13.2. FOURIER TRANSFORMS OF JUST ABOUT ANYTHING 379

Definition 13.2.3 For T ∈ G ∗, define FT,F−1T ∈ G ∗ by

FT (φ)≡ T (Fφ) , F−1T (φ)≡ T(F−1

φ)

Lemma 13.2.4 F and F−1 are both one to one, onto, and are inverses of each other.

Proof: First note F and F−1 are both linear. This follows directly from the definition.Suppose now FT = 0. Then FT (φ) = T (Fφ) = 0 for all φ ∈ G . But F and F−1 map Gonto G because if ψ ∈ G , then as shown above, ψ = F

(F−1 (ψ)

). Therefore, T = 0 and

so F is one to one. Similarly F−1 is one to one. Now

F−1 (FT )(φ)≡ (FT )(F−1

φ)≡ T

(F(F−1 (φ)

))= T φ .

Therefore, F−1 ◦F (T ) = T. Similarly, F ◦F−1 (T ) = T. Thus both F and F−1 are one toone and onto and are inverses of each other as suggested by the notation. ■

Probably the most interesting things in G ∗ are functions of various kinds. The followinglemma will be useful in considering this situation.

Lemma 13.2.5 If f ∈ L1loc (Rn) and

∫Rn f φdx = 0 for all φ ∈Cc (Rn), then f = 0 a.e.

Proof: Let E be bounded and Lebesgue measurable. By regularity, there exists a com-pact set Kk ⊆ E and an open set Vk ⊇ E such that mn (Vk \Kk)< 2−k. Let hk equal 1 on Kk,vanish on VC

k , and take values between 0 and 1. Let Ek ≡[∣∣hk−XEk

∣∣> ( 23

)k]. Then

mn (Ek)≤(

32

)k ∫Ek

∣∣hk−XEk

∣∣dmn <

(32

)k ∫Vk\Kk

2dmn = 2(

32

)k 12k = 2

(34

)k

and so ∑k mn (Ek)< ∞ so there is a set of measure zero such that off this set, hk→XEk .Hence, by the dominated convergence theorem,

∫f XEdmn = limk→∞

∫f hkdmn = 0. It

follows that for E an arbitrary Lebesgue measurable set,∫

f XB(0,R)XEdmn = 0. Let

sgn f =

{f| f | if | f | ̸= 00 if | f |= 0

By Theorem 9.1.6 applied to positive and negative parts, there exists {sk}, a sequence ofsimple functions converging pointwise to sgn f such that |sk| ≤ 1. Then by the dominatedconvergence theorem again,

∫| f |XB(0,R)dmn = limk→∞

∫f XB(0,R)skdmn = 0. Since R is

arbitrary, | f |= 0 a.e. ■

Corollary 13.2.6 Let f ∈ L1 (Rn) and suppose∫Rn f (x)φ (x)dx = 0 for all φ ∈ G .

Then f = 0 a.e.

Proof: Let ψ ∈Cc (Rn) . Then by the Stone Weierstrass approximation theorem, thereexists a sequence of functions, {φ k} ⊆ G such that φ k→ ψ uniformly. Then by the domi-nated convergence theorem,

∫f ψdx = limk→∞

∫f φ kdx = 0. By Lemma 13.2.5 f = 0. ■

The next theorem is the main result of this sort.

Theorem 13.2.7 Let f ∈ Lp (Rn) , p≥ 1, or suppose f is measurable and has poly-

nomial growth, | f (x)| ≤K(

1+ |x|2)m

for some m∈N. Then if∫

f ψdx= 0, for all ψ ∈ G ,then it follows f = 0.

13.2. FOURIER TRANSFORMS OF JUST ABOUT ANYTHING 379Definition 13.2.3 For T € Y*, define FT,F-'T € G* byFT (¢) =T (Fo), F-'T(¢) =T (F'@)Lemma 13.2.4 F and F~! are both one to one, onto, and are inverses of each other.Proof: First note F and F~! are both linear. This follows directly from the definition.Suppose now FT = 0. Then FT (@) = T (F@) = 0 for all 6 € Y. But F and F~! map Yonto Y because if yw € Y, then as shown above, yw = F (F 1 (w)) . Therefore, T = 0 andso F is one to one. Similarly F~! is one to one. NowF' (FT) (9) = (FT) (F-'6) =T (F (F' (@))) =T9.Therefore, F~! oF (T) = T. Similarly, Fo F~!(T) =T. Thus both F and F~! are one toone and onto and are inverses of each other as suggested by the notation.Probably the most interesting things in Y* are functions of various kinds. The followinglemma will be useful in considering this situation.Lemma 13.2.5 [f f € L},.(R") and fn fodx =0 for all @ € C,(R"), then f =O ae.locProof: Let E be bounded and Lebesgue measurable. By regularity, there exists a com-pact set K, C E and an open set Vk > E such that my (Vz \ Kx) < 2~*. Let ty equal 1 on Kx,vanish on VC, and take values between 0 and 1. Let Ey = |e — &p,| > (3)'| . Then3\* 3\* 3\* 1 3\*<[= _— _ = = —_ = =mn (Ex) < (5) [Ve Xp, | din < (5) [2am 2(5) 5k 2(7)and so ), my (Ex) <° so there is a set of measure zero such that off this set, hy — 2,.Hence, by the dominated convergence theorem, { f 2gdmy, = limy... f fhydmy, = 0. Itfollows that for E an arbitrary Lebesgue measurable set, { f 2g(0,n) 2zdimn = 0. Letif [FAO=? ff!sens Oif |f)/ =0By Theorem 9.1.6 applied to positive and negative parts, there exists {s,}, a sequence ofsimple functions converging pointwise to sgn f such that |s,;| < 1. Then by the dominatedconvergence theorem again, {|| -23(0,r) din = lim J f 2B(0,n)5kdIMn = 0. Since R isarbitrary, |f|=0a.e.Corollary 13.2.6 Let f € L'(R") and suppose rn f (x) (x) dx = 0 for all 9 EY.Then f =O a.e.Proof: Let yw € C, (R"). Then by the Stone Weierstrass approximation theorem, thereexists a sequence of functions, {¢,} C Y such that @, > y uniformly. Then by the domi-nated convergence theorem, f fydx = limg. f fo,dx = 0. By Lemma 13.2.5 f = 0.The next theorem is the main result of this sort.Theorem 13.2.7 Let f € LP (R"), p> 1, or suppose f is measurable and has poly-mnomial growth, |f (a)| <K (1 + je|*) for some m EN. Then if [ fwdx =0, forall y EY,then it follows f = 0.