Professional Documents
Culture Documents
edu/bban
lev.ban@ttu.edu
Real Analysis
Byeong Ho Ban
Mathematics and Statistics
Texas Tech University
Chapter 1. Measure
1.
Proof.
2.
Proof.
Proof. (−→)
Suppose that A is a σ−algebra and S∞that {Ei }∞
1 ⊂ A where E1 ⊂ E2 ⊂ · · · .
Then, by definition of σ−algebra, 1 Ei ∈ A.
(←−)
Suppose that ∞ ∞
S
1 Ei ∈ A whenever {Ei }1 ⊂ A and E1 ⊂ E2 ⊂ · · · . Since A is an algebra, A is closed
under complement. So it suffices to prove that A is closed under countable union. Suppose that {Fi }∞
1 is
S2 S3 Sk Sk
a sequence in A. Observe that F1 ⊂ 1 Fi ⊂ 1 Fi ⊂ · · · ⊂ 1 Fi ⊂ · · · . And note that 1 Fi ∈ A since
A is algebra so closed under finite union.
Thus, since A is closed under countable increasing unions, ∞
S Sk S∞
1 1 Fi = 1 Fi ∈ A. Therefore, A is closed
under countable union so is σ−algebra.
Let {Ei }∞1 ⊂ S. Note that, ∀i ∈ N ∃Fi ⊂ E such that each Fi are countable and Ei ∈ Fi . From
set theory, we know that
V= ∞
S
1 Fi
It is obvious that S ⊂ M since each M(F) are subsets of M. Now, observe that E ⊂ S because for
any E ∈ E, {E} is a countable set, so M({E}) ⊂ S and E should be in S. Thus, E ⊂ S ⊂ M. Therefore,
S = M(S) = M(E) = M.
∴S=M
6.
Proof.
7.
Proof.
8.
Proof.
3
10. Given a measure space (X, M, µ) and E ∈ M, define µE (A) = µ(A ∩ E) for A ∈ M. Then µE is a
measure.
Proof. In order to prove that µE is a measure, we need to show that µE satisfies the following two properties.
(1) µE (∅) = 0
S∞
(2) µE ( 1 Ei ) = Σ ∞
1 µE (Ei ) where Ei 6= Ej if i 6= j
µE ( ∞
S S∞ S∞
1 Ei ) = µ(E ∩ 1 Ei ) = µ( 1 E ∩ Ei )
Therefore,
µE ( ∞ ∞
S
1 Ei ) = Σ1 µE (E)
Thus, µE is a measure.
11.
Proof.
12.
Proof.
13.
Proof.
14.
Proof.
4
15.
Proof.
16.
Proof.
Thus, in order to prove the statement, we need to show the reverse inequality.
Reminding that each Aj are µ∗ −measurable, for any n ∈ N observe the below.
n
! n
! n
!
[ [ [
µ∗ E ∩ ( Aj ) = µ∗ E ∩ ( Aj ) ∩ A1 + µ∗ E ∩ ( Aj ) ∩ Ac1
1 1 1
n
!
[
= µ∗ (E ∩ A1 ) + µ∗ E ∩ ( Aj )
2
n
! n
!
[ [
= µ∗ (E ∩ A1 ) + µ∗ E ∩ ( Aj ) ∩ A2 + µ∗ E ∩ ( Aj ) ∩ Ac2
2 2
n
!
[
= Σ21 µ∗ (E ∩ Aj ) + µ∗ E ∩ ( Aj )
3
= · · · = Σn1 µ∗ (E ∩ Aj )
Due to monotonicity which is a property of outer measure,
∞
! n
!
[ [
µ∗ E ∩ ( Aj ) ≥ µ∗ E ∩ (Aj ) ≥ Σn1 µ∗ (E ∩ Aj )
1 1
18. Let A ⊂ P(X) be an algebra, Aσ the collection of countable unions of sets in A, and Aσδ the collec-
tion of countable intersection of sets in Aσ . Let µ0 be a premeasure on A and µ∗ the induced outer measure.
a. For any E ⊂ X and > 0, there exists A ∈ Aσ with E ⊂ A and µ∗ (A) ≤ µ∗ (E) +
b. If µ∗ (E) < ∞, then E is µ∗ − measurable iff there exists B ∈ Aσδ with E ⊂ B and µ∗ (B \ E) = 0.
c. If µ0 is σ-finite, the restriction µ∗ (E) < ∞ in (b) is superfluous.
5
Here, let A = j∈N Aj , then A ∈ Aσ and by the property of infimum, µ∗ (A) ≤ Σj∈N µ0 (Aj ), so
S
µ∗ (A) ≤ µ∗ (E) +
Therefore, the A is the set we were finding.
( ⇐= ) Suppose that there exists B ∈ Aσδ with E ⊂ B and µ∗ (B \ E) = 0. Observe that Aσδ ⊂
M(A). Since, by a theorem in our book, M(A) is s subset of a collection of all µ∗ −measurable sets, B is
µ∗ −measurable. Also, by Caratheodory’s Theorem, B \ E is also µ∗ −measurable. Thus,
B \ (B \ E) = B ∩ (B ∩ E c )c = B ∩ (B c ∪ E) = B ∩ E = E
is also µ∗ −measurable.
Proof for (c).
( =⇒ ).
Suppose that µ0 is σ−finite and E is µ∗ −measurable. Thus, X = n∈N En where µ∗ (En ) = µ0 (En ) < ∞
S
∀n ∈ N
By part (a), we can choose a sequence {Aij }i,j∈N ⊂ Aσ such that (En ∩ E) ⊂ Ank and µ∗ (En ∩ E) ≤
µ∗ (Ank ) < µ∗ (En ∩ E) + k21n ∀n, k ∈ N
Here, we know that µ∗ (En ∩ E) ≤ µ∗ (En ) < ∞ ∀n ∈ N, thus,
1
µ∗ (Ank \ (En ∩ E)) = µ∗ (Ank ) − µ∗ (En ∩ E) <
k2n
6
S S
Let Ak = n∈N Ank , then Ak ∈ Aσ and E ⊂ Ak . Now, note that Ak \ E ⊂ n∈N (Ank \ (E ∩ En )). Thus,
observe that
1
µ∗ (Ak \ E) ≤ Σn∈N µ∗ (Ank \ (E ∩ En )) <
k
T
Let B = k∈N Ak , then B ∈ Aσδ and E ⊂ B. Observe that
1
µ∗ (B \ E) ≤ µ∗ (Ak \ E) <
k
Therefore, if we send k to ∞, we get µ∗ (B \ E) = 0.
( ⇐= ).
Note that we have not used the condition either ”µ∗ (E) < ∞” or ”µ0 is a σ−finite” in the proof of part
(b). Thus, our proof in the part (b) is still valid. Below is just the copy of the proof.
Suppose that there exists B ∈ Aσδ with E ⊂ B and µ∗ (B \ E) = 0. Observe that Aσδ ⊂ M(A). Since, by
a theorem in our book, M(A) is s subset of a collection of all µ∗ −measurable sets, B is µ∗ −measurable.
Also, by Caratheodoy’s Theorem, B \ E is also µ∗ −measurable. Thus,
B \ (B \ E) = B ∩ (B ∩ E c )c = B ∩ (B c ∪ E) = B ∩ E = E
is also µ∗ −measurable.
19. Let µ∗ be an outer measure on X induced from a finite premeasure µ0 . If E ⊂ X, define the
inner measure of E to be µ∗ (E) = µ0 (X) − µ∗ (E c ). Then E is µ∗ −measurable iff µ∗ (E) = µ∗ (E)
Proof. Note that µ∗ is finite since µ0 (X) = µ∗ (X).
( =⇒ )
Suppose that E is µ∗ −measurable. Then
µ∗ (X) = µ∗ (X ∩ E) + µ∗ (X ∩ E c )
= µ∗ (E) + µ∗ (E c )
Since µ∗ (E c ) < ∞
( ⇐= )
Suppose that µ∗ (E) = µ∗ (E).
By part (a) of Homework question 2, for each n ∈ N, we can choose Bn ∈ Aσ with E ⊂ Bn such that
1
µ∗ (Bn ) ≤ µ∗ (E) +
n
T
Let B = n∈N Bn , then B ∈ Aσδ and by monotonocity,
1
µ∗ (B) ≤ µ∗ (Bn ) ≤ µ∗ (E) +
n
If we take a limit as n → ∞, µ∗ (B) ≤ µT
∗
(E).
Since E ⊂ Bn for any n ∈ N, E ⊂ B = n∈N Bn . So, by monotonicity, µ∗ (E) ≤ µ∗ (B)
Therefore, µ∗ (E) = µ∗ (B).
7
Note that Aσδ is a subset of the collection of all µ∗ −measurable sets. So B is µ∗ −measurable. Thus,
µ∗ (X) = µ∗ (X ∩ B) + µ∗ (X ∩ B c )
= µ∗ (B) + µ∗ (B c )
= µ∗ (E) + µ∗ (B c )
= µ∗ (E) + µ∗ (B c )
= µ∗ (X) − µ∗ (E c ) + µ∗ (B c )
Since µ∗ is finite,
µ∗ (E c ) = µ∗ (B c )
Observe that, since B is µ∗ −measurable,
µ∗ (E c ) = µ∗ (E c ∩ B) + µ∗ (E c ∩ B c )
= µ∗ (B \ E) + µ∗ (B c ) (∵ B c ⊂ E c )
= µ∗ (B \ E) + µ∗ (E c )
Since µ∗ (E c ) < ∞, µ∗ (B \ E) = 0.
By part (b) of homework question 2, since we found the set B, it is true that E is µ∗ −measurable.
20.
Proof.
21.
Proof.
22.
Proof.
23.
Proof.
24. Let µ be a finite measure on (X, M), and let µ∗ be the outer measure induced by µ. Suppose that
E ⊂ X satisfies µ∗ (E) = µ∗ (X) (but not that E ∈ M).
µ∗ (E ∪ B) = µ∗ (E ∪ A) = µ∗ (X)
µ∗ (X) = µ∗ (E ∩ B) + µ∗ (E ∪ B) − µ∗ (B)
= µ∗ (E ∩ A) + µ∗ (E ∪ A) − µ∗ (B)
= µ∗ (X) + µ∗ (A) − µ∗ (B)
Proof for b
Firstly, ME is closed under complement.(Set subtraction from E)
Let A ∩ E ∈ ME , and observe that
E \ (A ∩ E) = E ∩ (A ∩ E)c
= E ∩ (Ac ∪ E c ) = Ac ∩ E
Since Ac ∈ M, Ac ∩ E ∈ ME , so E \ (A ∩ E) ∈ ME
Secondly, it is closed under countable union.
Let {An }n∈N ⊂ M, then {An ∩ E}n∈N is a sequence of sets in ME . Now, observe that
!
[ [ [
(An ∩ E) = An ∩ E ∈ ME (∵ An ∈ M)
n∈N n∈N n∈N
We still need to prove the countable additivity. Let {An ∩ E}n∈N ⊂ ME be a sequence of disjoint sets.
S that{An }n∈N ⊂ M is not a sequence of disjoint sets.
However, it is possible
n−1
So let Bn = An \ j=1 Aj ∈ M.
Observe the below
n−1
!
\
Bn ∩ E = An ∩ E ∩ Acj
j=1
n−1
!
\
= (An ∩ E) ∩ (Acj ∩ E)
j=1
= An ∩ E (∵ (An ∩ E) ⊂ (Acj ∩ E) if n 6= j)
∴ An ∩ E = Bn ∩ E ∀n ∈ N
!
[
Σn∈N ν(An ∩ E) = Σn∈N ν(Bn ∩ E) = Σn∈N µ(Bn ) = µ Bn
n∈N
!! !! !
[ [ [
=ν E∩ Bn =ν E∩ An =ν An ∩ E
n∈N n∈N n∈N
25.
Proof.
27. Prove Proposition 1.22a. (Show that if x, y ∈ C and x < y, there exists z 6∈ C such that x < z < y)
Proposition 1.22a
C is compact, nowhere dense, and totally disconnected(i.e., the only connected subsets of C are single
points). Moreover, C has no isolated points.
Proof. .
1. C is compact.
(→) Note that C ⊂ [0, 1]. Also, note that C c is open since it is a countable disjoint unions of open intervals,
so C is closed. Since C is closed and bounded in R, C is compact.
10
2. C is totally disconnected.
(→) Let x, y ∈ C and x < y. By way of contradiction, assume that ∀z ∈ (x, y), z ∈ C. Note that
(x, y) ⊂ C . Also, note that y − x > 0 since x 6= y. Thus, by monotonicity, m(C) ≥ y − x > 0. But it is
contradiction to Proposition 1.22b. Thus, there should exist z ∈ (x, y) such that z 6∈ C
4. C is nowhere dense.
(→) Let x ∈ C and > 0. Since C has no isolated point, ∃y ∈ C such that y ∈ (x − , x + ). Also
observe that, since C is totally disconnected, ∃z ∈ (x, y) ⊂ (x − , x + ) such that z 6∈ C. Note that
(x, y) ⊂ (x − , x + ) and (x, y) 6⊂ C. Thus, (x − , x + ) cannot be a subset of C. Therefore, there is no
> 0 such that (x − , x + ) ⊂ C, so C has no interior point and C is nowhere dense.
28. Let F be increasing and right continuous, and let µF be the associated measure.
Then
µF ({a}) = F (a) − F (a−),
µF ([a, b)) = F (b−) − F (a−),
µF ([a, b]) = F (b) − F (a−),
and µF ((a, b)) = F (b−) − F (a)
Proof. .
Due to the continuity from above of µF , for any a, b ∈ R where a ≤ b, observe that
!
\ 1 1
µF ([a, b]) = µF (a − , b] = lim µF ((a − , b])
n∈N
n n→∞ n
1
= lim (F (b) − F (a − ))
n→∞ n
1
= F (b) − lim F (a − )
n→∞ n
= F (b) − F (a−)
Also, observe that
µF ({a}) = µF ([a, a]) = F (a) − F (a−)
Since ∞ ≤ 2 is impossible, it is contradiction. Thus, since E is Lebesgue measurable, the only possible
situation is when m(E) = 0.
Solution for b.
Note that, since
0 < m(E) = Σn∈Z m(E ∩ [n, n + 1])
∃n ∈ Z such that m(E ∩ [n, n + 1]) > 0.
Let E0 = {x − n : x ∈ E ∩ [n, n + 1]}, then E0 ⊂ [0, 1]. Since Lebesgue measure is translation invariant,
it is enough to prove it by assuming that E ⊂ [0, 1].
30.
Proof.
31.
12
Proof.
32.
Proof.
33.
Proof.
34.
Proof.
35.
Proof.
36.
Proof.
37.
Proof.
38.
Proof.
39.
Proof.
40.
Proof.
41.
Proof.
42.
Proof.
43.
13
Proof.
44.
Proof.
45.
Proof.
46.
Proof.
47.
Proof.
48.
Proof.
49.
Proof.
50.
Proof.
51.
Proof.
52.
Proof.
53.
Proof.
54.
Proof.
14
55.
Proof.
56.
Proof.
57.
Proof.
58.
Proof.