YukonModule.ToMathlib.Analysis.MeanInequalities.part0
DefinitionYukon_8321510fb5e60b1f0f309da5better-codes
Source module ToMathlib.Analysis.MeanInequalities.
Definition code
/-
Copyright (c) 2026 Quang Dao. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Quang Dao
-/
module
public import Mathlib.MeasureTheory.Integral.MeanInequalities
public import Mathlib.MeasureTheory.Integral.Lebesgue.Countable
public import Init
set_option backward.isDefEq.respectTransparency.types false
/-!
# Hölder Inequality for ENNReal-Valued Infinite Sums
This file provides a `tsum` version of Hölder's inequality for `ℝ≥0∞`-valued functions,
obtained by specializing the `lintegral` Hölder inequality to the counting measure.
The main advantage over `NNReal.inner_le_Lp_mul_Lq_tsum` in Mathlib is that no
`Summable` hypotheses are needed: `ℝ≥0∞` sums always converge (to `⊤` in the worst case).
## Main Results
- `ENNReal.inner_le_Lp_mul_Lq_tsum`: Hölder's inequality for `ℝ≥0∞`-valued `tsum`.
-/
@[expose] public section
noncomputable section
open MeasureTheory ENNReal
namespace ENNReal
variable {ι : Type*}
/-- Hölder's inequality for `ℝ≥0∞`-valued infinite sums:
`∑' i, f i * g i ≤ (∑' i, f i ^ p) ^ (1/p) * (∑' i, g i ^ q) ^ (1/q)`.
Obtained by specializing `ENNReal.lintegral_mul_le_Lp_mul_Lq` to the counting measure
with `DiscreteMeasurableSpace`. No `Summable` hypotheses are needed. -/
theorem inner_le_Lp_mul_Lq_tsum {p q : ℝ} (hpq : p.HolderConjugate q)
(f g : ι → ℝ≥0∞) :
∑' i, f i * g i ≤ (∑' i, f i ^ p) ^ (1 / p) * (∑' i, g i ^ q) ^ (1 / q) := by
letI : MeasurableSpace ι := ⊤
haveI : DiscreteMeasurableSpace ι := ⟨fun _ => trivial⟩
have := lintegral_mul_le_Lp_mul_Lq (α := ι) Measure.count hpq
(AEMeasurable.of_discrete (f := f)) (AEMeasurable.of_discrete (f := g))
simp only [lintegral_count, Pi.mul_apply] at this
exact this
end ENNReal
Source