Exceptional-center probability at most on good center atoms of stable points
ProvedOAI.Problem310.ConcreteWindowAtom.raw_failure_on_good_atomThis theorem bounds, on every good center atom of a stable point , the conditional probability that is an exceptional center of a random periodic set built from dyadic tables on a tree.
Tree and window data. Fix natural numbers . A word is a finite sequence over the alphabet ; the letter is the default child and are the nondefault children. Write for the length of a word and for followed by the letter . A node is a word of length , a leaf is a word of length exactly , and an edge is a nonempty word of length . Window data with parameters consist of functions and from words to such that, writing , the following hold for all edges :
- (W1) and ;
- (W2) and ;
- (W3) whenever is a proper prefix of ;
- (W4) whenever are letters, is a word, and and are edges;
- (W5) whenever is a prefix of ;
- (W6) .
Thus the edge owns the window of consecutive dyadic exponents.
Dyadic keys, tables and the law. For and put , where ; this is the index of the dyadic cell of length that contains modulo . An outcome consists of a selector table for every node and every , and a terminal table for every leaf ; different tables are different coordinates even when their cell indices coincide. Let be the finite set of all outcomes. For , denotes the product probability measure on under which every selector entry is a fair bit, every terminal entry equals with probability , and all entries are independent.
Routing and the routed set. Given selector tables , a word and a point , the selector fires at if and . The choice is the smallest firing , or the default child if no selector fires. Routing from a word for steps produces the words and for ; the routed leaf is the word obtained by routing from the empty word for steps. The routed set of is
the set of points whose routed leaf has terminal bit in the cell of the point.
Exceptional centers. For , a point is an exceptional center of if some dilate of the dyadic sequence based at misses entirely:
Center atoms, good exposures, stability. Fix . An exposure is a function assigning a bit to every node and every ; its center atom is the event
that the selector entries read at equal . The exposure is good if the route of takes the default child within depth when the selector tables are , the tables that agree with at the entries read at and vanish elsewhere: for some , where is the route of from the empty word. The point is stable for if for every node , every , every integer and every ,
so that dyadic perturbations at exponents beyond do not change the selector cells read at .
Theorem. Let be window data with parameters , where , and let . Assume the budget condition
If is stable for and is a good exposure, then
When , this says that, conditionally on the selector bits read at , the probability that some dilate with misses the random set is at most . Combined with a bound on the probability of the remaining atoms, on which the route of never takes the default child, it yields the pointwise failure bound that feeds the construction of an open periodic set meeting every such dilate.
Formalization Note Stability is StableAt W x, which asks that selectorValue at and at agree for every selector table; this is equivalent to the equality of dyadic keys displayed above. Goodness is GoodExposure, evaluated on centerCompletion, the selector table that extends by false off the exposed entries. The probability is ConcreteLabels.outcomeLaw, and the exceptional set is Auxiliary.exceptionalCenters with index set Ici 1 and the sequence . The budget hypothesis is stated in ENNReal with ENNReal.ofReal (1 - p / 2); since it is equivalent to the real inequality above.
import Mathlib
import Definitions.Def_DyadicAvoidance
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_PeriodicBasic
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_GridSeparation
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_FiniteTableModel
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_OrderedRouting
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_RoutingPath
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_CenterRouteExposure
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_ExceptionalCenters
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_CenterFailureBound
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_FirstDefault
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_ConcreteLabelMeasures
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_CenterFirstDefault
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_RandomSetDensity
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_RoutedSetDensity
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_RoutedExposure
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_FirstDefaultSelection
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_FiniteLocalPredicate
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_WindowData
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_WindowTableModel
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_StableRoutingBridge
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_WindowExceptionalTransfer
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_TrialAddresses
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_WindowTrialAddresses
import Definitions.Def_OAI_MeasureTheory_DyadicAvoidance_ConcreteWindowAtom
open OAI OAI.Problem310 OAI.Problem310.ConcreteWindowAtom
open Set MeasureTheory
open scoped ENNReal BigOperators
open FiniteTableModel RoutedSetDensity RoutingPath FiniteLocalPredicate
open CenterRouteExposure CenterFailureBound WindowExceptionalTransfer WindowTrialAddresses
variable {M d g r₀ : ℕ}theorem OAI.Problem310.ConcreteWindowAtom.raw_failure_on_good_atom
(W : WindowData (M + 1) d g r₀) (hM : 1 ≤ M)
(p : ℝ) (hp0 : 0 ≤ p) (hp1 : p ≤ 1)
(hbudget : ∀ r : ℕ, r₀ ≤ r → ∀ N : ℕ,
N ≤ 1 + M * r * 2 ^ (2 * r + 2) →
(N : ENNReal) * (ENNReal.ofReal (1 - p / 2)) ^ (M * r) ≤ ENNReal.ofReal p)
(x : ℝ) (hstable : StableAt W x)
(ξ : Node M d × Fin M → Bool) (hgood : GoodExposure (selectorEndpoints W) x ξ) :
ConcreteLabels.outcomeLaw (SelectorAddress (selectorEndpoints W))
(TerminalAddress (terminalEndpoints W)) p hp0 hp1
({ω | x ∈ Auxiliary.exceptionalCenters
(routedSet (selectorEndpoints W) (terminalEndpoints W) ω)
(Ici 1) (fun n => (2 : ℝ)⁻¹ ^ n)} ∩
centerAtom (selectorEndpoints W) (terminalEndpoints W) x ξ) ≤
ENNReal.ofReal p *
ConcreteLabels.outcomeLaw (SelectorAddress (selectorEndpoints W))
(TerminalAddress (terminalEndpoints W)) p hp0 hp1
(centerAtom (selectorEndpoints W) (terminalEndpoints W) x ξ) := by sorry