Each result cited below is universally quantified over the data appearing in its own statement, and is applied to the data named here. The measure space is ( Q , B Q , λ Q ) (Q,\mathcal{B}_{Q},\lambda_{Q}) ( Q , B Q , λ Q ) of clause 3 , with λ Q ( Q ) = 1 \lambda_{Q}(Q)=1 λ Q ( Q ) = 1 , unless ( R n , B ( R n ) , λ n ) (\mathbb{R}^{n},\mathcal{B}(\mathbb{R}^{n}),\lambda_{n}) ( R n , B ( R n ) , λ n ) is named instead. We use repeatedly that a map u ∈ C p e r u\in C_{\mathrm{per}} u ∈ C per , being continuous on R n \mathbb{R}^{n} R n , is measurable with respect to B ( R n ) \mathcal{B}(\mathbb{R}^{n}) B ( R n ) and B ( R ) \mathcal{B}(\mathbb{R}) B ( R ) , by claim 3(a) of The Borel Sigma-Algebra of a Euclidean Space as a Product, and Measurability of Projections, Sequentially Continuous Maps, and Open and Closed Sets together with claim 5 there.
Claim 1. For a real c c c we have
{ x ∈ Q : c < u ( x ) } = Q ∩ { x ∈ R n : c < u ( x ) } , \{x\in Q:c<u(x)\}=Q\cap\{x\in\mathbb{R}^{n}:c<u(x)\}, { x ∈ Q : c < u ( x )} = Q ∩ { x ∈ R n : c < u ( x )} ,
which belongs to B ( R n ) \mathcal{B}(\mathbb{R}^{n}) B ( R n ) and is contained in Q Q Q , hence belongs to B Q \mathcal{B}_{Q} B Q . By the criterion of clause 3 , u ∣ Q u|_{Q} u ∣ Q is measurable. Next, ∣ u ( x ) ∣ ≤ M |u(x)|\le M ∣ u ( x ) ∣ ≤ M for every x x x , so by Properties of Real Powers of Nonnegative Real Numbers §monotone the power ( ∣ u ( x ) ∣ ) p (|u(x)|)^{p} ( ∣ u ( x ) ∣ ) p is at most M p M^{p} M p for every x ∈ Q x\in Q x ∈ Q . The constant function on Q Q Q with value M p M^{p} M p is M p 1 Q M^{p}\,\mathbf{1}_{Q} M p 1 Q , whose integral is M p λ Q ( Q ) = M p M^{p}\lambda_{Q}(Q)=M^{p} M p λ Q ( Q ) = M p by The Integral of an Indicator Function is the Measure of the Set and claim 1 of Linearity and Monotonicity of the Lebesgue Integral . Monotonicity of the integral of nonnegative measurable functions (claim 1 of Linearity and Monotonicity of the Lebesgue Integral ) therefore gives
∫ Q ∣ u ∣ Q ∣ p d λ Q ≤ M p < ∞ , \int_{Q}\bigl|u|_{Q}\bigr|^{p}\,d\lambda_{Q}\le M^{p}<\infty , ∫ Q u ∣ Q p d λ Q ≤ M p < ∞ ,
so u ∣ Q ∈ L p ( T n ) u|_{Q}\in\mathcal{L}^{p}(\mathbb{T}^{n}) u ∣ Q ∈ L p ( T n ) by Power-Integrable Functions and the p-Seminorm §space . Applying the monotone power t ↦ t 1 / p t\mapsto t^{1/p} t ↦ t 1/ p (Properties of Real Powers of Nonnegative Real Numbers §monotone ) and the identity ( M p ) 1 / p = M (M^{p})^{1/p}=M ( M p ) 1/ p = M (Properties of Real Powers of Nonnegative Real Numbers §inverse ) gives ∥ u ∣ Q ∥ p ≤ M \lVert u|_{Q}\rVert_{p}\le M ∥ u ∣ Q ∥ p ≤ M for the p p p -seminorm , and ∥ [ u ∣ Q ] ∥ L p ( T n ) = ∥ u ∣ Q ∥ p \lVert[u|_{Q}]\rVert_{L^{p}(\mathbb{T}^{n})}=\lVert u|_{Q}\rVert_{p} ∥[ u ∣ Q ] ∥ L p ( T n ) = ∥ u ∣ Q ∥ p by The Lebesgue Space of Power-Integrable Functions §norm .
Claim 2. Let M M M be as in claim 1. The product 1 Q u \mathbf{1}_{Q}u 1 Q u is measurable with respect to B ( R n ) \mathcal{B}(\mathbb{R}^{n}) B ( R n ) : for a real c c c the set { x : c < 1 Q ( x ) u ( x ) } \{x:c<\mathbf{1}_{Q}(x)u(x)\} { x : c < 1 Q ( x ) u ( x )} equals Q ∩ { x : c < u ( x ) } Q\cap\{x:c<u(x)\} Q ∩ { x : c < u ( x )} when 0 ≤ c 0\le c 0 ≤ c , and equals the union of that set with the complement of Q Q Q when c < 0 c<0 c < 0 ; both lie in B ( R n ) \mathcal{B}(\mathbb{R}^{n}) B ( R n ) . Moreover ∣ 1 Q u ∣ ≤ M 1 Q |\mathbf{1}_{Q}u|\le M\mathbf{1}_{Q} ∣ 1 Q u ∣ ≤ M 1 Q pointwise, and ∫ R n M 1 Q d λ n = M λ n ( Q ) = M \int_{\mathbb{R}^{n}}M\mathbf{1}_{Q}\,d\lambda_{n}=M\lambda_{n}(Q)=M ∫ R n M 1 Q d λ n = M λ n ( Q ) = M is finite by The Integral of an Indicator Function is the Measure of the Set , The Half-Open Unit Cell Tiles Euclidean Space §cell and claim 1 of Linearity and Monotonicity of the Lebesgue Integral ; so 1 Q u \mathbf{1}_{Q}u 1 Q u is integrable by the criterion recorded in Integrable Function and the Lebesgue Integral .
Let u + u^{+} u + and u − u^{-} u − be the positive and negative parts of u u u in the sense of Integrable Function and the Lebesgue Integral . The zero extension of ( u ∣ Q ) + (u|_{Q})^{+} ( u ∣ Q ) + off Q Q Q is 1 Q u + \mathbf{1}_{Q}u^{+} 1 Q u + , and likewise for the negative parts. Claim 1 of Assembly of Measure Spaces: Restriction, Transport, One-Point Spaces, and Countable Disjoint Unions , applied to the measure space ( R n , B ( R n ) , λ n ) (\mathbb{R}^{n},\mathcal{B}(\mathbb{R}^{n}),\lambda_{n}) ( R n , B ( R n ) , λ n ) and to Q ∈ B ( R n ) Q\in\mathcal{B}(\mathbb{R}^{n}) Q ∈ B ( R n ) , therefore gives
∫ Q ( u ∣ Q ) ± d λ Q = ∫ R n 1 Q u ± d λ n , \int_{Q}(u|_{Q})^{\pm}\,d\lambda_{Q}=\int_{\mathbb{R}^{n}}\mathbf{1}_{Q}u^{\pm}\,d\lambda_{n}, ∫ Q ( u ∣ Q ) ± d λ Q = ∫ R n 1 Q u ± d λ n ,
both sides being finite. Subtracting the two identities and using Integrable Function and the Lebesgue Integral on each side yields ∫ T n u ∣ Q d x = ∫ R n 1 Q u d λ n \int_{\mathbb{T}^{n}}u|_{Q}\,dx=\int_{\mathbb{R}^{n}}\mathbf{1}_{Q}u\,d\lambda_{n} ∫ T n u ∣ Q d x = ∫ R n 1 Q u d λ n .
Claim 3. The maps u + v u+v u + v and c u cu c u lie in C p e r C_{\mathrm{per}} C per by Elementary Properties of Lattice-Periodic Functions §algebra , so their restrictions lie in L p ( T n ) \mathcal{L}^{p}(\mathbb{T}^{n}) L p ( T n ) by claim 1. Restriction commutes with the pointwise operations, that is ( u + v ) ∣ Q = u ∣ Q + v ∣ Q (u+v)|_{Q}=u|_{Q}+v|_{Q} ( u + v ) ∣ Q = u ∣ Q + v ∣ Q and ( c u ) ∣ Q = c ( u ∣ Q ) (cu)|_{Q}=c\,(u|_{Q}) ( c u ) ∣ Q = c ( u ∣ Q ) as maps on Q Q Q , and the operations on classes are defined by representatives in The Lebesgue Space of Power-Integrable Functions §space . The two identities follow.
Claim 4. We first prove the approximation statement. Let f ∈ L p ( T n ) f\in\mathcal{L}^{p}(\mathbb{T}^{n}) f ∈ L p ( T n ) and let ε \varepsilon ε be a positive real.
Step 1: reduction to a simple function. By Simple Functions are Dense in the Lebesgue Space §approximation there is a simple function s s s on ( Q , B Q ) (Q,\mathcal{B}_{Q}) ( Q , B Q ) lying in L p ( T n ) \mathcal{L}^{p}(\mathbb{T}^{n}) L p ( T n ) with ∥ f − s ∥ p ≤ ε / 2 \lVert f-s\rVert_{p}\le\varepsilon/2 ∥ f − s ∥ p ≤ ε /2 .
Step 2: approximating one indicator. We show: for every A ∈ B Q A\in\mathcal{B}_{Q} A ∈ B Q and every positive real η \eta η there is v ∈ C p e r v\in C_{\mathrm{per}} v ∈ C per with ∥ 1 A − v ∣ Q ∥ p ≤ η \lVert\mathbf{1}_{A}-v|_{Q}\rVert_{p}\le\eta ∥ 1 A − v ∣ Q ∥ p ≤ η .
Let β \beta β be a positive real with β ≤ 1 4 \beta\le\tfrac14 β ≤ 4 1 , to be fixed at the end, and put
V = { x ∈ R n : β < x i < 1 − β for every i ∈ [ n ] } . V=\{x\in\mathbb{R}^{n}:\beta<x_{i}<1-\beta\ \text{for every}\ i\in[n]\}. V = { x ∈ R n : β < x i < 1 − β for every i ∈ [ n ]} .
V V V is open: for x ∈ V x\in V x ∈ V let r r r be the least element of the finite set { x i − β : i ∈ [ n ] } ∪ { 1 − β − x i : i ∈ [ n ] } \{x_{i}-\beta:i\in[n]\}\cup\{1-\beta-x_{i}:i\in[n]\} { x i − β : i ∈ [ n ]} ∪ { 1 − β − x i : i ∈ [ n ]} of positive reals, positive by repeated use of claim 9 of Elementary Order Arithmetic in an Ordered Field ; if ∥ y − x ∥ < r \lVert y-x\rVert<r ∥ y − x ∥ < r then ∣ y i − x i ∣ < r |y_{i}-x_{i}|<r ∣ y i − x i ∣ < r for every i i i by claim 4 of Elementary Properties of the Euclidean Norm on R n \mathbb{R}^n R n , whence β < y i < 1 − β \beta<y_{i}<1-\beta β < y i < 1 − β . Also V ⊆ Q ˚ V\subseteq\mathring{Q} V ⊆ Q ˚ , since 0 < β 0<\beta 0 < β and 1 − β < 1 1-\beta<1 1 − β < 1 .
Put A β = A ∩ V A_{\beta}=A\cap V A β = A ∩ V , a member of B Q \mathcal{B}_{Q} B Q . Every x ∈ A ∖ A β x\in A\setminus A_{\beta} x ∈ A ∖ A β lies in Q Q Q and has some coordinate with x i ≤ β x_{i}\le\beta x i ≤ β or 1 − β ≤ x i 1-\beta\le x_{i} 1 − β ≤ x i , so
A ∖ A β ⊆ ⋃ i ∈ [ n ] ( S i ∪ T i ) , S i = { x ∈ Q : x i ≤ β } , T i = { x ∈ Q : 1 − β ≤ x i } . A\setminus A_{\beta}\subseteq\bigcup_{i\in[n]}\bigl(S_{i}\cup T_{i}\bigr),\qquad S_{i}=\{x\in Q:x_{i}\le\beta\},\quad T_{i}=\{x\in Q:1-\beta\le x_{i}\}. A ∖ A β ⊆ i ∈ [ n ] ⋃ ( S i ∪ T i ) , S i = { x ∈ Q : x i ≤ β } , T i = { x ∈ Q : 1 − β ≤ x i } .
Each S i S_{i} S i is the Borel rectangle whose i i i th factor is the interval { t : 0 ≤ t ≤ β } \{t:0\le t\le\beta\} { t : 0 ≤ t ≤ β } and whose other factors are { t : 0 ≤ t < 1 } \{t:0\le t<1\} { t : 0 ≤ t < 1 } , and each T i T_{i} T i is the Borel rectangle whose i i i th factor is { t : 1 − β ≤ t < 1 } \{t:1-\beta\le t<1\} { t : 1 − β ≤ t < 1 } and whose other factors are the same; by claim 4 of Existence of Lebesgue Measure on the Real Line and the rectangle identity of Lebesgue Measure on R n \mathbb{R}^n R n each has measure β \beta β . By the monotonicity and finite subadditivity of a measure recorded in Basic Properties of a Measure , λ n ( A ∖ A β ) ≤ 2 n β \lambda_{n}(A\setminus A_{\beta})\le2n\beta λ n ( A ∖ A β ) ≤ 2 n β .
Since λ n ( A β ) ≤ λ n ( Q ) = 1 < ∞ \lambda_{n}(A_{\beta})\le\lambda_{n}(Q)=1<\infty λ n ( A β ) ≤ λ n ( Q ) = 1 < ∞ , Outer and Inner Regularity of Lebesgue Measure on R n \mathbb{R}^n R n §inner-compact provides a compact K ⊆ A β K\subseteq A_{\beta} K ⊆ A β with λ n ( A β ∖ K ) ≤ β \lambda_{n}(A_{\beta}\setminus K)\le\beta λ n ( A β ∖ K ) ≤ β , and Outer and Inner Regularity of Lebesgue Measure on R n \mathbb{R}^n R n §outer provides an open U 0 ⊇ K U_{0}\supseteq K U 0 ⊇ K with λ n ( U 0 ∖ K ) ≤ β \lambda_{n}(U_{0}\setminus K)\le\beta λ n ( U 0 ∖ K ) ≤ β . Put U = U 0 ∩ V U=U_{0}\cap V U = U 0 ∩ V , an open set with K ⊆ U K\subseteq U K ⊆ U (because K ⊆ A β ⊆ V K\subseteq A_{\beta}\subseteq V K ⊆ A β ⊆ V ), with U ⊆ V ⊆ Q ˚ U\subseteq V\subseteq\mathring{Q} U ⊆ V ⊆ Q ˚ , and with λ n ( U ∖ K ) ≤ λ n ( U 0 ∖ K ) ≤ β \lambda_{n}(U\setminus K)\le\lambda_{n}(U_{0}\setminus K)\le\beta λ n ( U ∖ K ) ≤ λ n ( U 0 ∖ K ) ≤ β .
Apply Continuous Partition of Unity Subordinate to a Finite Open Cover of a Compact Set in a Metric Space in the metric space ( R n , d E ) (\mathbb{R}^{n},d_{E}) ( R n , d E ) to the compact set K K K and the one-member family consisting of the open set U U U , which covers K K K . It yields a continuous g : R n → R g:\mathbb{R}^{n}\to\mathbb{R} g : R n → R with 0 ≤ g ≤ 1 0\le g\le1 0 ≤ g ≤ 1 and a closed D ⊆ U D\subseteq U D ⊆ U such that g g g vanishes off D D D and g ( x ) = 1 g(x)=1 g ( x ) = 1 for every x ∈ K x\in K x ∈ K . In particular g g g vanishes off U U U , and U ⊆ Q ˚ U\subseteq\mathring{Q} U ⊆ Q ˚ , so Elementary Properties of Lattice-Periodic Functions §periodisation applies with W = U W=U W = U : the map v = g ∘ π v=g\circ\pi v = g ∘ π lies in C p e r C_{\mathrm{per}} C per , agrees with g g g on Q Q Q , and satisfies 0 ≤ v ≤ 1 0\le v\le1 0 ≤ v ≤ 1 .
Put E = ( ( A ∖ K ) ∪ ( U ∖ K ) ) ∩ Q E=\bigl((A\setminus K)\cup(U\setminus K)\bigr)\cap Q E = ( ( A ∖ K ) ∪ ( U ∖ K ) ) ∩ Q , a member of B Q \mathcal{B}_{Q} B Q . Note first that ( 1 E ) p = 1 E (\mathbf{1}_{E})^{p}=\mathbf{1}_{E} ( 1 E ) p = 1 E : indeed 0 p = 0 0^{p}=0 0 p = 0 by Real Power of a Nonnegative Real Number §power , while 1 p = 1 p ⋅ 1 p 1^{p}=1^{p}\cdot1^{p} 1 p = 1 p ⋅ 1 p by Properties of Real Powers of Nonnegative Real Numbers §product and 1 p ≠ 0 1^{p}\ne0 1 p = 0 by Properties of Real Powers of Nonnegative Real Numbers §values , so cancelling the nonzero factor gives 1 p = 1 1^{p}=1 1 p = 1 . Hence 1 E ∈ L p ( T n ) \mathbf{1}_{E}\in\mathcal{L}^{p}(\mathbb{T}^{n}) 1 E ∈ L p ( T n ) , its p p p th power having integral λ Q ( E ) ≤ λ Q ( Q ) = 1 < ∞ \lambda_{Q}(E)\le\lambda_{Q}(Q)=1<\infty λ Q ( E ) ≤ λ Q ( Q ) = 1 < ∞ by The Integral of an Indicator Function is the Measure of the Set and Basic Properties of a Measure . For x ∈ Q x\in Q x ∈ Q with x ∉ E x\notin E x ∈ / E we have 1 A ( x ) = v ( x ) \mathbf{1}_{A}(x)=v(x) 1 A ( x ) = v ( x ) : if x ∈ K x\in K x ∈ K then x ∈ A x\in A x ∈ A and g ( x ) = 1 g(x)=1 g ( x ) = 1 , so both values are 1 1 1 ; if x ∉ K x\notin K x ∈ / K then x ∉ A x\notin A x ∈ / A and x ∉ U x\notin U x ∈ / U , so 1 A ( x ) = 0 \mathbf{1}_{A}(x)=0 1 A ( x ) = 0 and g ( x ) = 0 g(x)=0 g ( x ) = 0 . For x ∈ E x\in E x ∈ E both 1 A ( x ) \mathbf{1}_{A}(x) 1 A ( x ) and v ( x ) v(x) v ( x ) lie between 0 0 0 and 1 1 1 , so ∣ 1 A ( x ) − v ( x ) ∣ ≤ 1 |\mathbf{1}_{A}(x)-v(x)|\le1 ∣ 1 A ( x ) − v ( x ) ∣ ≤ 1 . Hence ∣ 1 A − v ∣ Q ∣ ≤ 1 E |\mathbf{1}_{A}-v|_{Q}|\le\mathbf{1}_{E} ∣ 1 A − v ∣ Q ∣ ≤ 1 E pointwise on Q Q Q , and Elementary Properties of the p-Seminorm §comparison gives
∥ 1 A − v ∣ Q ∥ p ≤ ∥ 1 E ∥ p = ( λ Q ( E ) ) 1 / p , \lVert\mathbf{1}_{A}-v|_{Q}\rVert_{p}\le\lVert\mathbf{1}_{E}\rVert_{p}=\bigl(\lambda_{Q}(E)\bigr)^{1/p}, ∥ 1 A − v ∣ Q ∥ p ≤ ∥ 1 E ∥ p = ( λ Q ( E ) ) 1/ p ,
using ( 1 E ) p = 1 E (\mathbf{1}_{E})^{p}=\mathbf{1}_{E} ( 1 E ) p = 1 E as just noted, together with The Integral of an Indicator Function is the Measure of the Set . Finally
λ Q ( E ) ≤ λ n ( A ∖ K ) + λ n ( U ∖ K ) ≤ ( λ n ( A ∖ A β ) + λ n ( A β ∖ K ) ) + β ≤ 2 n β + β + β = ( 2 n + 2 ) β , \lambda_{Q}(E)\le\lambda_{n}(A\setminus K)+\lambda_{n}(U\setminus K)\le\bigl(\lambda_{n}(A\setminus A_{\beta})+\lambda_{n}(A_{\beta}\setminus K)\bigr)+\beta\le2n\beta+\beta+\beta=(2n+2)\beta, λ Q ( E ) ≤ λ n ( A ∖ K ) + λ n ( U ∖ K ) ≤ ( λ n ( A ∖ A β ) + λ n ( A β ∖ K ) ) + β ≤ 2 n β + β + β = ( 2 n + 2 ) β ,
again by Basic Properties of a Measure . Choosing β \beta β to be the least of 1 4 \tfrac14 4 1 and η p / ( 2 n + 2 ) \eta^{p}/(2n+2) η p / ( 2 n + 2 ) makes λ Q ( E ) ≤ η p \lambda_{Q}(E)\le\eta^{p} λ Q ( E ) ≤ η p and hence ∥ 1 A − v ∣ Q ∥ p ≤ ( η p ) 1 / p = η \lVert\mathbf{1}_{A}-v|_{Q}\rVert_{p}\le(\eta^{p})^{1/p}=\eta ∥ 1 A − v ∣ Q ∥ p ≤ ( η p ) 1/ p = η , by Properties of Real Powers of Nonnegative Real Numbers §monotone and Properties of Real Powers of Nonnegative Real Numbers §inverse .
Step 3: from indicators to s s s . Let s = ∑ l = 1 r c l 1 A l s=\sum_{l=1}^{r}c_{l}\mathbf{1}_{A_{l}} s = ∑ l = 1 r c l 1 A l be the standard representation of s s s , with distinct values c l c_{l} c l and A l = s − 1 ( { c l } ) ∈ B Q A_{l}=s^{-1}(\{c_{l}\})\in\mathcal{B}_{Q} A l = s − 1 ({ c l }) ∈ B Q . Let L = { l ∈ [ r ] : c l ≠ 0 } L=\{l\in[r]:c_{l}\ne0\} L = { l ∈ [ r ] : c l = 0 } ; since the terms with c l = 0 c_{l}=0 c l = 0 contribute nothing, s = ∑ l ∈ L c l 1 A l s=\sum_{l\in L}c_{l}\mathbf{1}_{A_{l}} s = ∑ l ∈ L c l 1 A l pointwise on Q Q Q . If L L L is empty then s s s is the zero map, and the map u u u with constant value 0 0 0 lies in C p e r C_{\mathrm{per}} C per with ∥ s − u ∣ Q ∥ p = 0 ≤ ε / 2 \lVert s-u|_{Q}\rVert_{p}=0\le\varepsilon/2 ∥ s − u ∣ Q ∥ p = 0 ≤ ε /2 . Otherwise put C = ∑ l ∈ L ∣ c l ∣ C=\sum_{l\in L}|c_{l}| C = ∑ l ∈ L ∣ c l ∣ , a positive real. Apply Step 2 to each A l A_{l} A l , l ∈ L l\in L l ∈ L , with η = ε / ( 2 C ) \eta=\varepsilon/(2C) η = ε / ( 2 C ) , obtaining v l ∈ C p e r v_{l}\in C_{\mathrm{per}} v l ∈ C per with ∥ 1 A l − v l ∣ Q ∥ p ≤ ε / ( 2 C ) \lVert\mathbf{1}_{A_{l}}-v_{l}|_{Q}\rVert_{p}\le\varepsilon/(2C) ∥ 1 A l − v l ∣ Q ∥ p ≤ ε / ( 2 C ) , and put u = ∑ l ∈ L c l v l u=\sum_{l\in L}c_{l}v_{l} u = ∑ l ∈ L c l v l , which lies in C p e r C_{\mathrm{per}} C per by Elementary Properties of Lattice-Periodic Functions §algebra . Then s − u ∣ Q = ∑ l ∈ L c l ( 1 A l − v l ∣ Q ) s-u|_{Q}=\sum_{l\in L}c_{l}\bigl(\mathbf{1}_{A_{l}}-v_{l}|_{Q}\bigr) s − u ∣ Q = ∑ l ∈ L c l ( 1 A l − v l ∣ Q ) pointwise on Q Q Q , so by Minkowski's Inequality and the Seminormed Space of Power-Integrable Functions §finite-sums and Elementary Properties of the p-Seminorm §homogeneous ,
∥ s − u ∣ Q ∥ p ≤ ∑ l ∈ L ∣ c l ∣ ∥ 1 A l − v l ∣ Q ∥ p ≤ C ⋅ ε 2 C = ε 2 . \lVert s-u|_{Q}\rVert_{p}\le\sum_{l\in L}|c_{l}|\,\bigl\lVert\mathbf{1}_{A_{l}}-v_{l}|_{Q}\bigr\rVert_{p}\le C\cdot\frac{\varepsilon}{2C}=\frac{\varepsilon}{2}. ∥ s − u ∣ Q ∥ p ≤ l ∈ L ∑ ∣ c l ∣ 1 A l − v l ∣ Q p ≤ C ⋅ 2 C ε = 2 ε .
Step 4: conclusion. By Minkowski's Inequality and the Seminormed Space of Power-Integrable Functions §minkowski ,
∥ f − u ∣ Q ∥ p ≤ ∥ f − s ∥ p + ∥ s − u ∣ Q ∥ p ≤ ε 2 + ε 2 = ε , \bigl\lVert f-u|_{Q}\bigr\rVert_{p}\le\lVert f-s\rVert_{p}+\bigl\lVert s-u|_{Q}\bigr\rVert_{p}\le\frac{\varepsilon}{2}+\frac{\varepsilon}{2}=\varepsilon , f − u ∣ Q p ≤ ∥ f − s ∥ p + s − u ∣ Q p ≤ 2 ε + 2 ε = ε ,
and ∥ [ f ] − [ u ∣ Q ] ∥ L p ( T n ) = ∥ f − u ∣ Q ∥ p \lVert[f]-[u|_{Q}]\rVert_{L^{p}(\mathbb{T}^{n})}=\lVert f-u|_{Q}\rVert_{p} ∥[ f ] − [ u ∣ Q ] ∥ L p ( T n ) = ∥ f − u ∣ Q ∥ p by The Lebesgue Space of Power-Integrable Functions §space and The Lebesgue Space of Power-Integrable Functions §norm . This is the approximation statement.
Density. Let F ∈ L p ( T n ) F\in L^{p}(\mathbb{T}^{n}) F ∈ L p ( T n ) and choose a representative f ∈ L p ( T n ) f\in\mathcal{L}^{p}(\mathbb{T}^{n}) f ∈ L p ( T n ) with F = [ f ] F=[f] F = [ f ] . For each k ∈ N k\in\mathbb{N} k ∈ N the approximation statement, applied with ε = 1 / k \varepsilon=1/k ε = 1/ k , provides u k ∈ C p e r u_{k}\in C_{\mathrm{per}} u k ∈ C per with ∥ F − [ u k ∣ Q ] ∥ L p ( T n ) ≤ 1 / k \lVert F-[u_{k}|_{Q}]\rVert_{L^{p}(\mathbb{T}^{n})}\le1/k ∥ F − [ u k ∣ Q ] ∥ L p ( T n ) ≤ 1/ k ; the choice of one such u k u_{k} u k for each k k k uses countable choice. Given a positive real ε \varepsilon ε , The Archimedean Property of the Real Numbers provides k 0 ∈ N k_{0}\in\mathbb{N} k 0 ∈ N with 1 < k 0 ε 1<k_{0}\varepsilon 1 < k 0 ε , and then ∥ F − [ u k ∣ Q ] ∥ L p ( T n ) ≤ 1 / k < ε \lVert F-[u_{k}|_{Q}]\rVert_{L^{p}(\mathbb{T}^{n})}\le1/k<\varepsilon ∥ F − [ u k ∣ Q ] ∥ L p ( T n ) ≤ 1/ k < ε for every k ≥ k 0 k\ge k_{0} k ≥ k 0 . So the sequence ( [ u k ∣ Q ] ) k ∈ N ([u_{k}|_{Q}])_{k\in\mathbb{N}} ([ u k ∣ Q ] ) k ∈ N in C \mathcal{C} C converges to F F F in the metric of L p ( T n ) L^{p}(\mathbb{T}^{n}) L p ( T n ) , and F F F lies in the closure of C \mathcal{C} C by Sequential Characterization of the Closure in a Metric Space . As F F F was arbitrary, the closure of C \mathcal{C} C is all of L p ( T n ) L^{p}(\mathbb{T}^{n}) L p ( T n ) , that is, C \mathcal{C} C is dense .