-
Notifications
You must be signed in to change notification settings - Fork 38
Expand file tree
/
Copy pathDefs.lean
More file actions
273 lines (216 loc) · 13.3 KB
/
Defs.lean
File metadata and controls
273 lines (216 loc) · 13.3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
import Carleson.ToMathlib.Annulus
import Carleson.ToMathlib.MeasureTheory.Measure.IsDoubling
import Carleson.ToMathlib.WeakType
import Mathlib.Analysis.CStarAlgebra.Classes
import Mathlib.Analysis.Fourier.AddCircle
open MeasureTheory Measure Metric Complex Set Function ENNReal
open scoped NNReal
noncomputable section
/-! # Main statements of the Carleson project
This file contains the statements of the main theorems from the Carleson formalization project:
Theorem 1.0.1 (classical Carleson), Theorem 1.1.1 (metric space Carleson) and Theorem 1.1.2
(linearised metric Carleson), as well as the definitions required to state these results.
## Main definitions
- `MeasureTheory.DoublingMeasure`: A metric space with a measure with some nice propreties,
including a doubling condition. This is called a "doubling metric measure space" in the blueprint.
- `FunctionDistances`: class stating that continuous functions have distances associated to
every ball.
- `IsOneSidedKernel K` states that `K` is a one-sided Calderon-Zygmund kernel.
- `KernelProofData`: Data common through most of chapters 2-7. These contain the minimal axioms
for `kernel-summand`'s proof.
- `ClassicalCarleson`: statement of Carleson's theorem asserting a.e. convergence of the partial
Fourier sums for continous functions (Theorem 1.0.1 in the blueprint).
- `MetricSpaceCarleson`: statement of Theorem 1.1.1 from the blueprint.
- `LinearizedMetricCarleson`: statement of Theorem 1.1.2 from the blueprint.
-/
section DoublingMeasure
universe u
/- # Doubling metric measure spaces -/
/-- A metric space with a measure with some nice propreties, including a doubling condition.
This is called a "doubling metric measure space" in the blueprint.
`A` will usually be `2 ^ a`. -/
class MeasureTheory.DoublingMeasure (X : Type*) (A : outParam ℝ≥0) [PseudoMetricSpace X] extends
CompleteSpace X, LocallyCompactSpace X,
MeasureSpace X, BorelSpace X,
IsLocallyFiniteMeasure (volume : Measure X),
IsDoubling (volume : Measure X) A, NeZero (volume : Measure X) where
variable {𝕜 X : Type*} {A : ℕ} [_root_.RCLike 𝕜] [PseudoMetricSpace X]
/-- The local oscillation of two functions w.r.t. a set `E`. This is `d_E` in the blueprint. -/
def localOscillation (E : Set X) (f g : C(X, 𝕜)) : ℝ≥0∞ :=
⨆ z ∈ E ×ˢ E, ENNReal.ofReal ‖f z.1 - g z.1 - f z.2 + g z.2‖
/-- A class stating that continuous functions have distances associated to every ball.
We use a separate type to conveniently index these functions. -/
class FunctionDistances (𝕜 : outParam Type*) (X : Type u)
[NormedField 𝕜] [TopologicalSpace X] where
/-- A type of continuous functions from `X` to `𝕜`. -/
Θ : Type u
/-- The coercion map from `Θ` to `C(X, 𝕜)`. -/
coeΘ : Θ → C(X, 𝕜)
/-- Injectivity of the coercion map from `Θ` to `C(X, 𝕜)`. -/
coeΘ_injective {f g : Θ} (h : ∀ x, coeΘ f x = coeΘ g x) : f = g
/-- For each `_x : X` and `_r : ℝ`, a `PseudoMetricSpace Θ`. -/
metric : ∀ (_x : X) (_r : ℝ), PseudoMetricSpace Θ
export FunctionDistances (Θ coeΘ)
section FunctionDistances
variable [FunctionDistances 𝕜 X]
instance : Coe (Θ X) C(X, 𝕜) := ⟨FunctionDistances.coeΘ⟩
instance : FunLike (Θ X) X 𝕜 where
coe := fun f ↦ (f : C(X, 𝕜))
coe_injective' _ _ hfg := FunctionDistances.coeΘ_injective fun x ↦ congrFun hfg x
set_option linter.unusedVariables false in
/-- Class used to endow `Θ X` with a pseudometric space structure. -/
@[nolint unusedArguments]
def WithFunctionDistance (x : X) (r : ℝ) := Θ X
instance {x : X} {r : ℝ} [d : FunctionDistances 𝕜 X] :
PseudoMetricSpace (WithFunctionDistance x r) := d.metric x r
end FunctionDistances
/-- The real-valued distance function on `WithFunctionDistance x r`. -/
notation3 "dist_{" x ", " r "}" => @dist (WithFunctionDistance x r) _
/-- The `ℝ≥0∞`-valued distance function on `WithFunctionDistance x r`. Preferred over `nndist`. -/
notation3 "edist_{" x ", " r "}" => @edist (WithFunctionDistance x r) _
/-- A set `Θ` of (continuous) functions is compatible. `A` will usually be `2 ^ a`. -/
class CompatibleFunctions (𝕜 : outParam Type*) (X : Type u) (A : outParam ℕ)
[RCLike 𝕜] [PseudoMetricSpace X] extends FunctionDistances 𝕜 X where
eq_zero : ∃ o : X, ∀ f : Θ, coeΘ f o = 0
/-- The distance is bounded below by the local oscillation. (1.1.4) -/
localOscillation_le_cdist {x : X} {r : ℝ} {f g : Θ} :
localOscillation (ball x r) (coeΘ f) (coeΘ g) ≤ ENNReal.ofReal (dist_{x, r} f g)
/-- The distance is monotone in the ball. (1.1.6) -/
cdist_mono {x₁ x₂ : X} {r₁ r₂ : ℝ} {f g : Θ}
(h : ball x₁ r₁ ⊆ ball x₂ r₂) : dist_{x₁, r₁} f g ≤ dist_{x₂, r₂} f g
/-- The distance of a ball with large radius is bounded above. (1.1.5) -/
cdist_le {x₁ x₂ : X} {r : ℝ} {f g : Θ} (h : dist x₁ x₂ < 2 * r) :
dist_{x₂, 2 * r} f g ≤ A * dist_{x₁, r} f g
/-- The distance of a ball with large radius is bounded below. (1.1.7) -/
le_cdist {x₁ x₂ : X} {r : ℝ} {f g : Θ} (h1 : ball x₁ r ⊆ ball x₂ (A * r)) :
/-(h2 : A * r ≤ Metric.diam (univ : Set X))-/
2 * dist_{x₁, r} f g ≤ dist_{x₂, A * r} f g
/-- Every ball of radius `2R` can be covered by `A` balls of radius `R`. (1.1.8) -/
allBallsCoverBalls {x : X} {r : ℝ} :
AllBallsCoverBalls (WithFunctionDistance x r) 2 A
variable [hXA : DoublingMeasure X A]
/-- The inhomogeneous Lipschitz norm on a ball. -/
def iLipENorm {𝕜 X : Type*} [NormedField 𝕜] [PseudoMetricSpace X]
(φ : X → 𝕜) (x₀ : X) (R : ℝ) : ℝ≥0∞ :=
(⨆ x ∈ ball x₀ R, ‖φ x‖ₑ) +
ENNReal.ofReal R * ⨆ (x ∈ ball x₀ R) (y ∈ ball x₀ R) (_ : x ≠ y), ‖φ x - φ y‖ₑ / edist x y
variable (X) in
/-- Θ is τ-cancellative. `τ` will usually be `1 / a` -/
class IsCancellative (τ : ℝ) [CompatibleFunctions ℝ X A] : Prop where
/- We register a definition with strong assumptions, which makes them easier to prove.
However, `enorm_integral_exp_le` removes them for easier application. -/
enorm_integral_exp_le' {x : X} {r : ℝ} {φ : X → ℂ} (hr : 0 < r) (h1 : iLipENorm φ x r ≠ ∞)
(h2 : support φ ⊆ ball x r) {f g : Θ X} :
‖∫ x, exp (I * (f x - g x)) * φ x‖ₑ ≤
(A : ℝ≥0∞) * volume (ball x r) * iLipENorm φ x r * (1 + edist_{x, r} f g) ^ (- τ)
/-- The "volume function" `V`. Preferably use `vol` instead. -/
protected def Real.vol {X : Type*} [PseudoMetricSpace X] [MeasureSpace X] (x y : X) : ℝ :=
volume.real (ball x (dist x y))
/-- `R_Q(θ, x)` defined in (1.1.17). -/
def upperRadius [FunctionDistances ℝ X] (Q : X → Θ X) (θ : Θ X) (x : X) : ℝ≥0∞ :=
⨆ (r : ℝ) (_ : dist_{x, r} θ (Q x) < 1), ENNReal.ofReal r
/-- The linearized maximally truncated nontangential Calderon–Zygmund operator `T_Q^θ`. -/
def linearizedNontangentialOperator [FunctionDistances ℝ X] (Q : X → Θ X) (θ : Θ X)
(K : X → X → ℂ) (f : X → ℂ) (x : X) : ℝ≥0∞ :=
⨆ (R₂ : ℝ) (R₁ ∈ Ioo 0 R₂) (x' ∈ ball x R₁),
‖∫ y in EAnnulus.oo x' (ENNReal.ofReal R₁) (min (ENNReal.ofReal R₂) (upperRadius Q θ x')),
K x' y * f y‖ₑ
/-- The maximally truncated nontangential Calderon–Zygmund operator `T_*`. -/
def nontangentialOperator (K : X → X → ℂ) (f : X → ℂ) (x : X) : ℝ≥0∞ :=
⨆ (R₂ : ℝ) (R₁ ∈ Ioo 0 R₂) (x' ∈ ball x R₁), ‖∫ y in Annulus.oo x' R₁ R₂, K x' y * f y‖ₑ
/-- The integrand in the (linearized) Carleson operator.
This is `G` in Lemma 3.0.1. -/
def carlesonOperatorIntegrand [FunctionDistances ℝ X] (K : X → X → ℂ)
(θ : Θ X) (R₁ R₂ : ℝ) (f : X → ℂ) (x : X) : ℂ :=
∫ y in Annulus.oo x R₁ R₂, K x y * f y * exp (I * θ y)
/-- The linearized generalized Carleson operator `T_Q`, taking values in `ℝ≥0∞`.
Use `ENNReal.toReal` to get the corresponding real number. -/
def linearizedCarlesonOperator [FunctionDistances ℝ X] (Q : X → Θ X) (K : X → X → ℂ)
(f : X → ℂ) (x : X) : ℝ≥0∞ :=
⨆ (R₁ : ℝ) (R₂ : ℝ) (_ : 0 < R₁) (_ : R₁ < R₂), ‖carlesonOperatorIntegrand K (Q x) R₁ R₂ f x‖ₑ
/-- The generalized Carleson operator `T`, taking values in `ℝ≥0∞`.
Use `ENNReal.toReal` to get the corresponding real number. -/
def carlesonOperator [FunctionDistances ℝ X] (K : X → X → ℂ) (f : X → ℂ) (x : X) : ℝ≥0∞ :=
⨆ (θ : Θ X), linearizedCarlesonOperator (fun _ ↦ θ) K f x
end DoublingMeasure
/-- The main constant in the blueprint, driving all the construction, is `D = 2 ^ (100 * a ^ 2)`.
It turns out that the proof is robust, and works for other values of `100`, giving better constants
in the end. We will formalize it using a parameter `𝕔` (that we fix equal to `100` to follow
the blueprint) and having `D = 2 ^ (𝕔 * a ^ 2)`. We register two lemmas `seven_le_c` and
`c_le_100` and will never unfold `𝕔` from this point on. -/
irreducible_def 𝕔 : ℕ := 100
/-- This is usually the value of the argument `A` in `DoublingMeasure`
and `CompatibleFunctions` -/
@[simp] abbrev defaultA (a : ℕ) : ℕ := 2 ^ a
/-- `defaultτ` is the inverse of `a`. -/
@[simp] def defaultτ (a : ℕ) : ℝ := a⁻¹
section Kernel
variable {X : Type*} {a : ℕ} {K : X → X → ℂ} [PseudoMetricSpace X] [MeasureSpace X]
open Function
/-- The constant used twice in the definition of the Calderon-Zygmund kernel. -/
@[simp] def C_K (a : ℝ) : ℝ≥0 := 2 ^ a ^ 3
/-- `K` is a one-sided Calderon-Zygmund kernel.
In the formalization `K x y` is defined everywhere, even for `x = y`. The assumptions on `K` show
that `K x x = 0`. -/
class IsOneSidedKernel (a : outParam ℕ) (K : X → X → ℂ) : Prop where
measurable_K : Measurable (uncurry K)
norm_K_le_vol_inv (x y : X) : ‖K x y‖ ≤ C_K a / Real.vol x y
norm_K_sub_le {x y y' : X} (h : 2 * dist y y' ≤ dist x y) :
‖K x y - K x y'‖ ≤ (dist y y' / dist x y) ^ (a : ℝ)⁻¹ * (C_K a / Real.vol x y)
end Kernel
/-- A constant used on the boundedness of `T_Q^θ` and `T_*`. We generally assume
`HasBoundedStrongType (linearizedNontangentialOperator Q θ K · ·) 2 2 volume volume (C_Ts a)`
throughout this formalization. -/
def C_Ts (a : ℕ) : ℝ≥0 := 2 ^ a ^ 3
/-- Data common through most of chapters 2-7.
These contain the minimal axioms for `kernel-summand`'s proof.
This is used in chapter 3 when we don't have all other fields from `ProofData`. -/
class KernelProofData {X : Type*} (a : outParam ℕ) (K : outParam (X → X → ℂ))
[PseudoMetricSpace X] where
d : DoublingMeasure X (defaultA a)
four_le_a : 4 ≤ a
cf : CompatibleFunctions ℝ X (defaultA a)
hcz : IsOneSidedKernel a K
export KernelProofData (four_le_a)
attribute [instance] KernelProofData.d KernelProofData.cf KernelProofData.hcz
section statements
/- ## Main statements
This section contains the statements of the main theorems from the project: Theorem 1.0.1
(classical Carleson), Theorem 1.1.1 (metric space Carleson) and Theorem 1.1.2 (linearised metric
Carleson). -/
set_option linter.unusedVariables false
open Real
/-- The Nᵗʰ partial Fourier sum of `f : ℝ → ℂ` for `N : ℕ`. -/
def partialFourierSum (N : ℕ) (f : ℝ → ℂ) (x : ℝ) : ℂ := ∑ n ∈ Finset.Icc (-(N : ℤ)) N,
fourierCoeffOn Real.two_pi_pos f n * fourier n (x : AddCircle (2 * π))
local notation "S_" => partialFourierSum
/-- Theorem 1.0.1: Carleson's theorem asserting a.e. convergence of the partial Fourier sums for
continous functions.
For the proof, see `classical_carleson` in the file `Carleson.Classical.ClassicalCarleson`. -/
def ClassicalCarleson : Prop :=
∀ {f : ℝ → ℂ} (cont_f : Continuous f) (periodic_f : f.Periodic (2 * π)),
∀ᵐ x, Filter.Tendsto (S_ · f x) Filter.atTop (nhds (f x))
/-- The constant used in `MetricSpaceCarleson` and `LinearizedMetricCarleson`.
Has value `2 ^ (443 * a ^ 3) / (q - 1) ^ 6` in the blueprint. -/
def C1_0_2 (a : ℕ) (q : ℝ≥0) : ℝ≥0 := 2 ^ ((3 * 𝕔 + 18 + 5 * (𝕔 / 4)) * a ^ 3) / (q - 1) ^ 6
/-- Theorem 1.1.1.
For the proof, see `metric_carleson` in the file `Carleson.MetricCarleson.Main`. -/
def MetricSpaceCarleson : Prop :=
∀ {X : Type*} {a : ℕ} [MetricSpace X] {q q' : ℝ≥0} {F G : Set X} {K : X → X → ℂ}
[KernelProofData a K] {f : X → ℂ} [IsCancellative X (defaultτ a)] (hq : q ∈ Ioc 1 2)
(hqq' : q.HolderConjugate q') (mF : MeasurableSet F) (mG : MeasurableSet G)
(mf : Measurable f) (nf : (‖f ·‖) ≤ F.indicator 1)
(hT : HasBoundedStrongType (nontangentialOperator K · ·) 2 2 volume volume (C_Ts a)),
∫⁻ x in G, carlesonOperator K f x ≤ C1_0_2 a q * volume G ^ (q' : ℝ)⁻¹ * volume F ^ (q : ℝ)⁻¹
/-- Theorem 1.1.2.
For the proof, see `linearized_metric_carleson` in the file `Carleson.MetricCarleson.Linearized`. -/
def LinearizedMetricCarleson : Prop :=
∀ {X : Type*} {a : ℕ} [MetricSpace X] {q q' : ℝ≥0} {F G : Set X} {K : X → X → ℂ}
[KernelProofData a K] {Q : SimpleFunc X (Θ X)} {f : X → ℂ} [IsCancellative X (defaultτ a)]
(hq : q ∈ Ioc 1 2) (hqq' : q.HolderConjugate q') (mF : MeasurableSet F) (mG : MeasurableSet G)
(mf : Measurable f) (nf : (‖f ·‖) ≤ F.indicator 1)
(hT : ∀ θ : Θ X, HasBoundedStrongType (linearizedNontangentialOperator Q θ K · ·)
2 2 volume volume (C_Ts a)),
∫⁻ x in G, linearizedCarlesonOperator Q K f x ≤
C1_0_2 a q * volume G ^ (q' : ℝ)⁻¹ * volume F ^ (q : ℝ)⁻¹
end statements