6 STEVE SHKOLLER
1.6. Approximation of Lp(X) by simple functions.
Lemma 1.23. If p ∈ [1, ∞), then the set of simple functions f =n
i=1ai1Ei, where each Ei is an element of the σ-algebraA and μ(Ei) <∞, is dense in Lp(X,A, μ).
Proof. If f ∈ Lp, then f is measurable; thus, there exists a sequence{φn}∞n=1 of simple functions, such that φn→ f a.e. with
0≤ |φ1| ≤ |φ2| ≤ · · · ≤ |f|, i.e., φn approximates f from below.
Recall that|φn− f|p→ 0 a.e. and |φn− f|p≤ 2p|f|p∈ L1, so by the Dominated Convergence Theorem,φn− fLp → 0.
Now, suppose that the set Ei are disjoint; then b definition of the Lebesgue integral,
X
φpndx =
n i=1
|ai|pμ(Ei) <∞ .
If ai= 0, then μ(Ei) <∞.
1.7. Approximation of Lp(Ω)by continuous functions.
Lemma 1.24. Suppose that Ω ⊂ Rn is bounded. Then C0(Ω) is dense in Lp(Ω) for p∈ [1, ∞).
Proof. Let K be any compact subset of Ω. The functions FK,n(x) = 1
1 + n dist(x, K) ∈ C0(Ω) satisfy FK,n≤ 1 ,
and decrease monotonically to the characteristic function1K. The Monotone Con- vergence Theorem gives
fK,n→ 1K in Lp(Ω), 1≤ p < ∞ .
Next, let A⊂ Ω be any measurable set, and let λ denote the Lebesgue measure.
Then
λ(A) = sup{μ(K) : K ⊂ A, K compact} .
It follows that there exists an increasing sequence of Kj of compact subsets of A such that λ(A\ ∪jKj) = 0. By the Monotone Convergence Theorem,1Kj → 1Ain Lp(Ω) for p∈ [1, ∞). According to Lemma 1.23, each function in Lp(Ω) is a norm limit of simple functions, so the lemma is proved. 1.8. Approximation of Lp(Ω) by smooth functions. For Ω ⊂ Rn open, for
> 0 taken sufficiently small, define the open subset of Ω by Ω:={x ∈ Ω | dist(x, ∂Ω) > } . Definition 1.25 (Mollifiers). Define η ∈ C∞(Rn) by
η(x) :=
Ce(|x|2−1)−1 if |x| < 1
0 if |x| ≥ 1 ,
with constant C > 0 chosen such that
Rnη(x)dx = 1.
For > 0, the standard sequence of mollifiers on Rn is defined by η(x) = −nη(x/) ,
and satisfy
Rnη(x)dx = 1 and spt(η)⊂ B(0, ).
NOTES ON Lp AND SOBOLEV SPACES 7
Definition 1.26. For Ω ⊂ Rn open, set
Lploc(Ω) ={u : Ω → R | u ∈ Lp( ˜Ω) ∀ ˜Ω ⊂⊂ Ω} ,
where ˜Ω⊂⊂ Ω means that there exists K compact such that ˜Ω ⊂ K ⊂ Ω. We say that ˜Ω is compactly contained in Ω.
Definition 1.27 (Mollification of L1). If f ∈ L1loc(Ω), define its mollification f= η∗ f in Ω,
so that
f(x) =
Ω
η(x− y)f(y)dy =
B(0,)
η(y)f (x− y)dy ∀x ∈ Ω. Theorem 1.28 (Mollification of Lp(Ω)).
(A) f∈ C∞(Ω).
(B) f→ f a.e. as → 0.
(C) If f ∈ C0(Ω), then f→ f uniformly on compact subsets of Ω.
(D) If p∈ [1, ∞) and f ∈ Lploc(Ω), then f→ f in Lploc(Ω).
Proof. Part (A). We rely on the difference quotient approximation of the partial derivative. Fix x ∈ Ω, and choose h sufficiently small so that x + hei ∈ Ω for i = 1, ..., n, and compute the difference quotient of f:
f(x + hei)− f(x)
= −n
Ω
1 h
η
x + hei− y
− η
x− y h
f (y)dy
= −n
Ω˜
1 h
η
x + hei− y
− η
x− y
f (y)dy for some open set ˜Ω⊂⊂ Ω. On ˜Ω,
h→0lim 1 h
η
x + hei− y
− η
x− y
= 1
∂η
∂xi
x− y
= n∂η
∂xi
x− y
, so by the Dominated Convergence Theorem,
∂f
∂xi(x) =
Ω
∂η
∂xi(x− y)f(y)dy .
A similar argument for higher-order partial derivatives proves (A).
Step 2. Part (B). By the Lebesgue differentiation theorem,
→0lim 1
|B(x, )|
B(x,)|f(y) − f(x)|dy for a.e. x ∈ Ω . Choose x∈ Ω for which this limit holds. Then
|f(x)− f(x)| ≤
B(x,)η(x− y)|f(y) − f(x)|dy
= 1
n
B(x,)
η((x− y)/)|f(y) − f(x)|dy
≤ C
|B(x, )|
B(x,)|f(x) − f(y)|dy −→ 0 as → 0 .
8 STEVE SHKOLLER
Step 3. Part (C). For ˜Ω ⊂ Ω, the above inequality shows that if f ∈ C0(Ω) and hence uniformly continuous on ˜Ω, then f(x)→ f(x) uniformly on ˜Ω.
Step 4. Part (D). For f ∈ Lploc(Ω), p∈ [1, ∞), choose open sets U ⊂⊂ D ⊂⊂ Ω;
then, for > 0 small enough,
fLp(U)≤ fLp(D). To see this, note that
|f(x)| ≤
B(x,)
η(x− y)|f(y)|dy
=
B(x,)η(x− y)(p−1)/pη(x− y)1/p|f(y)|dy
≤
B(x,)η(x− y)dy
(p−1)/p
B(x,)η(x− y)|f(y)|pdy 1/p
, so that for > 0 sufficiently small
U|f(x)|pdx≤
U
B(x,)η(x− y)|f(y)|pdydx
≤
D|f(y)|p
B(y,)η(x− y)dx
dy≤
D|f(y)|pdy . Since C0(D) is dense in Lp(D), choose g ∈ C0(D) such thatf − gLp(D) < δ;
thus
f− fLp(U)≤ f− gLp(U)+g− gLp(U)+g − fLp(U)
≤ 2f − gLp(D)+g− gLp(U)≤ 2δ + g− gLp(U).
1.9. Continuous linear functionals on Lp(X). Let Lp(X) denote the dual space of Lp(X). For φ ∈ Lp(X), the operator norm of φ is defined byφop = supLp(X)=1|φ(f)|.
Theorem 1.29. Let p ∈ (1, ∞], q = p−1p . For g∈ Lq(X), define Fg: Lp(X)→ R as
Fg(f ) =
X
f gdx .
Then Fg is a continuous linear functional on Lp(X) with operator normFgop=
gLp(X).
Proof. The linearity of Fgagain follows from the linearity of the Lebesgue integral.
Since
|Fg(f )| =
Xf gdx ≤
X|fg| dx ≤ fLpgLq ,
with the last inequality following from H¨older’s inequality, we have that supf
Lp=1|Fg(f )| ≤
gLq.