5.4. Mathlib candidates: Schwartz space and convolution
Weighted smoothing, Lizorkin and Schwartz-space auxiliaries, Young's convolution inequality, and a reusable temperate-growth closure theorem.
Weighted L¹ smoothing
-
abs_pow_sub_pow_le_mul[complete] -
norm_sub_taylorSum_le[complete] -
MeasureTheory.tendsto_setIntegral_norm_abs_gt[complete] -
MeasureTheory.tendsto_integral_norm_sub_comp_sub_right[complete] -
MeasureTheory.tendsto_integral_weight_norm_sub_comp_sub_right[complete] -
MeasureTheory.integrable_weight_norm_comp_sub[complete] -
MeasureTheory.integrable_weight_norm_sub_comp_sub[complete] -
MeasureTheory.continuous_integral_weight_norm_sub[complete] -
MeasureTheory.integral_comp_sub_div_smul[complete] -
MeasureTheory.tendsto_integral_weight_norm_smoothing_sub[complete] -
MeasureTheory.tendsto_integral_mul_smoothing_of_vanishing_moments[complete]
Polynomial-weighted L¹ smoothing estimates. The estimates driving the function-level pairing theory of tempered distributions modulo polynomials: a Lipschitz bound |x^n-y^n|\le nM^{n-1}|x-y| on [0,M]; a uniform Taylor remainder bound \|f(a+h)-\sum_{j\le k}(h^j/j!)f^{(j)}(a)\|\le C|h|^{k+1}/k! on unordered intervals (both endpoints of Mathlib's taylor_mean_remainder_bound, through reflection); the vanishing of integral tails \int_{|r|>c}\|g\|\to0; the continuity of translation in L¹(ℝ) — obtained from MeasureTheory.Lp.compMeasurePreserving_continuous, which Mathlib provides without this classical corollary — together with its polynomially weighted form; the weighted integrability and continuity of the translation modulus \omega(s)=\int(1+|z|)^k\|\Xi(z-s)-\Xi(z)\|dz; an E-valued affine change of variables \int F((w-r)/c)\,dw=c\cdot\int F; the scaled approximate identity in weighted L¹ (\int(1+|z|)^k\|\int K(u)\Xi(z-u/c)\,du-\Xi(z)\|dz\to0 as c\to\infty for kernels K of unit integral and finite k-th moment, by a Tonelli swap and dominated convergence against the modulus \omega); and the vanishing-moment cancellation for wide smoothing (last declaration): if \Xi has a finite k-th moment and vanishing moments up to k and \eta is strongly measurable with growth at most (1+|z|)^k, then \int\Xi(r)\,c^{-1}\!\int\eta(w)\theta((w-r)/c)\,dw\,dr\to0 as c → ∞ for every Schwartz kernel \theta. The proof Taylor-expands the kernel to order k, splits the line at |r|=c, bounds the outside by the integral tail, converts the truncated Taylor moments into tails through the vanishing moments, and dominates the remainder by \min(1,|r|/c); exactly k moments suffice.
Lean code for Theorem5.4.1●11 theorems
Associated Lean declarations
-
abs_pow_sub_pow_le_mul[complete]
-
norm_sub_taylorSum_le[complete]
-
MeasureTheory.tendsto_setIntegral_norm_abs_gt[complete]
-
MeasureTheory.tendsto_integral_norm_sub_comp_sub_right[complete]
-
MeasureTheory.tendsto_integral_weight_norm_sub_comp_sub_right[complete]
-
MeasureTheory.integrable_weight_norm_comp_sub[complete]
-
MeasureTheory.integrable_weight_norm_sub_comp_sub[complete]
-
MeasureTheory.continuous_integral_weight_norm_sub[complete]
-
MeasureTheory.integral_comp_sub_div_smul[complete]
-
MeasureTheory.tendsto_integral_weight_norm_smoothing_sub[complete]
-
MeasureTheory.tendsto_integral_mul_smoothing_of_vanishing_moments[complete]
-
abs_pow_sub_pow_le_mul[complete] -
norm_sub_taylorSum_le[complete] -
MeasureTheory.tendsto_setIntegral_norm_abs_gt[complete] -
MeasureTheory.tendsto_integral_norm_sub_comp_sub_right[complete] -
MeasureTheory.tendsto_integral_weight_norm_sub_comp_sub_right[complete] -
MeasureTheory.integrable_weight_norm_comp_sub[complete] -
MeasureTheory.integrable_weight_norm_sub_comp_sub[complete] -
MeasureTheory.continuous_integral_weight_norm_sub[complete] -
MeasureTheory.integral_comp_sub_div_smul[complete] -
MeasureTheory.tendsto_integral_weight_norm_smoothing_sub[complete] -
MeasureTheory.tendsto_integral_mul_smoothing_of_vanishing_moments[complete]
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem abs_pow_sub_pow_le_mul (n : ℕ) {x y M : ℝ} (hx : 0 ≤ x) (hy : 0 ≤ y) (hxM : x ≤ M) (hyM : y ≤ M) : |x ^ n - y ^ n| ≤ ↑n * M ^ (n - 1) * |x - y|
theorem abs_pow_sub_pow_le_mul (n : ℕ) {x y M : ℝ} (hx : 0 ≤ x) (hy : 0 ≤ y) (hxM : x ≤ M) (hyM : y ≤ M) : |x ^ n - y ^ n| ≤ ↑n * M ^ (n - 1) * |x - y|
Lipschitz bound for powers on `[0, M]`: `|x^n - y^n| ≤ n M^(n-1) |x - y|`.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem norm_sub_taylorSum_le.{u_1} {E : Type u_1} [NormedAddCommGroup E] [NormedSpace ℝ E] {f : ℝ → E} {k : ℕ} {C a h : ℝ} (hf : ContDiff ℝ (↑(k + 1)) f) (hC : ∀ y ∈ Set.uIcc a (a + h), ‖iteratedDeriv (k + 1) f y‖ ≤ C) : ‖f (a + h) - ∑ j ∈ Finset.range (k + 1), (h ^ j / ↑j.factorial) • iteratedDeriv j f a‖ ≤ C * |h| ^ (k + 1) / ↑k.factorial
theorem norm_sub_taylorSum_le.{u_1} {E : Type u_1} [NormedAddCommGroup E] [NormedSpace ℝ E] {f : ℝ → E} {k : ℕ} {C a h : ℝ} (hf : ContDiff ℝ (↑(k + 1)) f) (hC : ∀ y ∈ Set.uIcc a (a + h), ‖iteratedDeriv (k + 1) f y‖ ≤ C) : ‖f (a + h) - ∑ j ∈ Finset.range (k + 1), (h ^ j / ↑j.factorial) • iteratedDeriv j f a‖ ≤ C * |h| ^ (k + 1) / ↑k.factorial
**Uniform Taylor remainder bound**: for a `C^{k+1}` function whose `(k+1)`-st derivative is bounded by `C` on the segment from `a` to `a + h`, `‖f (a + h) - ∑_{j ≤ k} (h^j/j!) f^{(j)}(a)‖ ≤ C |h|^{k+1} / k!`. -
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.tendsto_setIntegral_norm_abs_gt.{u_1} {E : Type u_1} [NormedAddCommGroup E] {g : ℝ → E} (hg : MeasureTheory.Integrable g MeasureTheory.volume) : Filter.Tendsto (fun c ↦ ∫ (r : ℝ) in {r | c < |r|}, ‖g r‖) Filter.atTop (nhds 0)
theorem MeasureTheory.tendsto_setIntegral_norm_abs_gt.{u_1} {E : Type u_1} [NormedAddCommGroup E] {g : ℝ → E} (hg : MeasureTheory.Integrable g MeasureTheory.volume) : Filter.Tendsto (fun c ↦ ∫ (r : ℝ) in {r | c < |r|}, ‖g r‖) Filter.atTop (nhds 0)
The tail of a finite integral vanishes: for integrable `g`, `∫_{|r| > c} ‖g r‖ → 0` as `c → ∞`. -
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.tendsto_integral_norm_sub_comp_sub_right.{u_1} {E : Type u_1} [NormedAddCommGroup E] {Ξ : ℝ → E} (hΞ : MeasureTheory.Integrable Ξ MeasureTheory.volume) : Filter.Tendsto (fun s ↦ ∫ (z : ℝ), ‖Ξ (z - s) - Ξ z‖) (nhds 0) (nhds 0)
theorem MeasureTheory.tendsto_integral_norm_sub_comp_sub_right.{u_1} {E : Type u_1} [NormedAddCommGroup E] {Ξ : ℝ → E} (hΞ : MeasureTheory.Integrable Ξ MeasureTheory.volume) : Filter.Tendsto (fun s ↦ ∫ (z : ℝ), ‖Ξ (z - s) - Ξ z‖) (nhds 0) (nhds 0)
**Continuity of translation in `L¹(ℝ)`**: for integrable `Ξ`, `∫ ‖Ξ (z - s) - Ξ z‖ dz → 0` as `s → 0`.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.tendsto_integral_weight_norm_sub_comp_sub_right.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) {Ξ : ℝ → E} [NormedSpace ℝ E] (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) : Filter.Tendsto (fun s ↦ ∫ (z : ℝ), (1 + |z|) ^ k * ‖Ξ (z - s) - Ξ z‖) (nhds 0) (nhds 0)
theorem MeasureTheory.tendsto_integral_weight_norm_sub_comp_sub_right.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) {Ξ : ℝ → E} [NormedSpace ℝ E] (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) : Filter.Tendsto (fun s ↦ ∫ (z : ℝ), (1 + |z|) ^ k * ‖Ξ (z - s) - Ξ z‖) (nhds 0) (nhds 0)
**Continuity of translation in polynomially weighted `L¹(ℝ)`**: if `(1 + |z|)^k ‖Ξ z‖` is integrable, then `∫ (1 + |z|)^k ‖Ξ (z - s) - Ξ z‖ dz → 0` as `s → 0`.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.integrable_weight_norm_comp_sub.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) {Ξ : ℝ → E} (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (s : ℝ) : MeasureTheory.Integrable (fun z ↦ (1 + |z|) ^ k * ‖Ξ (z - s)‖) MeasureTheory.volume
theorem MeasureTheory.integrable_weight_norm_comp_sub.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) {Ξ : ℝ → E} (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (s : ℝ) : MeasureTheory.Integrable (fun z ↦ (1 + |z|) ^ k * ‖Ξ (z - s)‖) MeasureTheory.volume
Weighted integrability of translates.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.integrable_weight_norm_sub_comp_sub.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) {Ξ : ℝ → E} (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (s : ℝ) : MeasureTheory.Integrable (fun z ↦ (1 + |z|) ^ k * ‖Ξ (z - s) - Ξ z‖) MeasureTheory.volume
theorem MeasureTheory.integrable_weight_norm_sub_comp_sub.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) {Ξ : ℝ → E} (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (s : ℝ) : MeasureTheory.Integrable (fun z ↦ (1 + |z|) ^ k * ‖Ξ (z - s) - Ξ z‖) MeasureTheory.volume
Weighted integrability of translation differences.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.continuous_integral_weight_norm_sub.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) [NormedSpace ℝ E] {Ξ : ℝ → E} (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) : Continuous fun s ↦ ∫ (z : ℝ), (1 + |z|) ^ k * ‖Ξ (z - s) - Ξ z‖
theorem MeasureTheory.continuous_integral_weight_norm_sub.{u_1} {E : Type u_1} [NormedAddCommGroup E] (k : ℕ) [NormedSpace ℝ E] {Ξ : ℝ → E} (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) : Continuous fun s ↦ ∫ (z : ℝ), (1 + |z|) ^ k * ‖Ξ (z - s) - Ξ z‖
The polynomially weighted translation modulus is continuous.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.integral_comp_sub_div_smul.{u_2} {E : Type u_2} [NormedAddCommGroup E] [NormedSpace ℝ E] (F : ℝ → E) (r : ℝ) {c : ℝ} (hc : 0 < c) : ∫ (w : ℝ), F ((w - r) / c) = c • ∫ (v : ℝ), F v
theorem MeasureTheory.integral_comp_sub_div_smul.{u_2} {E : Type u_2} [NormedAddCommGroup E] [NormedSpace ℝ E] (F : ℝ → E) (r : ℝ) {c : ℝ} (hc : 0 < c) : ∫ (w : ℝ), F ((w - r) / c) = c • ∫ (v : ℝ), F v
Change of variables `w ↦ (w - r) / c` for vector-valued integrals on the line.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.tendsto_integral_weight_norm_smoothing_sub (k : ℕ) {Ξ K : ℝ → ℂ} (hΞm : MeasureTheory.StronglyMeasurable Ξ) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (hKm : MeasureTheory.AEStronglyMeasurable K MeasureTheory.volume) (hKk : MeasureTheory.Integrable (fun u ↦ (1 + |u|) ^ k * ‖K u‖) MeasureTheory.volume) (hK1 : ∫ (u : ℝ), K u = 1) : Filter.Tendsto (fun c ↦ ∫ (z : ℝ), (1 + |z|) ^ k * ‖(∫ (u : ℝ), K u * Ξ (z - u / c)) - Ξ z‖) Filter.atTop (nhds 0)
theorem MeasureTheory.tendsto_integral_weight_norm_smoothing_sub (k : ℕ) {Ξ K : ℝ → ℂ} (hΞm : MeasureTheory.StronglyMeasurable Ξ) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (hKm : MeasureTheory.AEStronglyMeasurable K MeasureTheory.volume) (hKk : MeasureTheory.Integrable (fun u ↦ (1 + |u|) ^ k * ‖K u‖) MeasureTheory.volume) (hK1 : ∫ (u : ℝ), K u = 1) : Filter.Tendsto (fun c ↦ ∫ (z : ℝ), (1 + |z|) ^ k * ‖(∫ (u : ℝ), K u * Ξ (z - u / c)) - Ξ z‖) Filter.atTop (nhds 0)
**Scaled approximate identity in polynomially weighted `L¹`**: for a kernel `K` with unit integral and finite `k`-th moment, the scaled smoothings `∫ K(u) Ξ(z - u/c) du` converge to `Ξ` in the polynomially weighted `L¹` norm as the scale `c → ∞`.
-
theoremdefined in LeanRidgelet/ToMathlib/WeightedL1Smoothing.leancomplete
theorem MeasureTheory.tendsto_integral_mul_smoothing_of_vanishing_moments (k : ℕ) {η Ξ : ℝ → ℂ} {Cη : ℝ} (θ : SchwartzMap ℝ ℂ) (hηm : MeasureTheory.StronglyMeasurable η) (hηk : ∀ (z : ℝ), ‖η z‖ ≤ Cη * (1 + |z|) ^ k) (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (hΞvm : ∀ j ≤ k, ∫ (r : ℝ), ↑r ^ j * Ξ r = 0) : Filter.Tendsto (fun c ↦ ∫ (r : ℝ), Ξ r * (↑c⁻¹ * ∫ (w : ℝ), η w * θ ((w - r) / c))) Filter.atTop (nhds 0)
theorem MeasureTheory.tendsto_integral_mul_smoothing_of_vanishing_moments (k : ℕ) {η Ξ : ℝ → ℂ} {Cη : ℝ} (θ : SchwartzMap ℝ ℂ) (hηm : MeasureTheory.StronglyMeasurable η) (hηk : ∀ (z : ℝ), ‖η z‖ ≤ Cη * (1 + |z|) ^ k) (hΞm : MeasureTheory.AEStronglyMeasurable Ξ MeasureTheory.volume) (hΞk : MeasureTheory.Integrable (fun r ↦ (1 + |r|) ^ k * ‖Ξ r‖) MeasureTheory.volume) (hΞvm : ∀ j ≤ k, ∫ (r : ℝ), ↑r ^ j * Ξ r = 0) : Filter.Tendsto (fun c ↦ ∫ (r : ℝ), Ξ r * (↑c⁻¹ * ∫ (w : ℝ), η w * θ ((w - r) / c))) Filter.atTop (nhds 0)
**Vanishing-moment cancellation for wide smoothing**: if `Ξ` has a finite `k`-th moment and vanishing moments up to `k`, and `η` is strongly measurable with growth at most `(1 + |z|)^k`, then the pairing of `Ξ` with the `c`-scale smoothing of `η` by a Schwartz kernel `θ` tends to zero as the scale `c → ∞`.
Schwartz space
-
IsLizorkin[complete] -
LizorkinSpace[complete] -
mem_lizorkinSpace_iff[complete] -
integrable_pow_smul_schwartz[complete] -
Real.fourier_apply_zero[complete] -
Real.iteratedDeriv_fourier_zero[complete] -
mem_lizorkinSpace_iff_fourier_flat[complete] -
integral_polynomial_mul_eq_zero_of_mem_lizorkinSpace[complete] -
LizorkinDistribution[complete]
The Lizorkin space and Lizorkin distributions. Mathlib has the Schwartz space and its Fourier theory but not the Lizorkin space \mathcal S_0(\mathbb R), the submodule of Schwartz functions all of whose moments vanish. It is defined here, together with the identity moments = derivatives of the Fourier transform at the origin, whence the Fourier characterization: a Schwartz function is Lizorkin exactly when its Fourier transform vanishes to infinite order at the origin. Polynomials annihilate it, the precise form of the statement that a point mass \delta^{(j)} at the origin is invisible. The Lizorkin distributions \mathcal S_0'(\mathbb R)=\mathcal S_0(\mathbb R)\to_L\mathbb C are tempered distributions modulo polynomials.
Lean code for Definition5.4.2●9 declarations
Associated Lean declarations
-
IsLizorkin[complete]
-
LizorkinSpace[complete]
-
mem_lizorkinSpace_iff[complete]
-
integrable_pow_smul_schwartz[complete]
-
Real.fourier_apply_zero[complete]
-
Real.iteratedDeriv_fourier_zero[complete]
-
mem_lizorkinSpace_iff_fourier_flat[complete]
-
integral_polynomial_mul_eq_zero_of_mem_lizorkinSpace[complete]
-
LizorkinDistribution[complete]
-
IsLizorkin[complete] -
LizorkinSpace[complete] -
mem_lizorkinSpace_iff[complete] -
integrable_pow_smul_schwartz[complete] -
Real.fourier_apply_zero[complete] -
Real.iteratedDeriv_fourier_zero[complete] -
mem_lizorkinSpace_iff_fourier_flat[complete] -
integral_polynomial_mul_eq_zero_of_mem_lizorkinSpace[complete] -
LizorkinDistribution[complete]
-
defdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
def IsLizorkin (φ : SchwartzMap ℝ ℂ) : Prop
def IsLizorkin (φ : SchwartzMap ℝ ℂ) : Prop
Implementation after
:=:= ∀ n : ℕ, (∫ z : ℝ, (z : ℂ) ^ n * φ z) = 0
A Schwartz function is a **Lizorkin test function** when all of its moments vanish.
-
defdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
def LizorkinSpace : Submodule ℂ (SchwartzMap ℝ ℂ)
def LizorkinSpace : Submodule ℂ (SchwartzMap ℝ ℂ)
Implementation after
:=:= {φ | IsLizorkin φ} add_mem' := by intro φ ψ hφ hψ n have h : (∫ z : ℝ, (z : ℂ) ^ n * (φ + ψ) z) = (∫ z : ℝ, (z : ℂ) ^ n * φ z) + ∫ z : ℝ, (z : ℂ) ^ n * ψ z := by rw [← integral_add (integrable_pow_smul_schwartz φ n) (integrable_pow_smul_schwartz ψ n)] refine integral_congr_ae (Filter.Eventually.of_forall fun z => ?_) simp only [add_apply] ring rw [h, hφ n, hψ n, add_zero] zero_mem' := by intro n simp smul_mem' := by intro c φ hφ n have h : (∫ z : ℝ, (z : ℂ) ^ n * (c • φ) z) = c * ∫ z : ℝ, (z : ℂ) ^ n * φ z := by rw [← integral_const_mul] refine integral_congr_ae (Filter.Eventually.of_forall fun z => ?_) simp only [smul_apply, smul_eq_mul] ring rw [h, hφ n, mul_zero]The **Lizorkin space** `𝒮₀(ℝ)`: the Schwartz functions all of whose moments vanish. Its annihilator contains every polynomial, which is why the Lizorkin dual `𝒮₀'(ℝ)` sees tempered distributions only modulo polynomials.
-
theoremdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
theorem mem_lizorkinSpace_iff {φ : SchwartzMap ℝ ℂ} : φ ∈ LizorkinSpace ↔ ∀ (n : ℕ), ∫ (z : ℝ), ↑z ^ n * φ z = 0
theorem mem_lizorkinSpace_iff {φ : SchwartzMap ℝ ℂ} : φ ∈ LizorkinSpace ↔ ∀ (n : ℕ), ∫ (z : ℝ), ↑z ^ n * φ z = 0
-
theoremdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
theorem integrable_pow_smul_schwartz (φ : SchwartzMap ℝ ℂ) (n : ℕ) : MeasureTheory.Integrable (fun z ↦ ↑z ^ n * φ z) MeasureTheory.volume
theorem integrable_pow_smul_schwartz (φ : SchwartzMap ℝ ℂ) (n : ℕ) : MeasureTheory.Integrable (fun z ↦ ↑z ^ n * φ z) MeasureTheory.volume
Integrability of the moment integrands of a Schwartz function.
-
theoremdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
theorem Real.fourier_apply_zero (g : ℝ → ℂ) : FourierTransform.fourier g 0 = ∫ (z : ℝ), g z
theorem Real.fourier_apply_zero (g : ℝ → ℂ) : FourierTransform.fourier g 0 = ∫ (z : ℝ), g z
The Fourier transform at the origin is the total integral.
-
theoremdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
theorem Real.iteratedDeriv_fourier_zero (φ : SchwartzMap ℝ ℂ) (n : ℕ) : iteratedDeriv n (FourierTransform.fourier ⇑φ) 0 = (-2 * ↑Real.pi * Complex.I) ^ n * ∫ (z : ℝ), ↑z ^ n * φ z
theorem Real.iteratedDeriv_fourier_zero (φ : SchwartzMap ℝ ℂ) (n : ℕ) : iteratedDeriv n (FourierTransform.fourier ⇑φ) 0 = (-2 * ↑Real.pi * Complex.I) ^ n * ∫ (z : ℝ), ↑z ^ n * φ z
The moments of a Schwartz function are the derivatives of its Fourier transform at the origin.
-
theoremdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
theorem mem_lizorkinSpace_iff_fourier_flat {φ : SchwartzMap ℝ ℂ} : φ ∈ LizorkinSpace ↔ ∀ (n : ℕ), iteratedDeriv n (FourierTransform.fourier ⇑φ) 0 = 0
theorem mem_lizorkinSpace_iff_fourier_flat {φ : SchwartzMap ℝ ℂ} : φ ∈ LizorkinSpace ↔ ∀ (n : ℕ), iteratedDeriv n (FourierTransform.fourier ⇑φ) 0 = 0
**Fourier characterization of the Lizorkin space**: a Schwartz function has all moments vanishing exactly when its Fourier transform vanishes to infinite order at the origin.
-
theoremdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
theorem integral_polynomial_mul_eq_zero_of_mem_lizorkinSpace {φ : SchwartzMap ℝ ℂ} (hφ : φ ∈ LizorkinSpace) (Q : Polynomial ℂ) : ∫ (z : ℝ), Polynomial.eval (↑z) Q * φ z = 0
theorem integral_polynomial_mul_eq_zero_of_mem_lizorkinSpace {φ : SchwartzMap ℝ ℂ} (hφ : φ ∈ LizorkinSpace) (Q : Polynomial ℂ) : ∫ (z : ℝ), Polynomial.eval (↑z) Q * φ z = 0
**Polynomials annihilate the Lizorkin space.** This is the precise form of the statement that a point mass `δ^{(j)}` at the origin — equivalently a polynomial summand on the space side — is invisible to a Lizorkin distribution. -
abbrevdefined in LeanRidgelet/ToMathlib/Lizorkin.leancomplete
abbrev LizorkinDistribution : Type
abbrev LizorkinDistribution : Type
Implementation after
:=:= LizorkinSpace →L[ℂ] ℂ
The **Lizorkin distributions** `𝒮₀'(ℝ)`: the continuous linear functionals on the Lizorkin space. Restricting a tempered distribution to `𝒮₀(ℝ)` forgets exactly its polynomial part, by `integral_polynomial_mul_eq_zero_of_mem_lizorkinSpace`.
-
MeasureTheory.PolynomiallyBounded[complete] -
MeasureTheory.polynomiallyBounded_nonneg_const[complete] -
MeasureTheory.PolynomiallyBounded.add[complete] -
MeasureTheory.polynomiallyBounded_polynomial_eval[complete] -
MeasureTheory.PolynomiallyBounded.integrable_mul_schwartz[complete] -
one_add_abs_add_le_mul[complete] -
MeasureTheory.integrable_one_add_abs_pow_mul_schwartz[complete] -
MeasureTheory.integral_pow_mul_conj_comp_sub_eq_zero[complete]
Polynomially bounded functions. Mathlib has Function.HasTemperateGrowth, which additionally demands smoothness with polynomially bounded derivatives of every order, but no bare growth predicate \|f(z)\|\le C(1+|z|)^k for merely measurable functions — the classical criterion, together with local integrability, for defining a tempered distribution. It is introduced here with its closure under addition, its validity for polynomials, and the resulting integrability against every Schwartz function. Alongside: submultiplicativity 1+|z+t|\le(1+|z|)(1+|t|) of the weight, weighted integrability (1+|s|)^k\|\varphi(s)\|\in L^1 of Schwartz functions, and the invariance of vanishing moments under shift and conjugation, \int r^j\overline{\psi(r-c)}\,dr=0 for j\le k.
Lean code for Theorem5.4.3●8 declarations
Associated Lean declarations
-
MeasureTheory.PolynomiallyBounded[complete]
-
MeasureTheory.polynomiallyBounded_nonneg_const[complete]
-
MeasureTheory.PolynomiallyBounded.add[complete]
-
MeasureTheory.polynomiallyBounded_polynomial_eval[complete]
-
MeasureTheory.PolynomiallyBounded.integrable_mul_schwartz[complete]
-
one_add_abs_add_le_mul[complete]
-
MeasureTheory.integrable_one_add_abs_pow_mul_schwartz[complete]
-
MeasureTheory.integral_pow_mul_conj_comp_sub_eq_zero[complete]
-
MeasureTheory.PolynomiallyBounded[complete] -
MeasureTheory.polynomiallyBounded_nonneg_const[complete] -
MeasureTheory.PolynomiallyBounded.add[complete] -
MeasureTheory.polynomiallyBounded_polynomial_eval[complete] -
MeasureTheory.PolynomiallyBounded.integrable_mul_schwartz[complete] -
one_add_abs_add_le_mul[complete] -
MeasureTheory.integrable_one_add_abs_pow_mul_schwartz[complete] -
MeasureTheory.integral_pow_mul_conj_comp_sub_eq_zero[complete]
-
defdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
def MeasureTheory.PolynomiallyBounded (η : ℝ → ℂ) : Prop
def MeasureTheory.PolynomiallyBounded (η : ℝ → ℂ) : Prop
Implementation after
:=:= ∃ (C : ℝ) (k : ℕ), ∀ z : ℝ, ‖η z‖ ≤ C * (1 + |z|) ^ k
Polynomial growth bound: `‖η z‖ ≤ C (1 + |z|)^k`. Together with local integrability this is the classical criterion for a function to define a tempered distribution.
-
theoremdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
theorem MeasureTheory.polynomiallyBounded_nonneg_const {η : ℝ → ℂ} {C : ℝ} {k : ℕ} (h : ∀ (z : ℝ), ‖η z‖ ≤ C * (1 + |z|) ^ k) : 0 ≤ C
theorem MeasureTheory.polynomiallyBounded_nonneg_const {η : ℝ → ℂ} {C : ℝ} {k : ℕ} (h : ∀ (z : ℝ), ‖η z‖ ≤ C * (1 + |z|) ^ k) : 0 ≤ C
-
theoremdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
theorem MeasureTheory.PolynomiallyBounded.add {η₁ η₂ : ℝ → ℂ} (h₁ : MeasureTheory.PolynomiallyBounded η₁) (h₂ : MeasureTheory.PolynomiallyBounded η₂) : MeasureTheory.PolynomiallyBounded fun z ↦ η₁ z + η₂ z
theorem MeasureTheory.PolynomiallyBounded.add {η₁ η₂ : ℝ → ℂ} (h₁ : MeasureTheory.PolynomiallyBounded η₁) (h₂ : MeasureTheory.PolynomiallyBounded η₂) : MeasureTheory.PolynomiallyBounded fun z ↦ η₁ z + η₂ z
Polynomial growth bounds are stable under addition.
-
theoremdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
theorem MeasureTheory.polynomiallyBounded_polynomial_eval (Q : Polynomial ℂ) : MeasureTheory.PolynomiallyBounded fun z ↦ Polynomial.eval (↑z) Q
theorem MeasureTheory.polynomiallyBounded_polynomial_eval (Q : Polynomial ℂ) : MeasureTheory.PolynomiallyBounded fun z ↦ Polynomial.eval (↑z) Q
Every polynomial, evaluated along the real line, has polynomial growth.
-
theoremdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
theorem MeasureTheory.PolynomiallyBounded.integrable_mul_schwartz {η : ℝ → ℂ} (hb : MeasureTheory.PolynomiallyBounded η) (hm : MeasureTheory.AEStronglyMeasurable η MeasureTheory.volume) (φ : SchwartzMap ℝ ℂ) : MeasureTheory.Integrable (fun z ↦ η z * φ z) MeasureTheory.volume
theorem MeasureTheory.PolynomiallyBounded.integrable_mul_schwartz {η : ℝ → ℂ} (hb : MeasureTheory.PolynomiallyBounded η) (hm : MeasureTheory.AEStronglyMeasurable η MeasureTheory.volume) (φ : SchwartzMap ℝ ℂ) : MeasureTheory.Integrable (fun z ↦ η z * φ z) MeasureTheory.volume
A polynomially bounded measurable function is integrable against any Schwartz function.
-
theoremdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
theorem one_add_abs_add_le_mul (z t : ℝ) : 1 + |z + t| ≤ (1 + |z|) * (1 + |t|)
theorem one_add_abs_add_le_mul (z t : ℝ) : 1 + |z + t| ≤ (1 + |z|) * (1 + |t|)
Submultiplicativity of the polynomial weight: `1 + |z + t| ≤ (1 + |z|) (1 + |t|)`.
-
theoremdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
theorem MeasureTheory.integrable_one_add_abs_pow_mul_schwartz (φ : SchwartzMap ℝ ℂ) (k : ℕ) : MeasureTheory.Integrable (fun s ↦ (1 + |s|) ^ k * ‖φ s‖) MeasureTheory.volume
theorem MeasureTheory.integrable_one_add_abs_pow_mul_schwartz (φ : SchwartzMap ℝ ℂ) (k : ℕ) : MeasureTheory.Integrable (fun s ↦ (1 + |s|) ^ k * ‖φ s‖) MeasureTheory.volume
Weighted integrability of a one-dimensional Schwartz function.
-
theoremdefined in LeanRidgelet/ToMathlib/PolynomialGrowth.leancomplete
theorem MeasureTheory.integral_pow_mul_conj_comp_sub_eq_zero {ψ : ℝ → ℂ} {k : ℕ} (hψm : MeasureTheory.AEStronglyMeasurable ψ MeasureTheory.volume) (hψk : MeasureTheory.Integrable (fun s ↦ (1 + |s|) ^ k * ‖ψ s‖) MeasureTheory.volume) (hψvm : ∀ j ≤ k, ∫ (s : ℝ), ↑s ^ j * ψ s = 0) {j : ℕ} (hj : j ≤ k) (c : ℝ) : ∫ (r : ℝ), ↑r ^ j * (starRingEnd ℂ) (ψ (r - c)) = 0
theorem MeasureTheory.integral_pow_mul_conj_comp_sub_eq_zero {ψ : ℝ → ℂ} {k : ℕ} (hψm : MeasureTheory.AEStronglyMeasurable ψ MeasureTheory.volume) (hψk : MeasureTheory.Integrable (fun s ↦ (1 + |s|) ^ k * ‖ψ s‖) MeasureTheory.volume) (hψvm : ∀ j ≤ k, ∫ (s : ℝ), ↑s ^ j * ψ s = 0) {j : ℕ} (hj : j ≤ k) (c : ℝ) : ∫ (r : ℝ), ↑r ^ j * (starRingEnd ℂ) (ψ (r - c)) = 0
Vanishing moments are inherited by shifted conjugate sections: if `ψ` has finite `k`-th moment and vanishing moments up to `k`, then `∫ r^j conj (ψ (r - c)) dr = 0` for every `j ≤ k` and every shift `c`. This is the mechanism by which the polynomial part of the function — the kernel of the Lizorkin quotient — becomes invisible to a pairing against a function with vanishing moments.
-
coe_iterate_schwartz_derivCLM[complete] -
iteratedDeriv_eq_zero_of_tsupport_subset_compl[complete] -
exists_pos_le_abs_of_tsupport_subset[complete] -
MeasureTheory.integrable_mul_of_bound_on_tsupport[complete] -
hasTemperateGrowth_const_mul[complete] -
antilipschitzWith_const_mul[complete]
Auxiliary lemmas on Schwartz functions and supports. Iterating SchwartzMap.derivCLM computes iteratedDeriv; all derivatives of a Schwartz function supported away from a point vanish at that point; a function supported away from the origin is uniformly separated from it; multiplying an integrable function by a factor bounded on its support preserves integrability; and the scaling \zeta\mapsto c\zeta is of temperate growth and, for c\ne0, antilipschitz — the two hypotheses of SchwartzMap.compCLMOfAntilipschitz, which is how a Fourier rescaling is recognized as an endomorphism of the Schwartz space.
Lean code for Theorem5.4.4●6 theorems
Associated Lean declarations
-
coe_iterate_schwartz_derivCLM[complete]
-
iteratedDeriv_eq_zero_of_tsupport_subset_compl[complete]
-
exists_pos_le_abs_of_tsupport_subset[complete]
-
MeasureTheory.integrable_mul_of_bound_on_tsupport[complete]
-
hasTemperateGrowth_const_mul[complete]
-
antilipschitzWith_const_mul[complete]
-
coe_iterate_schwartz_derivCLM[complete] -
iteratedDeriv_eq_zero_of_tsupport_subset_compl[complete] -
exists_pos_le_abs_of_tsupport_subset[complete] -
MeasureTheory.integrable_mul_of_bound_on_tsupport[complete] -
hasTemperateGrowth_const_mul[complete] -
antilipschitzWith_const_mul[complete]
-
theoremdefined in LeanRidgelet/ToMathlib/SchwartzAux.leancomplete
theorem coe_iterate_schwartz_derivCLM (n : ℕ) (φ : SchwartzMap ℝ ℂ) : ⇑((⇑(SchwartzMap.derivCLM ℂ ℂ))^[n] φ) = iteratedDeriv n ⇑φ
theorem coe_iterate_schwartz_derivCLM (n : ℕ) (φ : SchwartzMap ℝ ℂ) : ⇑((⇑(SchwartzMap.derivCLM ℂ ℂ))^[n] φ) = iteratedDeriv n ⇑φ
The coercion of an iterated Schwartz derivative is the iterated derivative.
-
theoremdefined in LeanRidgelet/ToMathlib/SchwartzAux.leancomplete
theorem iteratedDeriv_eq_zero_of_tsupport_subset_compl (φ : SchwartzMap ℝ ℂ) (hφ : tsupport ⇑φ ⊆ {0}ᶜ) (n : ℕ) : iteratedDeriv n (⇑φ) 0 = 0
theorem iteratedDeriv_eq_zero_of_tsupport_subset_compl (φ : SchwartzMap ℝ ℂ) (hφ : tsupport ⇑φ ⊆ {0}ᶜ) (n : ℕ) : iteratedDeriv n (⇑φ) 0 = 0
A Schwartz function supported away from the origin has all iterated derivatives vanishing at the origin.
-
theoremdefined in LeanRidgelet/ToMathlib/SchwartzAux.leancomplete
theorem exists_pos_le_abs_of_tsupport_subset {g : ℝ → ℂ} (hg : tsupport g ⊆ {0}ᶜ) : ∃ δ > 0, ∀ ζ ∈ tsupport g, δ ≤ |ζ|
theorem exists_pos_le_abs_of_tsupport_subset {g : ℝ → ℂ} (hg : tsupport g ⊆ {0}ᶜ) : ∃ δ > 0, ∀ ζ ∈ tsupport g, δ ≤ |ζ|
A test function supported away from the origin is uniformly separated from it.
-
theoremdefined in LeanRidgelet/ToMathlib/SchwartzAux.leancomplete
theorem MeasureTheory.integrable_mul_of_bound_on_tsupport {c g : ℝ → ℂ} {C : ℝ} (hc : MeasureTheory.AEStronglyMeasurable c MeasureTheory.volume) (hg : MeasureTheory.Integrable g MeasureTheory.volume) (hbound : ∀ ζ ∈ tsupport g, ‖c ζ‖ ≤ C) : MeasureTheory.Integrable (fun ζ ↦ c ζ * g ζ) MeasureTheory.volume
theorem MeasureTheory.integrable_mul_of_bound_on_tsupport {c g : ℝ → ℂ} {C : ℝ} (hc : MeasureTheory.AEStronglyMeasurable c MeasureTheory.volume) (hg : MeasureTheory.Integrable g MeasureTheory.volume) (hbound : ∀ ζ ∈ tsupport g, ‖c ζ‖ ≤ C) : MeasureTheory.Integrable (fun ζ ↦ c ζ * g ζ) MeasureTheory.volume
Multiplying an integrable function by a factor bounded on its support preserves integrability.
-
theoremdefined in LeanRidgelet/ToMathlib/SchwartzAux.leancomplete
theorem hasTemperateGrowth_const_mul (c : ℝ) : Function.HasTemperateGrowth fun ζ ↦ c * ζ
theorem hasTemperateGrowth_const_mul (c : ℝ) : Function.HasTemperateGrowth fun ζ ↦ c * ζ
-
theoremdefined in LeanRidgelet/ToMathlib/SchwartzAux.leancomplete
theorem antilipschitzWith_const_mul {c : ℝ} (hc : c ≠ 0) : AntilipschitzWith (Real.nnabs c⁻¹) fun ζ ↦ c * ζ
theorem antilipschitzWith_const_mul {c : ℝ} (hc : c ≠ 0) : AntilipschitzWith (Real.nnabs c⁻¹) fun ζ ↦ c * ζ
Convolution
-
MeasureTheory.enorm_integral_mul_sub_le[complete] -
MeasureTheory.lintegral_lintegral_mul_comp_sub[complete] -
MeasureTheory.eLpNorm_convolution_le[complete] -
MeasureTheory.Integrable.convolution_memLp[complete]
Young's convolution inequality L¹ ⋆ L^p ⊆ L^p. On a measurable additive commutative group with an invariant σ-finite measure, \|f\ast g\|_{L^p}\le\|f\|_{L^1}\|g\|_{L^p} for 1\le p\le\infty with no finiteness hypotheses, and the convolution of an integrable function with an L^p function is in L^p. The proof avoids the (also missing) continuous Minkowski integral inequality through the Hölder splitting \|f(t)\|\,\|g(x-t)\|=\|f(t)\|^{1/q}\cdot(\|f(t)\|^{1/p}\|g(x-t)\|), the two-function Hölder inequality ENNReal.lintegral_mul_le_Lp_mul_Lq, Tonelli's theorem, and translation invariance.
Lean code for Theorem5.4.5●4 theorems
Associated Lean declarations
-
MeasureTheory.enorm_integral_mul_sub_le[complete]
-
MeasureTheory.lintegral_lintegral_mul_comp_sub[complete]
-
MeasureTheory.eLpNorm_convolution_le[complete]
-
MeasureTheory.Integrable.convolution_memLp[complete]
-
MeasureTheory.enorm_integral_mul_sub_le[complete] -
MeasureTheory.lintegral_lintegral_mul_comp_sub[complete] -
MeasureTheory.eLpNorm_convolution_le[complete] -
MeasureTheory.Integrable.convolution_memLp[complete]
-
theoremdefined in LeanRidgelet/ToMathlib/YoungConvolution.leancomplete
theorem MeasureTheory.enorm_integral_mul_sub_le.{u_1} {G : Type u_1} [MeasurableSpace G] [Sub G] {μ : MeasureTheory.Measure G} (f g : G → ℂ) (x : G) : ‖∫ (t : G), f t * g (x - t) ∂μ‖ₑ ≤ ∫⁻ (t : G), ‖f t‖ₑ * ‖g (x - t)‖ₑ ∂μ
theorem MeasureTheory.enorm_integral_mul_sub_le.{u_1} {G : Type u_1} [MeasurableSpace G] [Sub G] {μ : MeasureTheory.Measure G} (f g : G → ℂ) (x : G) : ‖∫ (t : G), f t * g (x - t) ∂μ‖ₑ ≤ ∫⁻ (t : G), ‖f t‖ₑ * ‖g (x - t)‖ₑ ∂μ
The convolution integral is dominated by the `lintegral` of the pointwise product of extended norms.
-
theoremdefined in LeanRidgelet/ToMathlib/YoungConvolution.leancomplete
theorem MeasureTheory.lintegral_lintegral_mul_comp_sub.{u_1} {G : Type u_1} [MeasurableSpace G] [AddCommGroup G] [MeasurableAdd₂ G] [MeasurableNeg G] {μ : MeasureTheory.Measure G} [MeasureTheory.SFinite μ] [μ.IsAddRightInvariant] (F K : G → ENNReal) (hF : AEMeasurable F μ) (hK : AEMeasurable K μ) : ∫⁻ (x : G), ∫⁻ (t : G), F t * K (x - t) ∂μ ∂μ = (∫⁻ (t : G), F t ∂μ) * ∫⁻ (y : G), K y ∂μ
theorem MeasureTheory.lintegral_lintegral_mul_comp_sub.{u_1} {G : Type u_1} [MeasurableSpace G] [AddCommGroup G] [MeasurableAdd₂ G] [MeasurableNeg G] {μ : MeasureTheory.Measure G} [MeasureTheory.SFinite μ] [μ.IsAddRightInvariant] (F K : G → ENNReal) (hF : AEMeasurable F μ) (hK : AEMeasurable K μ) : ∫⁻ (x : G), ∫⁻ (t : G), F t * K (x - t) ∂μ ∂μ = (∫⁻ (t : G), F t ∂μ) * ∫⁻ (y : G), K y ∂μ
Tonelli's theorem combined with translation invariance: the iterated `lintegral` of a convolution-shaped product of nonnegative functions factors into the product of the total integrals.
-
theoremdefined in LeanRidgelet/ToMathlib/YoungConvolution.leancomplete
theorem MeasureTheory.eLpNorm_convolution_le.{u_1} {G : Type u_1} [MeasurableSpace G] [AddCommGroup G] [MeasurableAdd₂ G] [MeasurableNeg G] {μ : MeasureTheory.Measure G} [MeasureTheory.SFinite μ] [μ.IsAddLeftInvariant] [μ.IsAddRightInvariant] [μ.IsNegInvariant] {p : ENNReal} (hp : 1 ≤ p) {f g : G → ℂ} (hf : MeasureTheory.AEStronglyMeasurable f μ) (hg : MeasureTheory.AEStronglyMeasurable g μ) : MeasureTheory.eLpNorm (fun x ↦ ∫ (t : G), f t * g (x - t) ∂μ) p μ ≤ MeasureTheory.eLpNorm f 1 μ * MeasureTheory.eLpNorm g p μ
theorem MeasureTheory.eLpNorm_convolution_le.{u_1} {G : Type u_1} [MeasurableSpace G] [AddCommGroup G] [MeasurableAdd₂ G] [MeasurableNeg G] {μ : MeasureTheory.Measure G} [MeasureTheory.SFinite μ] [μ.IsAddLeftInvariant] [μ.IsAddRightInvariant] [μ.IsNegInvariant] {p : ENNReal} (hp : 1 ≤ p) {f g : G → ℂ} (hf : MeasureTheory.AEStronglyMeasurable f μ) (hg : MeasureTheory.AEStronglyMeasurable g μ) : MeasureTheory.eLpNorm (fun x ↦ ∫ (t : G), f t * g (x - t) ∂μ) p μ ≤ MeasureTheory.eLpNorm f 1 μ * MeasureTheory.eLpNorm g p μ
**Young's convolution inequality**, `L¹ ⋆ L^p` case, `eLpNorm` form: for `1 ≤ p`, `‖f ⋆ g‖_p ≤ ‖f‖₁ ‖g‖_p`. Neither side is assumed finite.
-
theoremdefined in LeanRidgelet/ToMathlib/YoungConvolution.leancomplete
theorem MeasureTheory.Integrable.convolution_memLp.{u_1} {G : Type u_1} [MeasurableSpace G] [AddCommGroup G] [MeasurableAdd₂ G] [MeasurableNeg G] {μ : MeasureTheory.Measure G} [MeasureTheory.SFinite μ] [μ.IsAddLeftInvariant] [μ.IsAddRightInvariant] [μ.IsNegInvariant] {p : ENNReal} (hp : 1 ≤ p) {f g : G → ℂ} (hf : MeasureTheory.Integrable f μ) (hg : MeasureTheory.MemLp g p μ) : MeasureTheory.MemLp (fun x ↦ ∫ (t : G), f t * g (x - t) ∂μ) p μ
theorem MeasureTheory.Integrable.convolution_memLp.{u_1} {G : Type u_1} [MeasurableSpace G] [AddCommGroup G] [MeasurableAdd₂ G] [MeasurableNeg G] {μ : MeasureTheory.Measure G} [MeasureTheory.SFinite μ] [μ.IsAddLeftInvariant] [μ.IsAddRightInvariant] [μ.IsNegInvariant] {p : ENNReal} (hp : 1 ≤ p) {f g : G → ℂ} (hf : MeasureTheory.Integrable f μ) (hg : MeasureTheory.MemLp g p μ) : MeasureTheory.MemLp (fun x ↦ ∫ (t : G), f t * g (x - t) ∂μ) p μ
**Young's convolution inequality**, membership form: the convolution of an integrable function with an `L^p` function lies in `L^p` for `1 ≤ p`.
-
Function.norm_iteratedFDerivWithin_rpow_le[complete] -
Function.HasTemperateGrowth.rpow_of_le[complete]
A real power of a temperate function bounded away from zero has temperate growth. Mathlib has this for the Bessel potential multiplier x\mapsto(1+\|x\|^2)^a, whose proof composes u\mapsto u^a — temperate on any ray bounded away from the origin — with the inner function 1+\|x\|^2, whose range is bounded away from the origin; its own comment observes that the argument works for any such ray. That generalization is proved here: if q has temperate growth and q\geq c for some c>0, so does q^a, for every real a.
Lean code for Theorem5.4.6●2 theorems
Associated Lean declarations
-
Function.norm_iteratedFDerivWithin_rpow_le[complete]
-
Function.HasTemperateGrowth.rpow_of_le[complete]
-
Function.norm_iteratedFDerivWithin_rpow_le[complete] -
Function.HasTemperateGrowth.rpow_of_le[complete]
-
theoremdefined in LeanRidgelet/ToMathlib/TemperateGrowth.leancomplete
theorem Function.norm_iteratedFDerivWithin_rpow_le {b : ℝ} (hb : 0 < b) (a : ℝ) (N : ℕ) : ∃ k C, 0 ≤ C ∧ ∀ n ≤ N, ∀ u ∈ Set.Ioi b, ‖iteratedFDerivWithin ℝ n (fun u ↦ u ^ a) (Set.Ioi b) u‖ ≤ C * (1 + ‖u‖) ^ k
theorem Function.norm_iteratedFDerivWithin_rpow_le {b : ℝ} (hb : 0 < b) (a : ℝ) (N : ℕ) : ∃ k C, 0 ≤ C ∧ ∀ n ≤ N, ∀ u ∈ Set.Ioi b, ‖iteratedFDerivWithin ℝ n (fun u ↦ u ^ a) (Set.Ioi b) u‖ ≤ C * (1 + ‖u‖) ^ k
The derivatives of `u ↦ u^a` are bounded by a polynomial on any ray `(b, ∞)` with `b > 0`. This is the estimate behind `Function.HasTemperateGrowth.rpow_of_le`, and the only analytic content of it: the `n`-th derivative is a constant times `u^{a-n}`, which on `u > b` is bounded by a fixed power of `1 + u` when `a - n ≥ 0` and by a constant when `a - n < 0`. -
theoremdefined in LeanRidgelet/ToMathlib/TemperateGrowth.leancomplete
theorem Function.HasTemperateGrowth.rpow_of_le.{u_1} {E : Type u_1} [NormedAddCommGroup E] [NormedSpace ℝ E] {q : E → ℝ} (hq : Function.HasTemperateGrowth q) {c : ℝ} (hc : 0 < c) (hqc : ∀ (x : E), c ≤ q x) (a : ℝ) : Function.HasTemperateGrowth fun x ↦ q x ^ a
theorem Function.HasTemperateGrowth.rpow_of_le.{u_1} {E : Type u_1} [NormedAddCommGroup E] [NormedSpace ℝ E] {q : E → ℝ} (hq : Function.HasTemperateGrowth q) {c : ℝ} (hc : 0 < c) (hqc : ∀ (x : E), c ≤ q x) (a : ℝ) : Function.HasTemperateGrowth fun x ↦ q x ^ a
**A real power of a temperate function bounded away from zero has temperate growth.** This generalizes `Function.hasTemperateGrowth_one_add_norm_sq_rpow`, which is the case `q = 1 + ‖·‖²` and `c = 1`, to an arbitrary inner function; the point of the generalization is that one may then *choose* the inner function, which is what a multiplier with a singularity at the origin needs.
What the generalization buys is the freedom to choose the inner function, which is what a multiplier with a singularity at the origin needs: \|x\|^a is not of temperate growth — for a non-even it is not smooth at the origin, and for a negative not even bounded there — but outside a ball it agrees with q^{a/2} for a q that is \|x\|^2 corrected by a bump, and that is temperate by the statement here. The correction goes inside the ball, where the multiplier is not needed, rather than on the multiplier itself.
The estimate is separated out as the first declaration, and it is the only analytic content: the n-th derivative of u^a is a descending Pochhammer factor times u^{a-n}, which on a ray u>b>0 is bounded by a fixed power of 1+u when a-n\geq0 and by a constant when a-n<0.