thus we take  = n1 ,

E = X+n ).

T

AF
Proof of Theorem 18.12. We prove this only for the case that µ, m are finite; and only

outline the extension to the σ-finite case.

As before, f is selected by a ‘greedy algorithm. Let S denote the set of unsigned f

DR

such that mf ≤ µ and observe 0 ∈ S . Let M be the supremum of the set {

R
X
f dm :

f ∈ S }. Note that M is finite, since µ is. Choose a sequence fn so that

R
X
fn dm → M .

Define gn = max1≤k≤n fk and note that fn ≤ gn and the gn are increasing. An exercise

shows if g, h ∈ S , then mϕ ≤ µ where ϕ = max{g, h} from which mgn ≤ µ follows.

Hence each gn ∈ S . Since (gn ) is pointwise increasing, it converges pointwise (in [0, ∞])
R
to some unsigned measurable g. Since, E
gn dm ≤ µ(E) for each E and n, the MCT

implies g ∈ S (and in particular g is finite a.e.). Since

R R
X
fn dm ≤ X
gn dm ≤ M for
R
each n, it follows that the sequence ( X gn dm) converges to M and hence

Z
M= g dm.
X
 MAA6616 COURSE NOTES FALL 2016 169

Now choose any f ∈ S that achieves this maximum M . In particular, f is unsigned

and in L1 (m) (its integral is finite). By construction ν = µ − mf is totally positive since

f ∈ S and ν is absolutely continuous with respect to m since both µ and mf are. If

ν 6= 0, then then by Lemma 18.14 there is an  > 0 and an E such that m(E) > 0 and

ν − m|E ≥ 0, equivalently ν ≥ m|E . Hence µ ≥ mf +1E , contradicting the maximality

R R
of f since X
(f + 1E ) dm = X
f dm + m(E) > M .

Still assuming the measures are finite, to see that f is unique, suppose µ = mf = mg

T
for some unsigned g. It follows that g ∈ L1 (m) and hence, for all measurable sets E,
Z
AF 0 = mf (E) − mg (E) = (f − g) dm
E

and thus f = g a.e. m.

Now suppose µ and m are sigma-finite. In this case there exists set Xn ⊂ Xn+1 such
DR

that X = ∪Xn and µ(X), m(X) < ∞. Letting µn (E) = µ(E ∩ Xn ) and mn (E) = m(E ∩

Xn ), it follows that µn << mn and hence there exists a unique unsigned measurable

function fn supported on Xn such that if E ⊂ Xn is measurable, then

Z
µ(E) = fn dm.
E

Since also, for E ⊂ Xn ⊂ Xn+1 ,

Z
µ(E) = fn+1 dm,
E

it follows that fn+1 = fn a.e. m on Xn (by uniqueness). Hence the sequence (fn ) is

pointwise increasing. Choose f to be its limit. By the Monotone Convergence Theorem,

170 MAA6616 COURSE NOTES FALL 2016

for measurable sets E ⊂ X,

Z
µ(E) = lim µ(E ∩ Xn ) = lim fn dm
E∩Xn
Z Z
= lim fn dm = f dm = mf (E).
E E


The function f is called the Radon-Nikodym derivative of µ with respect to m,

dµ
denoted dm
= f . The basic manipulations suggested by the derivative notation are

valid. For example it is easy to check that d(µ1 + µ2 )/dm = dµ1 /dm + dµ2 /dm. We will

T
see in a moment that the chain rule is valid.
AF
Corollary 18.15. Assuming µ and ν are (unsigned) σ-finite measures on the measurable

space (X, M ), µ << ν if and only if there is an unsigned measurable function f such

that µ = νf . †
DR

One important corollary of the Lebesgue-Radon-Nikodym theorem is the existence

of conditional expectations.

Proposition 18.16. Let (X, M , µ) be a σ-finite measure space (µ a positive measure),

N a sub-σ-algebra of M , and suppose ν = µ|N is σ-finite. If f ∈ L1 (µ) then there

exists g ∈ L1 (ν) (unique modulo ν-null sets) such that

Z Z
f dµ = g dν
E E

for all E ∈ N (g is called the conditional expectation of f on N ). †

Proof. Problem 19.27 

 MAA6616 COURSE NOTES FALL 2016 171

Proposition 18.17 (Chain rule for Radon-Nikodym derivatives). Suppose µ, m, λ are

σ-finite positive measures on the measure space (X, M ), and µ  λ  m.

(i) If g ∈ L1 (|µ|), then g dµ

dλ
∈ L1 (λ) and
Z Z
dµ
g dµ = g dλ. (52)
X X dλ

(ii)
dµ dµ dλ
= m-a.e. (53)
dm dλ dm

T
AF †

dµ dλ
Proof. In this case the Radon-Nikodym derivatives dλ
and dm
are unsigned. In particu-

lar, for measurable sets E,

Z Z
dµ
µ(E) = dµ = dλ.
E E dλ
DR

By Problem 13.18, if either g is unsigned or g ∈ L1 (µ), then

Z Z
dµ
g dµ = g dλ (54)
X X dλ

proving (i). As a warm up for (ii), note that (i) gives, for h either unsigned or h ∈ L1 (λ),
Z Z
dλ
h dλ = h dm.
X X dm

With h = 1E dµ
dλ
for a measurable set E,
Z Z
dµ dµ dλ
µ(E) = dλ = dm.
E dλ E dλ dm
dµ dµ
R
On the other hand, µ  m, and µ(E) = E dm
dm by the definition of dm
. Hence (ii)

follows by uniqueness of the Radon-Nikodym derivative. 

172 MAA6616 COURSE NOTES FALL 2016

Theorem 18.18 (Lebesgue Decomposition). If ν and µ are unsigned σ-finite measures

on the measure space (X, M ), then there exist unique measures µa and µs such that

µa << ν and µs ⊥ν and µ = µa + µs .

As before, using the Jordan decomposition one can allow ν to be a signed measures

in the Lebesgue decomposition.

Proof. Let m = ν +µ. In particular, m is σ-finite and µ << m. Hence there is a uniquely

T
(a.e. m) determined unsigned function f such that µ = mf . Thus, for all measurable E,
Z Z
AF µ(E) =
E
f d(µ + ν) =
E
f dm.

From here we prove the result under the added assumption that µ and ν are finite,

leaving the proof of the general case to the interested reader.

DR

Let Fa = {f < 1} and Fs = {f = 1} and define

µt (E) = µ(E ∩ Ft ),

for t = a, s. Both are measures

An easy argument shows f ≤ 1 a.e. m. Hence µ = µa + µs . Next,

Z
µ(Fs ) = f d(µ + ν) = µ(Fs ) + ν(Fs ).
Fs

Thus ν(Fs ) = 0 and µa ⊥ν. Now suppose E is measurable and ν(E) = 0. Letting

Fn = {f < 1 − n1 } (for positive integers n),

1
µ(E ∩ Fn ) ≤ (1 − )µ(E ∩ Fn ).
n
 MAA6616 COURSE NOTES FALL 2016 173

Hence µ(E ∩ Fn ) = 0. Since E ∩ Fs = ∪(E ∩ Fn ) it follows that µ(E ∩ Fs ) = 0 and

therefore µs << ν.

To prove uniqueness, suppose µ = ρa + ρs with ρa << ν and ρs ⊥ν. In particular,

there is a set G such that ρs (G) = 0 and ν(Gc ) = 0. In particular, ρa (Gc ) = 0 = µa (Gc )

and further

ρa (E) = ρa (E ∩ G), ρs (E) = ρs (E ∩ Gc ).

T
It follows that
AF
ρa (E) = ρa (E ∩ G ∩ F ) = µa (E ∩ F ∩ G) = µa (E).

Thus, if µ(E) < ∞, then ρs (E) = µs (E) too. The general case now follows from the
DR

σ-finiteness assumption. 

Example 18.19. This example shows the σ-finiteness hypothesis is needed in the

Lebesgue decomposition. On (R, 2R ), let c denote counting measure and let ν denote

the measure defined by ν(E) = 1 if E is uncountable and ν(E) = 0 otherwise. Arguing

by contradiction, suppose c = ca + cs where ca << ν and cs ⊥ν. It follows that there is a

set B such that cs (B c ) = 0 and ν(B) = 0. In particular, B is at most countable. There

is an x ∈ B c (since R is not countable). Now ν({x}) = 0 and hence ca ({x}) = 0 and

also cs ({x}) = 0 since x ∈ B c . Hence ca ({x}) + cs ({x}) = 0 6= c({x}) = 1. 4

174 MAA6616 COURSE NOTES FALL 2016

18.3. Lebesgue differentiation revisited. Finally, we describe the connection be-

tween Radon-Nikodym derivatives and Lebesgue differentiation on Rn . Recall a positive

measure µ is regular if

i) µ(K) < ∞ for every compact K ⊂ Rn , and

ii) for every Borel set E ⊂ Rn , we have µ(E) = inf{µ(U ) : U open , E ⊂ U }.

Theorem 18.20. Let µ be a regular Borel measure on Rn with Lebesgue decomposition

T
AF µ = mf + µs

with respect to Lebesgue measure m. For m-a.e. x ∈ Rn .

µ(Br (x))
lim = f (x) (55)
r→0 m(Br (x))

Proof. By the regularity of µ, we see that the measure mf is locally finite, so f ∈ L1loc .
DR

One may verify that the measure mf is regular, and so µs is as well. Applying the

Lebesgue differentiation theorem, (55) holds already with µ = mf , so it suffices to prove

that

µs (Br (x))
lim = 0 m − a.e. (56)
r→0 m(Br (x))

for the singular part µs .

Fix a Borel set E such that µs (E) = m(E c ) = 0 and let

 
µs (Br (x)) 1
Ek = x ∈ E : ∀t > 0 ∃ 0 < r < t such that > .
m(Br (x)) k

It will suffice to prove that m(Ek ) = 0 for each integer k ≥ 1.

 MAA6616 COURSE NOTES FALL 2016 175

By regularity, for given  > 0 there is an open set U containing E such that µs (U ) <

. By the definition of Ek , for each x ∈ Ek there is a ball Bx centered at x such

m(Bx ) S
that Bx ⊂ U and µs (Bx ) > k
. Let V = x∈Ek Bx be the union of these balls.

Fix a number c < m(V ) and apply Wiener’s covering lemma 17.6 to obtain points

x1 , . . . xm ∈ Ek such that the balls B1 , . . . Bm are disjoint and

m
X m
X
n n
c<3 m(Bj ) ≤ 3 k µs (Bj ) ≤ 3n kµs (V ) ≤ 3n kµs (U ) < 3n k.
j=1 j=1

Thus m(V ) ≤ 3n k, and since Ek ⊂ V and  was arbitrary, we conclude m(Ek ) = 0. 

T
19. Problems
AF
19.1. Product measures.

Problem 19.1. Let µX denote counting measure on X. Prove that if X, Y are both at
DR

most countable, then 2X ⊗ 2Y = 2X×Y and µX × µY = µX×Y .

Problem 19.2. Prove that the product measure construction is associative: that is, if

(Xj , Mj , µj ), j = 1, 2, 3 are σ-finite measure spaces, then (M1 ⊗ M2 ) ⊗ M3 = M1 ⊗

(M2 ⊗ M3 ), and (µ1 × µ2 ) × µ3 = µ1 × (µ2 × µ3 ).

Problem 19.3. Let X = Y = [0, 1], M = B[0,1] , N = 2R , let µ Lebesgue measure on

M , and let ν counting measure on N . Let ∆ denote the diagonal ∆ = {(x, x) : x ∈

[0, 1]} ⊂ [0, 1] × [0, 1]. Prove that ∆ ∈ M ⊗ N and

Z Z  Z Z 
1E (x, y) dν(y) dµ(x), 1E (x, y) dµ(x) dν(y) (57)
X Y Y X
176 MAA6616 COURSE NOTES FALL 2016

are unequal. Show that, for each P ∈ M ⊗ N , the functions f (x) = ν(Px ) and

g(y) = µ(P y ) are measurable (with respect to M and N respectively) and

Z Z
τ (P ) = ν(Px ) dµ, ρ(P ) = µ(P y ) dν
X Y

are both measures M ⊗ N . (Note that Theorem 15.7 does not (directly) apply).

Problem 19.4. Prove Proposition 15.11.

Problem 19.5. Prove Corollary 15.8.

T
Problem 19.6. Prove Corollary 15.14.
AF
19.2. Integration on Rn .

Problem 19.7. Compare the three integrals

DR

ZZ Z 1 Z 1  Z 1 Z 1 
2
f dm , f (x, y) dx dy, f (x, y) dy dx (58)
[0,1]2 0 0 0 0

for the functions

x2 − y 2
a) f (x, y) =
(x2 + y 2 )2
b) f (x, y) = (1 − xy)−s , s > 0

R1
Problem 19.8. Prove that if f ∈ L1 [0, 1] and g(x) = x
t−1 f (t) dt, then g ∈ L1 [0, 1]
R1 R1
and 0
g(x) dx = 0
f (x) dx.

R∞ Rb
Problem 19.9. Prove that 0
| sinx x | dx = +∞, but the limit limb→+∞ 0
sin x
x
dx exists

and is finite. (For a bigger challenge, show that the value of the limit is π2 .)
 MAA6616 COURSE NOTES FALL 2016 177

Problem 19.10. Prove Theorem 16.1.

Problem 19.11. Complete the proof of Theorem 16.6.

Problem 19.12. [The Gamma function] Define

Z ∞
Γ(x) := tx−1 e−t dt (59)
0

a) Prove that the function t → tx−1 e−t is absolutely integrable for all fixed x > 0

(thus Γ(x) is defined for all x > 0).

T
b) Prove that Γ(x + 1) = xΓ(x) for all x > 0.
AF
c) Compute Γ(1/2). (Hint: if you haven’t seen this before, first make the change of
√
variables u = t, then evaluate the square of the resulting integral using Tonelli’s

theorem and polar coordinates.)

√
DR

d) Using (b) and (c), conclude that Γ(n+ 12 ) = (n− 12 )(n− 32 ) · · · ( 12 ) π for all n ≥ 1.

Problem 19.13. Complete the following outline to prove that

2π n/2
σ(S n−1 ) = . (60)
Γ(n/2)

a) Show that if f ∈ L1 (Rn ) and f is a radial function (that is, f (x) = g(|x|) for

some function g : [0, ∞) → C), then

Z Z ∞
n−1
f (x) dx = σ(S ) g(r)rn−1 dr. (61)
Rn 0

b) Show that for all c > 0,

Z  π n/2
2
e−c|x| dx = . (62)
Rn c
178 MAA6616 COURSE NOTES FALL 2016

2 Qn 2
(Hint: write e−c|x| = j=1 e−c|xj | and use Tonelli’s theorem.)

c) Finish by combining (a) and (b). (Using the results on the Gamma function

from the previous exercise, one finds that σ(S n−1 ) is always a rational multiple

of an integer power of π.)

19.3. Differentiation theorems.

Problem 19.14. Prove that if 0 6= f ∈ L1 (R), then there exist constants C, R > 0

T
(depending on f ) such that

C
AF Mf (x) ≥
|x|
for all |x| > R. (63)

(Hint: reduce to the case f = 1E where E is a bounded set of positive measure.)

Conclude that Mf never belongs to L1 (R) if f ∈ L1 is not a.e. 0.

DR

Problem 19.15. The Lebesgue differentiation theorem says that for f ∈ L1 (Rn ), we

have Ar,f → f pointwise a.e. as r → 0. Prove that also Ar,f → f in the L1 norm.

(Hint: the proof can be done in three steps: first prove this under the assumption that

f is continuous with compact support. Then prove that for all f ∈ L1 and r > 0, the

functions Ar,f ∈ L1 ; in fact kAr,f k1 ≤ kf k1 for all r. Tonelli’s theorem will help. Finally,

to pass to general L1 functions, use a density argument.)

Problem 19.16. Let E be a Borel set in R. Define the density of E at x to be

m(E ∩ B(x, r))

DE (x) = lim
r→0 m(B(x, r))
 MAA6616 COURSE NOTES FALL 2016 179

whenever the limit exists.

a) Show that DE (x) = 1 for a.e. x ∈ E and DE (x) = 0 for a.e. x ∈

/ E.

b) Give examples of E and x for which DE (x) = α (0 < α < 1) and for which

DE (x) does not exist.

Problem 19.17. Define the decentered Hardy-Littlewood maximal function for f ∈

L1 (Rn ) by
Z
1

T
Mf∗ (x) = sup |f (x)| dx (64)
B m(B) B

where the supremum is taken over all open balls containing x (not just those centered
AF
at x). Prove that

Mf ≤ Mf∗ ≤ 2n Mf . (65)
DR

19.4. Signed measures and the Lebesgue-Radon-Nikodym theorem.

Problem 19.18. Prove Proposition 18.4.

Problem 19.19. Complete the proof of Theorem 18.5.

Problem 19.20. Prove the uniqueness statement in the Jordan decomposition theorem.

(Hint: if also ρ = σ+ − σ− , use σ± to obtain another Hahn decomposition of X.)

P
Problem 19.21. Prove that if νj ⊥µ for j ∈ N then ( j νj )⊥µ, and if νj  µ for j ∈ N
P
then ( j νj )  µ.
180 MAA6616 COURSE NOTES FALL 2016

Problem 19.22. Complete the proof of Proposition 18.9.

Problem 19.23. Prove Proposition 18.10.

Problem 19.24. Complete the proof of Theorem 18.12 in the σ-finite case.

Problem 19.25. Complete the proof of the (i) =⇒ (iii) implication in Corollary 18.15

Problem 19.26. Suppose ρ is a signed measure on (X, M ) and E ∈ M . Prove that

T
a) ρ+ (E) = sup{ρ(F ) : F ∈ M , F ⊂ E} and ρ− (E) = − inf{ρ(F ) : F ∈ M , F ⊂

E}
AF
b) |ρ|(E) = sup{ n1 |ρ(Ej )| : E1 , . . . En are disjoint and ∪n1 Ej = E}
P

Problem 19.27. a) Prove Proposition 18.16. b) In the case µ = Lebesgue measure on

DR

[0, 1), fix a positive integer k and let N be the sub-σ-algebra generated by the intervals

[ kj , j+1
k
) for j = 0, . . . k − 1. Give an explicit formula for the conditional expectation g

in terms of f . c) Show that the σ-finite hypothesis on ν is needed.

19.5. The Riesz-Markov Theorem.

Problem 19.28. Explain how to construct Lebesgue measure on [0, 1] from the Riemann

integral and Theorem 14.2.

Problem 19.29. Suppose X is a locally compact abelian topological group (the defini-

tions are available online). Given y ∈ X, let ty : X → X denote translation by y so that

 MAA6616 COURSE NOTES FALL 2016 181

ty (x) = x + y (the group is abelian so the group operation is written as +). A linear

functional λ : Cc (X) → C is translation invariant if λ(f ) = λ(f ◦ ty ) for each y ∈ X and

f ∈ C(X). Prove, if λ is a positive linear functional that is translation invariant, then

the representing measure µ for λ from Theorem 14.2 is translation invariant.

Problem 19.30. Let X be a compact Hausdorff space. Fix p ∈ X and consider the

linear functional Ep : C(X) → C defined by Ep (f ) = f (p). Show Ep is positive and

determine the representing measure for Ep .

T
AF
DR