Folland: Real Analysis, Chapter 6

Sébastien Picard

Problem 6.8
Suppose µ(X) = 1R and f ∈ Lp for some p > 0, so that f ∈ Lq for 0 < q < p.
a. log
R ||fq||q ≥ log |f |. (Use Exercise
R 42d in §3.5, with F (t) = et ).
b. ( |f | − 1)/q ≥ logR||f ||q , and ( |f |q − 1)/q → log |f | as q → 0.
R

c. limq→0 ||f ||q = exp( log |f |).

Solution:
(a) Define g = q log |f |.

/ L1 . Then | log |f |q | = ∞. We can split up the integral

R
Case 1: Suppose g ∈
Z Z Z Z Z
| log |f | | = log |f | χ|f |>1 − log |f | χ|f |<1 ≤ |f | − log |f |q χ|f |<1 .
q q q q

q q
log |f |q χ|f |>1 is finite,
R R
R Since f ∈ L , we must have log |f | χ|f |<1 = −∞. Therefore, since
q log |f | = −∞ and the inequality holds trivially.

Case 2: Suppose g ∈ L1 . Let F (t) = et . We know that F (t) is convex on the real line since
the exponential function is absolutely continuous on every compact interval and F ′′ (t) = F (t) > 0.
Thus we can apply Jensen’s inequality:
1
Z Z
log |f | = log |f |q
q
1
Z
= g
q
1 R
= log e g
q
1
Z
≤ log eg
q
1
Z
= log |f |q = log ||f ||q .
q

(b)
The first inequality follows directly from the fact that log(x) ≤ x − 1.
1 1
Z Z
log ||f ||q = log |f | ≤ ( |f |q − 1).
q
q q

In order to prove the limit, we will need the fact that the map h : (0, ∞) → R defined by
h(x) = (ax − 1)/x is a monotone increasing function for any a ≥ 0. If a = 0, this is immediate. If
a > 0, we prove monotonicity by showing that h′ (x) ≥ 0. Indeed,

1
 ′ ax log ax − ax + 1
h (x) = ,
x2

hence the result will follow if we show that w : (0, ∞) → R, w(x) = x log x − x is greater than or
equal to −1. Since w ′ only vanishes at x = 1, and w ′′(1) > 0, we see that the minimum of w occurs
at x = 1 and hence w ≥ −1.

We now prove the limit. By l’Hopital’s rule, pointwise we have

|f |q − 1 log |f | · |f |q
lim = lim = log |f |.
q→0 q q→0 1

As shown before, (|f |q − 1)/q is monotone increasing, hence the limit as q → 0 is monotone
decreasing. By the Monotone Convergence Theorem, we have
1 1
Z Z Z
q q
lim (|f | − 1) = lim (|f | − 1) = log |f |.
q→0 q q→0 q

(c) From
R part (a), after composing both sides with the exponential function, we see that ||f ||q ≥
exp( log |f |) for all 0 < q < p. From part (b), we see that
Z Z
q


lim sup log ||f ||q ≤ lim sup |f | − 1 /q → log |f |.
q→0 q→0

R
Therefore,
R lim sup q→0 ||f || q ≤ exp( log |f |). Therefore, the limit exists, and limq→0 ||f ||p =
exp( log |f |).

Problem 6.10
Suppose 1 ≤ p < ∞. If fn , f ∈ Lp and fn → f a.e., then ||fn − f ||p → 0 iff ||fn ||p → ||f ||p. (Use
Exercise 20 in §2.3.)

Solution:
Suppose ||fn − f ||p → 0. Then by the reverse triangle inequality,

| ||fn ||p − ||f ||p | ≤ ||fn − f ||p → 0

as n → ∞.

On the other hand, suppose ||fn ||p → ||f ||p. Notice

|fn − f |p ≤ 2p (|f |p + |fn |p ).

We can define gn = 2p (|f |p + |fn |p ) and g = 2p+1|f |p ∈ L1 . Then gn → g a.e, and furthermore,

2
 Z Z Z Z Z
p p p p p+1 p


lim gn = 2 |f | + lim 2 |fn | = 2 |f | = g.

By Exercise 20 in §2.3,
Z Z
p
lim |fn − f | = lim |fn − f |p = 0.

Problem 6.20
Suppose supn ||fn ||p < ∞ and fn → f a.e.
a. If 1 < p < ∞, then fn → Rf weakly in Lp . (Given g ∈ Lq , where q is conjugate to p, and ǫ > 0,
there exists (i) δ > 0 such that E |g|q < ǫ whenever µ(E) < δ, (ii) A ⊂ X such that µ(A) < ∞ and
|g|q < ǫ, and (iii) B ⊂ A such that µ(A\B) < δ and fn → f uniformly on B.)
R
X\A
b. The result of (a) is false in general for p = 1. (Find counterexamples in L1 (R, m) and l1 .)
It is, however, true for p = ∞ if µ is σ−finite and weak convergence is replaced by weak* convergence.

Solution:
a. Claim (i) follows from Corollary 3.6, and claim (iii) is Egoroff’s theorem (2.33). We first prove
claim (ii). By Proposition 2.20, P = {x : |g|q > 0} is σ−finite. Hence we can assume that P = ∪∞ i Pi
where µ(Pi ) < ∞ and Pi are disjoint. Therefore,
Z Z ∞ Z
X
q q
|g| = |g| = |g|q < ∞.
X P i=1 Pi

P∞ R
Since the sequence converges, there exists an N ∈ N such that i=N Pi
|g|q < ǫ. Therefore,
A = ∪Ni=1 Pi has the required properties, which proves (ii).

We now prove that fn → f weakly in Lp . Take g ∈ Lq and any ǫ > 0. Since supn ||fn ||p ≤ C0 < ∞,
we have that f ∈ Lp by Fatou’s lemma:
Z Z
|f | ≤ lim inf |fn |p ≤ C0p < ∞.
p

By (i), choose δ > 0 such that

 q
ǫ
Z
q
|g| < ,
E 6C0

when µ(E) < δ. By (ii), choose A ⊂ X such that µ(A) < ∞ and
 q
ǫ
Z
q
|g| < .
X\A 6C0

By (iii), choose B ⊂ A and N ∈ N such that µ(A\B) < δ and

3
 ǫ
|fn (x) − f (x)| < ,
3µ(B)1/p ||g|| q

for all x ∈ B and n ≥ N.

With this setup, when n ≥ N, we apply Holder’s inequality to obtain

Z Z  Z Z Z
 
 fn g − f g ≤ |fn − f | |g| + |fn − f | |g| + |fn − f | |g|
 
X X X\A A\B B
Z 1/q Z 1/q
q q
≤ ||fn − f ||p |g| + ||fn − f ||p |g|
X\A A\B
Z 1/q Z 1/p
ǫ q p
+ |g| 1
3µ(B)1/p ||g||q B B
Z 1/q Z 1/q
q q ǫ
≤ 2C0 |g| + 2C0 |g| +
X\A A\B 3
< ǫ/3 + ǫ/3 + ǫ/3 = ǫ.

b. First, a counterexample in L1 (R, m). Consider

fn = n χ(0,1/n) .

We
R have ||fn ||1 R= 1 for all positive integers n, and fn → f a.e., where1 f = 0. However,
lim fn · 1 = 1 but f · 1 = 0. Therefore fn does not converge weakly to 0 in L .

Next, a counterexample in l1 . Consider


0 if i 6= n,
fn (i) =
1 if i = n.

R see that ||fn ||l1 =

We R 1 for all positive integers n, and fn → f a.e., where f = 0. But
lim fn · 1 dµ = 1 and f · 1 dµ = 0, where µ is the counting measure. Therefore fn does not
converge to 0 weakly in l1 .

Now we show that if supn ||fn ||∞ = C0 < ∞, fn → f a.e., and µ is σ−finite, then fn → f
in the weak* sense on L∞ . The first thing we show is that ||f ||∞ < ∞. Indeed, for a.e. x, there exists
a positive integer n such that |fn (x) − f (x)| < 1. Then

|f (x)| ≤ |f (x) − fn (x)| + |fn (x)| ≤ 1 + C0 .

Hence ||f ||∞ ≤ 1 + C0 .

4
 By the σ−finite condition, L∞ = (L1 )∗ . Let g ∈ L1 and ǫ > 0.

By fact (i), choose δ > 0 such that

ǫ
Z
|g| < ,
E 3(2C0 + 1)

when µ(E) < δ. By fact (ii), choose A ⊂ X such that µ(A) < ∞ and
ǫ
Z
|g| < .
X\A 3(2C0 + 1)

By fact (iii), choose B ⊂ A and N ∈ N such that µ(A\B) < δ and

ǫ
|fn (x) − f (x)| < ,
3||g||1

for all x ∈ B and n ≥ N. With this setup, when n ≥ N we have

Z Z  Z Z Z
 
 fn g − f g  ≤ |fn − f | |g| + |fn − f | |g| + |fn − f | |g|

X X X\A A\B B
ǫ
Z Z Z
≤ ||fn − f ||∞ |g| + ||fn − f ||∞ |g| + |g|
X\A A\B 3||g||1 B
ǫ ǫ ǫ
< (2C0 + 1) + (2C0 + 1) + ||g||1 = ǫ.
3(2C0 + 1) 3(2C0 + 1) 3||g||1

Problem 6.22
Let X = [0, 1], with Lebesgue measure.
a. Let fn (x) = cos 2πnx. Then fn → 0 weakly in L2 (see Exercise 63 in §5.5), but fn 9 0 a.e. or
in measure.
b. Let fn (x) = nχ(0,1/n) . Then fn → 0 a.e. and in measure, but fn 9 0 weakly in Lp for any p.

Solution: √
a. Let un (x) = 2 cos 2πnx. We note the following two identities, where n 6= m.
Z 1 Z 1 Z 1
2 2 1 1
un = 2 (cos 2πnx) dx = 2 ( + cos 4πnx)dx = 1.
0 0 0 2 2

Z 1 Z 1 Z 1
un um = 2 (cos 2πnx)(cos 2πmx)dx = (cos(2πx(n − m)) + cos(2πx(n + m)))dx = 0.
0 0 0

Therefore {un } is orthonormal

√ in the Hilbert space L2 . By Exercise 63(a), {un } converges weakly
to 0 in L2 . Therefore, {(1/ 2)un } = {fn } converges weakly to 0 in L2 .

5
 We show that fn 9 0 a.e. by contradiction. If fn → 0 a.e., then (fn · fn ) → 0 a.e. Since
fn2≤ 1 is dominated by an integrable function on [0, 1], by the Dominated Convergence Theorem we
have
Z Z
lim (cos 2πnx) dx = lim(cos 2πnx)2 dx = 0.
2

R1
This is clearly nonsense since 0
cos(2πnx) = 1/2.

Finally, we show fn 9 0 in measure. We explicitely calculate the measure of the set of all
x ∈ [0, 1] such that | cos 2πnx| ≥ 1/2. Denote

π
3π
5π
7π

Ek = 2πk, + 2πk ∪ + 2πk, π + 2πk ∪ π + 2πk, + 2πk ∪ + 2πk, 2π(k + 1) .
4 4 4 4

Then | cos 2πnx| ≥ 1/2 iff 2πnx ∈ Ek for some k ∈ N. Therefore, using the 2π periodicity of
cosine, we see that | cos 2πnx| ≥ 1/2 when x ∈ [0, 1] lies in one of the 4n disjoint intervals of length
1/8n. Hence
1
{x ∈ [0, 1] : | cos 2πnx| ≥ 1/2} = .
2

Since the measure is constant at 1/2 for all n, fn 9 0 in measure.

(b) First we show that fn (x) = nχ(0,1/n) → 0 a.e. Fix x ∈ (0, 1]. Then fn (x) = 0 for all n > 1/x.

Next, we show that fn (x) = nχ(0,1/n) → 0 in measure. For all ǫ > 0, we have

µ{x ∈ [0, 1] : |fn (x)| ≥ ǫ} ≤ µ(0, 1/n) = 1/n → 0

as n → ∞. Lastly, we show fn 9 0 weakly in Lp for any p. Since 1 ∈ Lq ([0, 1]), let φ be the bounded
R1
linear functional on Lp defined by φ(f ) = 0 f . Then
Z 1
1
lim φ(fn ) = lim fn (x)dx = lim n = 1.
0 n

Therefore, fn does not converge weakly to zero in Lp .

Problem 6.27 (Hilbert’s R ∞Inequality)

−1
The operator
R ∞ −1/p T f (x) −1= 0 (x + y) f (y)dy satisfies ||T f ||p ≤ Cp ||f ||p for 1 < p < ∞, where
Cp = 0 x (x+ 1) dx. (For those who know about contour integrals: Show that Cp = π csc(π/p).)

Solution:

6
 Define a function K on (0, ∞) × (0, ∞) by K(x, y) = (x + y)−1 . Then K is a Lebesgue measurable
function such that K(λx, λy) = λ−1 K(x, y) for all λ > 0. In order to apply Theorem 6.20, we must
R∞ R1
verify that 0 |K(x, R1)|x−1/p dx is finite. This follows from p > 1, and the facts that 0 x−α < ∞
∞
when 0 < α < 1 and 1 x−β < ∞ when 1 < β.

∞ 1 ∞ 1 ∞
1 1 1 1
Z Z Z Z Z
−1/p
|K(x, 1)|x dx = + ≤ + < ∞.
0 0 x1/p (x + 1) 1 x1/p (x + 1) 0 x1/p 1 x1/p x

Hence the conditions of Theorem 6.20 are satisfied, and the desired inequality follows directly from
the theorem.

Problem 6.29
Suppose that 1 ≤ p < ∞, r > 0, and h is a nonnegative measurable function on (0, ∞). Then:
Z ∞ Z x p  p Z ∞
−r−1 p
x h(y)dy dx ≤ xp−r−1 h(x)p dx,
0 0 r 0
Z ∞ Z ∞ p  p Z ∞
p
xr−1 h(y)dy dx ≤ xp+r−1 h(x)p dx.
0 x r 0

(Apply Theorem 6.20 with K(x, y) = xβ−1 y −β χ(0,∞) (y − x), f (x) = xγ h(x), and g(x) = xδ h(x) for
suitable β, γ, δ.)

Solution:
We start by showing the first inequality. Consider K(x, y) = xβ−1 y −β χ(0,∞) (y − x) and f (x) =
xγ h(x), where β = (r + 1)/p and γ = 1 − (r + 1)/p. If f ∈ / Lp (0, ∞), the right-hand side of the
inequality is infinite and the inequality holds trivially. Hence we assume that f ∈ Lp (0, ∞). It is clear
that K is a Lebesgue measurable function on (0, ∞) × (0, ∞) such that K(λx, λy) = λ−1 K(x, y) for
all λ > 0. Using the notation from Theorem 6.20, we evaluate C:
Z ∞ Z ∞ Z 1
−1/p β−1 −1/p p
C= |K(x, 1)|x dx = x χ(0,∞) (1 − x)x dx = x−1+(r/p) dx = .
0 0 0 r

By Theorem 6.20, ||T f ||p ≤ C||f ||p, where

Z ∞ Z y
β−1 −β γ −(r+1)/p
Tf = x y χ(0,∞) (y − x)x h(x)dx = y h(x)dx.
0 0

Then we have the following inequalities:

 −(r+1)/p y
Z
p p
h(x)dx p ≤ C p  x1−(r+1)/p h(x) p .

 y
0
∞ Z y p  p Z ∞
p
Z
−(r+1)
y h(x)dx dy ≤ xp−(r+1) h(x)p dx.
0 0 r 0

7
 To show the second inequality, we reapply the theorem with different functions K̃ and f˜. Consider
K̃(x, y) = xα−1 y −αχ(0,∞) (x − y) and f˜(x) = xδ h(x), where α = (1 − r)/p and δ = 1 + (r − 1)/p. If
f˜ ∈
/ Lp (0, ∞), the right-hand side of the inequality is infinite and the inequality holds trivially. Hence
we assume that f˜ ∈ Lp (0, ∞).

We evaluate C:
Z ∞ ∞ ∞
p
Z Z
−1/p α−1 −1/p
C= |K̃(x, 1)|x dx = x χ(0,∞) (x − 1)x dx = x−1−(r/p) dx = .
0 0 1 r

By Theorem 6.20, ||T f˜||p ≤ C||f˜||p , where

Z ∞ Z ∞
T f˜ = α−1 −α δ
x y χ(0,∞) (x − y)x h(x)dx = y (−1+r)/p
h(x)dx.
0 y

Then we have the following inequalities:

 (−1+r)/p ∞
Z
p p
h(x)dx p ≤ C p  x1+(r−1)/p h(x) p .

 y
y
∞ Z ∞ p  p Z ∞
p
Z
r−1
y h(x)dx dy ≤ xp+r−1 h(x)p dx.
0 y r 0

Problem 6.31 (A Generalized Pn Holder’s Inequality) Qn

−1 −1 pj r
Suppose
Qn that 1
Qn≤ p j ≤ ∞ and 1 p j = r = 1. If fj ∈ L for j = 1, . . . , n, then 1 fj ∈ L and
|| 1 fj ||r ≤ 1 ||fj ||pj . (First do the case n=2.)

Solution:
We do the case n = 2 and then proceed by induction. When n = 2, we apply Holder’s inequality
with conjugate exponents p1 /r and p2 /r.
Z
||f1 f2 ||r = |f1 f2 |r ≤ || |f1 |r ||p1 /r || |f2 |r ||p2 /r = ||f1 ||rp1 ||f2||rp2 .
r

The result follows by taking rth roots of both sides.

Suppose thePresult is true for (n − 1). First we Holder’s inequality with conjugate exponents
pn /r and (1/r)( 1n−1 p−1 −1
j ) .

n Z n−1
 Y r r
Y
 fj r = |fn | |fj |r
1 j=1

r 
 n−1
Y
|fj |r (1/r)(Pn−1 p−1 )−1
  
≤ |fn | pn /r 

1 j
j=1

 r  n−1
Y r
= fn   fj  Pn−1
pn ( p−1
.
j )
−1
1
j=1

8
 We can now take rth roots of both sides, and apply the induction hypothesis with r ′ such that
n−1
(r ′ )−1 = 1 p−1
P
j .

n
 Y     n−1
Y  n
Y
 fj r ≤ fn pn
      fj r′ ≤
 ||fj ||pj .
1 j=1 j=1

Problem P6.38
f ∈ L iff ∞
p kp k
−∞ 2 λf (2 ) < ∞.

Solution:
Suppose f ∈ Lp . Using that αf is decreasing, and Theorem 2.15 to exchange an infinite sum and
an integral, we compute the following:
∞ ∞ Z 2k
X
k kp
X
k p
λf (2 )2 = λf (2 ) −p
αp−1 dα
−∞ −∞
1−2 2k−1
∞ Z 2k
X p
≤ −p
λf (α)αp−1dα
−∞
1−2 2k−1
Z ∞
p
= −p
λf (α)αp−1dα
1−2
Z0
1
= −p
|f |p < ∞
1−2

P∞
where the last step follows from Proposition 6.24. On the other hand, if −∞ 2kp λf (2k ) < ∞, we
have the following:
Z Z ∞
p
|f | = p λf (α)αp−1dα
0
∞ Z
X 2k+1
=p λf (α)αp−1dα
−∞ 2k
∞
X Z 2k+1
k
≤p λf (2 ) αp−1 dα
−∞ 2k
∞
X
p
= (2 − 1) λf (2k )2kp < ∞
−∞

Problem 6.39
If f ∈ Lp , then limα→0 αp λf (α) = limα→∞ αp λf (α) = 0. (First suppose f is simple.)

Solution:
First, suppose f = n1 aj χEj , where the Ej are disjoint P
P
and µ(Ej ) < ∞ for all j. Let amax =
max{aj } and amin = min{aj }. Then when α < amin , λf (α) = µ(Ej ) < ∞. Hence

9
 X
lim αp λf (α) = lim αp


µ(Ej ) = 0.
α→0 α→0

On the other hand, when α > amax , then λf (α) = 0. Therefore, limα→∞ αp λf (α) = 0.

We now take an arbitrary f ∈ Lp . We show that there exists a simple function φ = n1 aj χEj ,
P
where Ej are disjoint and µ(Ej ) < ∞ for all j, such that ||f −φ||pp < ǫ for every ǫ > 0. By Theorem 2.10,
there is a sequence {φ
Pnn} of simple functions such that φn → f a.e. Pand |φn | ≤ |f |. RThen every φn ∈ Lp ,
n
and assuming φn = 1 aj χEj where the Ej are disjoint, we see 1 |aj |p µ(Ej ) = |φn |p ≤ ||f ||pp < ∞.
Hence µ(Ej ) < ∞ for all j. Since

|f − φn |p ≤ 2p (|f |p + |φn |p ) ≤ 2p+1 |f |p,

we have that |f − φn |p is dominated by an L1 function, hence by dominated convergence,

Z Z
lim |f − φn | = lim |f − φn |p = 0.
p

Therefore, we can choose φ = n1 aj χEj , where µ(Ej ) < ∞ for all j, such that ||f − φ||pp < ǫ/2p+1 .
P
By the above result, for all α small enough, we have αp λφ (α) < ǫ/2p+1. Now, we can write
f = (f − φ) + φ and apply Proposition 6.22(d) and Chebyshev’s Inequality:
1 1
lim sup αp λf (α) ≤ lim sup { αp λf −φ ( α) + αp λφ ( α) }
α→0+ α→0+ 2 2
p p p p
≤ lim sup { 2 ||f − φ||p + 2 α λφ (α) }
α→0+
< ǫ/2 + ǫ/2 = ǫ.

Therefore, limα→0 αp λf (α) = 0. Similarly, for all α large enough, we have αp λφ (α) < ǫ/2p+1 . The
same calculation shows:
1 1
lim sup αp λf (α) ≤ lim sup { αp λf −φ ( α) + αp λφ ( α) }
α→∞ α→∞ 2 2
p p p p
≤ lim sup { 2 ||f − φ||p + 2 α λφ (α) }
α→∞
< ǫ/2 + ǫ/2 = ǫ.

Additional Problem from Class

Let f : X → C be measurable, and let E = {p : X |f |p = ||f ||pp < ∞}. It follows from a result in
R

Folland that the set E is connected.

i) Prove that the function φ(p) = log(||f ||pp) is convex in the interior of E, and that φ is continuous
on E.
ii) Can E consist of a single point? Can it be any connected subset of (0, ∞)?

10
 iii) Let ||f ||r < ∞ for some r < ∞. Prove that limr→∞ ||f ||r = ||f ||∞ .

Solution:
(i) Let t ∈ (0, 1), and p, q ∈ E. By Holder’s Inequality, we know that
Z
|f |pt|f |q(1−t) ≤ || |f |pt ||1/t || |f |q(1−t) ||1/(1−t) .

Therefore, we can take the logarithm and obtain

Z
|f |tp+(1−t)q


φ(tp + (1 − t)q) = log
 Z Z 
p t q 1−t



≤ log |f | |f |
Z Z
p
|f |q



= t log |f | + (1 − t) log

= tφ(p) + (1 − t)φ(q).

Next, we show continuity of φ by showing sequential continuity. Suppose {qn } is a sequence in E

converging to a point p ∈ E. If p is in the interior of E, there exists a δ > 0 such that (p−δ, p+δ) ⊂ E
and qn ∈ (p − δ, p + δ) for all n > N for some N ∈ N. Then we have for all n > N:

|f |qn = |f |qn χ|f |≥1 + |f |qn χ|f |≤1 ≤ |f |p+δ χ|f |≥1 + |f |p−δ χ|f |≤1 ≤ |f |p+δ + |f |p−δ .

By the Dominated Convergence Theorem, lim |f |qn = |f |p . Therefore, by continuity of the

R R

logarithm, lim φ(qn ) = φ(p) and φ is continuous in the interior of E.

The case when p is at an endpoint of E is similar. First we do the case when qn ≤ p (ie p is
a right endpoint of E). Then there exists a δ > 0 such that (p − δ, p] ⊂ E and qn ∈ (p − δ, p] for all
n > N for some N ∈ N. Then we have for all n > N:

|f |qn = |f |qn χ|f |≥1 + |f |qn χ|f |≤1 ≤ |f |p χ|f |≥1 + |f |p−δ χ|f |≤1 ≤ |f |p + |f |p−δ .

By the Dominated Convergence Theorem, lim φ(qn ) = φ(p).

When p is a left endpoint of E, then qn ≥ p, and there exists a δ > 0 such that [p, p + δ) ⊂ E and
qn ∈ [p, p + δ) for all n > N for some N ∈ N. Then we have for all n > N:

|f |qn = |f |qn χ|f |≥1 + |f |qn χ|f |≤1 ≤ |f |p+δ χ|f |≥1 + |f |p χ|f |≤1 ≤ |f |p+δ + |f |p .

By the Dominated Convergence Theorem, lim φ(qn ) = φ(p).

(ii) We construct a function f : (0, ∞) → C such that E = {1}. We first use the fact that for α > 0,
we have

11
 ∞
1
Z
χ(0,1) < ∞
0 xα

if and only if α < 1. We then consider the function

∞
X an 1
f= n (1+1/n)−1 χ(0,1) ,
n=1
2 x

where
 1 
an =  χ(0,1) 1 .
x(1+1/n)−1

Then f is a series of functions in L1 that converges absolutely, so by completeness, f ∈ L1 . How-

/ Lp for any p > 1.
ever, f ∈

Next, we use the fact that for β > 0, we have

Z ∞
1
χ(1,∞) < ∞
0 xβ

if and only if β > 1. We then consider the function

∞
X bn 1
g= n x(1+1/n)
χ(1,∞) ,
n=1
2

where
 1 
bn =  χ(1,∞) 1 .
x(1+1/n)

Then g is a series of functions in L1 that converges absolutely, so by completeness, g ∈ L1 . How-

/ Lp for any p < 1.
ever, g ∈

Therefore, the function f + g is in L1 , but not in Lp for any p ∈ (0, 1) ∪ (1, ∞).

We now show how to get any interval E = (a, b), again using X = (0, ∞). Let f (x) = x−1/b
when 0 < x < 1 and identically zero otherwise. Let g(x) = x−1/a when x > 1 and identically zero
otherwise. Then f + g is such that E = (a, b).

12
(iii) First, suppose f ∈ L∞ . Then
Z

1/q
lim ||f ||q = lim |f |q−r |f |r
q→∞ q→∞
Z
q−r

1/q
≤ lim ||f ||∞ |f |r
q→∞
Z
(q−r)/q

1/q
= lim ||f ||∞ |f |r
q→∞

= ||f ||∞.

Therefore, if the limit exists we have limq→∞ ||f ||q ≤ ||f ||∞. Now, for all ǫ > 0, there is a set
E ⊂ X of positive measure such that |f (x)| ≥ ||f ||∞ − ǫ for all x ∈ E. Then when q > r we have
Z

1/q
||f ||q ≥ |f |q ≥ (||f ||∞ − ǫ)(µ(E))1/q .
E

Now µ(E) is finite, so taking the limit as q → ∞ we get limq→∞ ||f ||q ≥ ||f ||∞ − ǫ. Since this
holds for all ǫ > 0, we conclude that limq→∞ ||f ||q = ||f ||∞.

Next, suppose ||f ||∞ = ∞. I have a marvelous proof of this case, which this margin is too
narrow to contain.

13