35.3. WEAK DERIVATIVES IN Lploc 1237

Therefore, m(Em) = 0, a contradiction. Thus

m(E)≤∞

∑m=1

m(Em) = 0

and so, since ε > 0 is arbitrary,

m({x : f (x)> 0}) = 0.

Similarly m({x : f (x)< 0}) = 0. This proves the lemma.This lemma allows the following definition.

Definition 35.3.4 Let U be an open subset of Rn and let u∈ L1loc (U) . Then Dα u∈ L1

loc (U)if there exists a function g ∈ L1

loc (U), necessarily unique by Lemma 35.3.3, such that forall φ ∈C∞

c (U), ∫U

gφdx = Dα u(φ)≡∫

U(−1)|α| u(Dα

φ)dx.

Then Dα u is defined to equal g when this occurs.

Lemma 35.3.5 Let u ∈ L1loc (Rn) and suppose u,i ∈ L1

loc (Rn), where the subscript on the ufollowing the comma denotes the ith weak partial derivative. Then if φ ε is a mollifier anduε ≡ u∗φ ε , it follows uε,i ≡ u,i ∗φ ε .

Proof: If ψ ∈C∞c (Rn), then∫

u(x−y)ψ ,i (x)dx =∫

u(z)ψ ,i (z+y)dz

= −∫

u,i (z)ψ (z+y)dz

= −∫

u,i (x−y)ψ (x)dx.

Therefore,

uε,i (ψ) = −∫

uε ψ ,i =−∫ ∫

u(x−y)φ ε (y)ψ ,i (x)d ydx

= −∫ ∫

u(x−y)ψ ,i (x)φ ε (y)dxdy

=∫ ∫

u,i (x−y)ψ (x)φ ε (y)dxdy

=∫

u,i ∗φ ε (x)ψ (x)dx.

The technical questions about product measurability in the use of Fubini’s theorem may beresolved by picking a Borel measurable representative for u. This proves the lemma.

What about the product rule? Does it have some form in the context of weak deriva-tives?

35.3. WEAK DERIVATIVES IN Lh, ‘ 1237Therefore, m (Em) = 0, a contradiction. Thusm(E)< y m(Em) =0m=1and so, since € > 0 is arbitrary,m({x: f (x) > 0}) =0.Similarly m({x : f (x) < 0}) =0. This proves the lemma.This lemma allows the following definition.Definition 35.3.4 Let U be an open subset of R" and letu€ Li, (U). Then D“u€ Li, (U)loc locif there exists a function g € Lhe (U), necessarily unique by Lemma 35.3.3, such that forall@ €Ce(U),| gddx = D*u(o) = | (—1)!@! u(D%o) dx.U JuThen Du is defined to equal g when this occurs.Lemma 35.3.5 Let u € Lhe (IR") and suppose uj € Loe (IR”), where the subscript on the ufollowing the comma denotes the i‘" weak partial derivative. Then if - is a mollifier andUg =U*Q,, it follows Ug jj =Uj* Og.Proof: If y ¢ C? (R"), then[uc-yvie)ax = [u@)yjeryaz|/ui(z) W(z+y) dzui (X—y) W(x) dx.Therefore,wei(W) = —[uews=— | [uex—y) ely) vj (xa vax= =| [ulx—y)y,(x)o¢(y)ardy= | [use—y) ve) oe (y) day= Ju * - (x) W(x) dx.The technical questions about product measurability in the use of Fubini’s theorem may beresolved by picking a Borel measurable representative for u. This proves the lemma.What about the product rule? Does it have some form in the context of weak deriva-tives?