TheoremBase

Linearity is the bilinearity of the inner product; injectivity is the completeness of the trigonometric system; Parseval, the expansion and the Riesz-Fischer statement are the orthonormal expansion theorem applied to the trigonometric orthonormal basis along the enumeration; and the realisation clause follows from injectivity together with the invertibility of the weights.

Proof

Each result cited is universally quantified over the data in its own statement. Two coefficient families are equal exactly when their values agree at every k∈Znk\in\mathbb{Z}^{n}, and the pointwise operations of Map(Zn,R)\mathrm{Map}(\mathbb{Z}^{n},\mathbb{R}) are those of The Real Vector Space of Real-Valued Functions on a Set (i). The inner-product norm of L2(Tn)L^{2}(\mathbb{T}^{n}) is ∥ ⋅ ∥L2\lVert\,\cdot\,\rVert_{L^{2}} by L²(μ) Is a Real Hilbert Space (i), and differences in it are those of Elementary Identities in a Vector Space. The inner product identities of Elementary Identities in a Real Inner Product Space (i) and the conditions of Real Inner Product Spaces (i) are used for ⟨ ⋅ ,⋅ ⟩L2\langle\,\cdot\,,\cdot\,\rangle_{L^{2}}. For an enumeration κ\kappa, the sequence (Eκ(j))j∈N(E_{\kappa(j)})_{j\in\mathbb{N}} is an orthonormal basis of L2(Tn)L^{2}(\mathbb{T}^{n}) by The Trigonometric System is an Orthonormal Basis of the Square-Integrable Space of the Torus (ii), so that Orthonormal Expansions in a Real Hilbert Space, in the setting of Real Hilbert Spaces: Standing Notation and Background, applies to H=L2(Tn)H=L^{2}(\mathbb{T}^{n}) with ej=Eκ(j)e_{j}=E_{\kappa(j)}.

Claim 1. Let U,U′∈L2(Tn)U,U'\in L^{2}(\mathbb{T}^{n}), t∈Rt\in\mathbb{R} and k∈Znk\in\mathbb{Z}^{n}. By condition (b) of Real Inner Product Spaces (i), U+U′^(k)=⟨U+U′,Ek⟩L2=⟨U,Ek⟩L2+⟨U′,Ek⟩L2=U^(k)+U^′(k)=(U^+U^′)(k)\widehat{U+U'}(k)=\langle U+U',E_{k}\rangle_{L^{2}}=\langle U,E_{k}\rangle_{L^{2}}+\langle U',E_{k}\rangle_{L^{2}}=\hat{U}(k)+\hat{U}'(k)=(\hat{U}+\hat{U}')(k), and by condition (c), tU^(k)=t⟨U,Ek⟩L2=(tU^)(k)\widehat{tU}(k)=t\langle U,E_{k}\rangle_{L^{2}}=(t\hat{U})(k). As kk was arbitrary, U+U′^=U^+U^′\widehat{U+U'}=\hat{U}+\hat{U}' and tU^=tU^\widehat{tU}=t\hat{U}, which are the two conditions of Linear Map.

Claim 2. Let U^=U^′\hat{U}=\hat{U}'. For every k∈Znk\in\mathbb{Z}^{n}, Elementary Identities in a Real Inner Product Space (i) gives ⟨U−U′,Ek⟩L2=⟨U,Ek⟩L2−⟨U′,Ek⟩L2=U^(k)−U^′(k)=0\langle U-U',E_{k}\rangle_{L^{2}}=\langle U,E_{k}\rangle_{L^{2}}-\langle U',E_{k}\rangle_{L^{2}}=\hat{U}(k)-\hat{U}'(k)=0. By The Trigonometric System is an Orthonormal Basis of the Square-Integrable Space of the Torus (i), U−U′=0L2U-U'=0_{L^{2}}, and adding U′U' to both sides, using the conditions of Vector Space over a Field and claim 2 of Elementary Identities in a Vector Space, gives U=U′U=U'.

Claim 3. E^k(k′)=⟨Ek,Ek′⟩L2\hat{E}_{k}(k')=\langle E_{k},E_{k'}\rangle_{L^{2}}, which equals 11 if k=k′k=k' and 00 otherwise by The Trigonometric System on the Torus is Orthonormal (iii).

Claim 4. Let κ\kappa be an enumeration and U,U′∈L2(Tn)U,U'\in L^{2}(\mathbb{T}^{n}). Since ⟨U,Eκ(j)⟩L2=U^(κ(j))\langle U,E_{\kappa(j)}\rangle_{L^{2}}=\hat{U}(\kappa(j)), Orthonormal Expansions in a Real Hilbert Space (iv), applied with x=Ux=U and y=U′y=U', states that ∑j=1∞U^(κ(j))U^′(κ(j))\sum_{j=1}^{\infty}\hat{U}(\kappa(j))\hat{U}'(\kappa(j)) converges with sum ⟨U,U′⟩L2\langle U,U'\rangle_{L^{2}} and that ∑j=1∞U^(κ(j))2\sum_{j=1}^{\infty}\hat{U}(\kappa(j))^{2} has sum ⟨U,U⟩L2=(∥U∥L2)2\langle U,U\rangle_{L^{2}}=(\lVert U\rVert_{L^{2}})^{2}, the last equality by Real Inner Product Spaces (ii); and Orthonormal Expansions in a Real Hilbert Space (iii) states that ∑j=1∞U^(κ(j))Eκ(j)\sum_{j=1}^{\infty}\hat{U}(\kappa(j))E_{\kappa(j)} converges in L2(Tn)L^{2}(\mathbb{T}^{n}) with sum UU.

Claim 5. Let κ\kappa be an enumeration and c∈Map(Zn,R)c\in\mathrm{Map}(\mathbb{Z}^{n},\mathbb{R}) with ∑j=1∞c(κ(j))2\sum_{j=1}^{\infty}c(\kappa(j))^{2} convergent. Applying Orthonormal Expansions in a Real Hilbert Space (ii) to the sequence cj=c(κ(j))c_{j}=c(\kappa(j)) of real numbers, the series ∑j=1∞c(κ(j))Eκ(j)\sum_{j=1}^{\infty}c(\kappa(j))E_{\kappa(j)} converges in L2(Tn)L^{2}(\mathbb{T}^{n}); writing UU for its sum, (∥U∥L2)2=∑j=1∞c(κ(j))2(\lVert U\rVert_{L^{2}})^{2}=\sum_{j=1}^{\infty}c(\kappa(j))^{2} and ⟨U,Eκ(j)⟩L2=c(κ(j))\langle U,E_{\kappa(j)}\rangle_{L^{2}}=c(\kappa(j)) for every j∈Nj\in\mathbb{N}. Given k∈Znk\in\mathbb{Z}^{n}, by Bijection of Sets there is j∈Nj\in\mathbb{N} with κ(j)=k\kappa(j)=k, so U^(k)=c(k)\hat{U}(k)=c(k); hence U^=c\hat{U}=c. If U′∈L2(Tn)U'\in L^{2}(\mathbb{T}^{n}) also satisfies U^′=c\hat{U}'=c, then U′=UU'=U by claim 2. This proves claim 5.

Claim 6. Fix m∈Nm\in\mathbb{N}. Since ρk≠0\rho_{k}\ne0, claim 4 of Properties of Natural Number Powers in a Field gives ρkm≠0\rho_{k}^{m}\ne0 for every k∈Znk\in\mathbb{Z}^{n}, so (ρkm)−1(\rho_{k}^{m})^{-1} exists.

Uniqueness. If W,W′∈L2(Tn)W,W'\in L^{2}(\mathbb{T}^{n}) both satisfy W^(k)=ρkmc(k)=W^′(k)\hat{W}(k)=\rho_{k}^{m}c(k)=\hat{W}'(k) for every kk, then W^=W^′\hat{W}=\hat{W}' and W=W′W=W' by claim 2.

Subspace. By Elementary Identities in a Real Inner Product Space (ii), 0L2^(k)=⟨0L2,Ek⟩L2=0=ρkm⋅0\widehat{0_{L^{2}}}(k)=\langle0_{L^{2}},E_{k}\rangle_{L^{2}}=0=\rho_{k}^{m}\cdot0 for every kk, so the zero family lies in Hm\mathcal{H}_{m}, with Λm\Lambda_{m} of it equal to 0L20_{L^{2}}. Let c,d∈Hmc,d\in\mathcal{H}_{m}, t∈Rt\in\mathbb{R}, W=ΛmcW=\Lambda_{m}c and W′=ΛmdW'=\Lambda_{m}d. By claim 1 and distributivity in R\mathbb{R}, for every kk,

W+W′^(k)=ρkmc(k)+ρkmd(k)=ρkm (c+d)(k),tW^(k)=t ρkmc(k)=ρkm (tc)(k),\widehat{W+W'}(k)=\rho_{k}^{m}c(k)+\rho_{k}^{m}d(k)=\rho_{k}^{m}\,(c+d)(k),\qquad \widehat{tW}(k)=t\,\rho_{k}^{m}c(k)=\rho_{k}^{m}\,(tc)(k),

so c+dc+d and tctc lie in Hm\mathcal{H}_{m}, with Λm(c+d)=W+W′=Λmc+Λmd\Lambda_{m}(c+d)=W+W'=\Lambda_{m}c+\Lambda_{m}d and Λm(tc)=tW=tΛmc\Lambda_{m}(tc)=tW=t\Lambda_{m}c by uniqueness. Thus Hm\mathcal{H}_{m} satisfies the three conditions of Linear Subspace, so it is a linear subspace of Map(Zn,R)\mathrm{Map}(\mathbb{Z}^{n},\mathbb{R}) and, by The Real Vector Space of Real-Valued Functions on a Set (ii), a real vector space under the pointwise operations; and Λm\Lambda_{m} satisfies the two conditions of Linear Map for that vector space structure.

Bijection. Let W∈L2(Tn)W\in L^{2}(\mathbb{T}^{n}). Define c∈Map(Zn,R)c\in\mathrm{Map}(\mathbb{Z}^{n},\mathbb{R}) by c(k)=(ρkm)−1W^(k)c(k)=(\rho_{k}^{m})^{-1}\hat{W}(k); then ρkmc(k)=W^(k)\rho_{k}^{m}c(k)=\hat{W}(k) for every kk, so c∈Hmc\in\mathcal{H}_{m} and Λmc=W\Lambda_{m}c=W. If d∈Hmd\in\mathcal{H}_{m} also satisfies Λmd=W\Lambda_{m}d=W, then ρkmd(k)=W^(k)=ρkmc(k)\rho_{k}^{m}d(k)=\hat{W}(k)=\rho_{k}^{m}c(k) for every kk, and multiplying by (ρkm)−1(\rho_{k}^{m})^{-1} gives d(k)=c(k)d(k)=c(k), so d=cd=c. Hence every WW has exactly one preimage under Λm\Lambda_{m}, which is the condition of Bijection of Sets.

Citations

Loading…

Dependencies

Uses0

Loading…

Comments

Log in to comment.

Loading…