Documentation

Carleson.ToMathlib.MeasureTheory.Integral.MeanInequalities

theorem ENNReal.Lp_add_le_sum {ι : Type u_2} {κ : Type u_3} {s : Finset ι} {t : Finset κ} {f : ι → κ → ENNReal} {p : ℝ} (hp : 1 ≤ p) :
(∑ i ∈ s, (∑ j ∈ t, f i j) ^ p) ^ (1 / p) ≤ ∑ j ∈ t, (∑ i ∈ s, f i j ^ p) ^ (1 / p)

Minkowski inequality for finite sums of ENNReal-valued functions.

theorem ENNReal.lintegral_prod_norm_pow_le' {α : Type u_2} {ι : Type u_3} [MeasurableSpace α] {μ : MeasureTheory.Measure α} {s : Finset ι} {f : ι → α → ENNReal} (hf : ∀ i ∈ s, AEMeasurable (f i) μ) {p : ι → ENNReal} (hp : ∑ i ∈ s, (p i)⁻¹ = 1) :
∫⁻ (a : α), ∏ i ∈ s, f i a ∂μ ≤ ∏ i ∈ s, MeasureTheory.eLpNorm (f i) (p i) μ

A version of Hölder with multiple arguments, allowing ∞ as an exponent.

theorem ENNReal.lintegral_mul_le_eLpNorm_mul_eLqNorm {α : Type u_1} [MeasurableSpace α] {μ : MeasureTheory.Measure α} {p q : ENNReal} (hpq : p.HolderConjugate q) {f g : α → ENNReal} (hf : AEMeasurable f μ) (hg : AEMeasurable g μ) :
∫⁻ (a : α), (f * g) a ∂μ ≤ MeasureTheory.eLpNorm f p μ * MeasureTheory.eLpNorm g q μ

Hölder's inequality for functions α → ℝ≥0∞, using exponents in ℝ≥0∞

theorem ENNReal.sq_lintegral_mul_le_mul_lintegral_sq {α : Type u_1} [MeasurableSpace α] {μ : MeasureTheory.Measure α} {f g : α → ENNReal} (hf : AEMeasurable f μ) (hg : AEMeasurable g μ) :
(∫⁻ (a : α), f a * g a ∂μ) ^ 2 ≤ (∫⁻ (a : α), f a ^ 2 ∂μ) * ∫⁻ (a : α), g a ^ 2 ∂μ

Cauchy–Schwarz inequality for functions α → ℝ≥0∞ (Hölder's inequality squared).

Special case of Young's convolution inequality when r = ∞.

theorem ENNReal.enorm_convolution_le_eLpNorm_mul_eLpNorm_mul_eLpNorm {𝕜 : Type u𝕜} {G : Type uG} [MeasurableSpace G] {μ : MeasureTheory.Measure G} {E : Type uE} {E' : Type uE'} {F : Type uF} [NormedAddCommGroup E] [NormedAddCommGroup E'] [NormedAddCommGroup F] [NontriviallyNormedField 𝕜] [NormedSpace 𝕜 E] [NormedSpace 𝕜 E'] [NormedSpace 𝕜 F] [NormedSpace ℝ F] {L : E →L[𝕜] E' →L[𝕜] F} [AddGroup G] [TopologicalSpace G] [IsTopologicalAddGroup G] [BorelSpace G] [μ.IsAddHaarMeasure] [LocallyCompactSpace G] [SecondCountableTopology G] [μ.IsNegInvariant] {p q r : ℝ} (hp : 1 ≤ p) (hq : 1 ≤ q) (hr : 1 ≤ r) (hpqr : p⁻¹ + q⁻¹ = r⁻¹ + 1) {f : G → E} {g : G → E'} (hf : MeasureTheory.AEStronglyMeasurable f μ) (hg : MeasureTheory.AEStronglyMeasurable g μ) (c : ℝ) (hL : ∀ (x y : G), ‖(L (f x)) (g y)‖ ≤ c * ‖f x‖ * ‖g y‖) (x : G) :

This inequality is used in the proof of Young's convolution inequality eLpNorm_convolution_le_ofReal.

theorem ENNReal.eLpNorm_convolution_le_ofReal {𝕜 : Type u𝕜} {G : Type uG} [MeasurableSpace G] {μ : MeasureTheory.Measure G} {E : Type uE} {E' : Type uE'} {F : Type uF} [NormedAddCommGroup E] [NormedAddCommGroup E'] [NormedAddCommGroup F] [NontriviallyNormedField 𝕜] [NormedSpace 𝕜 E] [NormedSpace 𝕜 E'] [NormedSpace 𝕜 F] [NormedSpace ℝ F] {L : E →L[𝕜] E' →L[𝕜] F} [AddGroup G] [TopologicalSpace G] [IsTopologicalAddGroup G] [BorelSpace G] [μ.IsAddHaarMeasure] [LocallyCompactSpace G] [SecondCountableTopology G] [μ.IsAddRightInvariant] {p q r : ℝ} (hp : 1 ≤ p) (hq : 1 ≤ q) (hr : 1 ≤ r) (hpqr : p⁻¹ + q⁻¹ = r⁻¹ + 1) {f : G → E} {g : G → E'} (hf : MeasureTheory.AEStronglyMeasurable f μ) (hg : MeasureTheory.AEStronglyMeasurable g μ) (c : ℝ) (hL : ∀ (x y : G), ‖(L (f x)) (g y)‖ ≤ c * ‖f x‖ * ‖g y‖) :
theorem ENNReal.eLpNorm_convolution_le_of_norm_le_mul {𝕜 : Type u𝕜} {G : Type uG} [MeasurableSpace G] {μ : MeasureTheory.Measure G} {E : Type uE} {E' : Type uE'} {F : Type uF} [NormedAddCommGroup E] [NormedAddCommGroup E'] [NormedAddCommGroup F] [NontriviallyNormedField 𝕜] [NormedSpace 𝕜 E] [NormedSpace 𝕜 E'] [NormedSpace 𝕜 F] [NormedSpace ℝ F] (L : E →L[𝕜] E' →L[𝕜] F) [AddGroup G] [TopologicalSpace G] [IsTopologicalAddGroup G] [BorelSpace G] [μ.IsAddHaarMeasure] [LocallyCompactSpace G] [SecondCountableTopology G] [μ.IsAddRightInvariant] {p q r : ENNReal} (hp : 1 ≤ p) (hq : 1 ≤ q) (hr : 1 ≤ r) (hpqr : p⁻¹ + q⁻¹ = r⁻¹ + 1) {f : G → E} {g : G → E'} (hf : MeasureTheory.AEStronglyMeasurable f μ) (hg : MeasureTheory.AEStronglyMeasurable g μ) (c : ℝ) (hL : ∀ (x y : G), ‖(L (f x)) (g y)‖ ≤ c * ‖f x‖ * ‖g y‖) :

Young's convolution inequality: the L^r seminorm of a convolution (f ⋆[L, μ] g) is bounded by ‖L‖ₑ times the product of the L^p and L^q seminorms, where 1 / p + 1 / q = 1 / r + 1. Here ‖L‖ₑ is replaced with a bound for L restricted to the ranges of f and g; see eLpNorm_convolution_le_enorm_mul for a version using ‖L‖ₑ explicitly.

Young's convolution inequality: the L^r seminorm of a convolution (f ⋆[L, μ] g) is bounded by ‖L‖ₑ times the product of the L^p and L^q seminorms, where 1 / p + 1 / q = 1 / r + 1.

theorem ENNReal.eLpNorm_Ioc_convolution_le_of_norm_le_mul_aux {𝕜 : Type u𝕜} {E : Type uE} {E' : Type uE'} {F : Type uF} [NormedAddCommGroup E] [NormedAddCommGroup E'] [NormedAddCommGroup F] [NontriviallyNormedField 𝕜] [NormedSpace 𝕜 E] [NormedSpace 𝕜 E'] [NormedSpace 𝕜 F] [NormedSpace ℝ F] (L : E →L[𝕜] E' →L[𝕜] F) (a : ℝ) {T : ℝ} [hT : Fact (0 < T)] {p q r : ENNReal} (hp : 1 ≤ p) (hq : 1 ≤ q) (hr : 1 ≤ r) (hpqr : p⁻¹ + q⁻¹ = r⁻¹ + 1) {f : ℝ → E} {g : ℝ → E'} (hfT : Function.Periodic f T) (hgT : Function.Periodic g T) (hf : MeasureTheory.AEStronglyMeasurable f MeasureTheory.volume) (hg : MeasureTheory.AEStronglyMeasurable g MeasureTheory.volume) (c : ℝ) (hL : ∀ (x y : ℝ), ‖(L (f x)) (g y)‖ ≤ c * ‖f x‖ * ‖g y‖) :
theorem ENNReal.eLpNorm_Ioc_convolution_le_of_norm_le_mul {𝕜 : Type u𝕜} {E : Type uE} {E' : Type uE'} {F : Type uF} [NormedAddCommGroup E] [NormedAddCommGroup E'] [NormedAddCommGroup F] [NontriviallyNormedField 𝕜] [NormedSpace 𝕜 E] [NormedSpace 𝕜 E'] [NormedSpace 𝕜 F] [NormedSpace ℝ F] (L : E →L[𝕜] E' →L[𝕜] F) (a : ℝ) {T : ℝ} [hT : Fact (0 < T)] {p q r : ENNReal} (hp : 1 ≤ p) (hq : 1 ≤ q) (hr : 1 ≤ r) (hpqr : p⁻¹ + q⁻¹ = r⁻¹ + 1) {f : ℝ → E} {g : ℝ → E'} (hgT : Function.Periodic g T) (hf : MeasureTheory.AEStronglyMeasurable f MeasureTheory.volume) (hg : MeasureTheory.AEStronglyMeasurable g MeasureTheory.volume) (c : ℝ) (hL : ∀ (x y : ℝ), ‖(L (f x)) (g y)‖ ≤ c * ‖f x‖ * ‖g y‖) :

Young's convolution inequality on (a, a + T]: the L^r seminorm of the convolution of T-periodic functions over (a, a + T] is bounded by ‖L‖ₑ times the product of the L^p and L^q seminorms on that interval, where 1 / p + 1 / q = 1 / r + 1. Here ‖L‖ₑ is replaced with a bound for L restricted to the ranges of f and g; see eLpNorm_Ioc_convolution_le_enorm_mul for a version using ‖L‖ₑ explicitly.

theorem ENNReal.eLpNorm_Ioc_convolution_le_enorm_mul {𝕜 : Type u𝕜} {E : Type uE} {E' : Type uE'} {F : Type uF} [NormedAddCommGroup E] [NormedAddCommGroup E'] [NormedAddCommGroup F] [NontriviallyNormedField 𝕜] [NormedSpace 𝕜 E] [NormedSpace 𝕜 E'] [NormedSpace 𝕜 F] [NormedSpace ℝ F] (L : E →L[𝕜] E' →L[𝕜] F) (a : ℝ) {T : ℝ} [hT : Fact (0 < T)] {p q r : ENNReal} (hp : 1 ≤ p) (hq : 1 ≤ q) (hr : 1 ≤ r) (hpqr : p⁻¹ + q⁻¹ = r⁻¹ + 1) {f : ℝ → E} {g : ℝ → E'} (hgT : Function.Periodic g T) (hf : MeasureTheory.AEStronglyMeasurable f MeasureTheory.volume) (hg : MeasureTheory.AEStronglyMeasurable g MeasureTheory.volume) :

Young's convolution inequality on (a, a + T]: the L^r seminorm of the convolution of T-periodic functions over (a, a + T] is bounded by ‖L‖ₑ times the product of the L^p and L^q seminorms on that interval, where 1 / p + 1 / q = 1 / r + 1.