A conditional unmatched-mass recursion on good pairs
ProvedAnalyticCombinatorics.Ch8.Partitions.Erdos.umass_succ_le_of_goodanalytic-combinatoricscoupling-hittinglean4partitions-support
Per-step unmatched-mass recursion conditional on minorization on a Good predicate; it is not global coalescence.
The full literal source contract is:
theorem umass_succ_le_of_good (Good : α → α → Prop) [DecidableRel Good] (δ : ℝ) (hδ0 : 0 ≤ δ)
(hminor_good : ∀ x y, Good x y → δ ≤ cmass P rnk W x y) (t : ℕ) :
umass P rnk W i j (t + 1)
≤ umass P rnk W i j t
- δ * ∑ x, ∑ y, (if Good x y then Umat P rnk W i j t x y else 0)
Preamble
import Definitions.Def_AnalyticCombinatorics_Ch8_Partitions_ITERCoupling
import Mathlib
set_option autoImplicit true
set_option relaxedAutoImplicit false
set_option maxSynthPendingDepth 3
/-!
# R7 Fact B via Doeblin: minorization on a sub-predicate ⟹ umass recursion
For the KILLED chain, the Doeblin minorization `δ ≤ cmass` holds only on the *active high-rank* window
`GoodHi` (where `Pker_window_minor`/brick 79 applies), NOT on absorbed/low-rank pairs. This brick
generalizes the per-step contraction to any sub-predicate `Good` on which the minorization holds:
`umass (t+1) ≤ umass t − δ · (Good-window mass of Umat t)`.
It follows from `umass_succ_eq` (`umass(t+1) = umass(t) − ∑ Umat·cmass`) since the coalescence mass
`∑ Umat·cmass ≥ δ · ∑_{Good} Umat`. Then the (banked) CoalesceBridge consumes this with `g = ĝ` the
conditioned `Good`-fraction. Deterministic finite-sum. Opus-authored.
-/
open Finset
open AnalyticCombinatorics AnalyticCombinatorics.Ch8 AnalyticCombinatorics.Ch8.Partitions AnalyticCombinatorics.Ch8.Partitions.Erdos
variable {α : Type*} [Fintype α] [DecidableEq α]
variable (P : α → α → ℝ) (rnk : α → ℕ) (W : ℕ)
variable (hPnn : ∀ x y, 0 ≤ P x y) (hPmass : ∀ x, ∑ y, P x y = 1)
variable (i j : α)
include hPnn hPmass
Formal statement
theorem AnalyticCombinatorics.Ch8.Partitions.Erdos.umass_succ_le_of_good (Good : α → α → Prop) [DecidableRel Good] (δ : ℝ) (hδ0 : 0 ≤ δ)
(hminor_good : ∀ x y, Good x y → δ ≤ cmass P rnk W x y) (t : ℕ) :
umass P rnk W i j (t + 1)
≤ umass P rnk W i j t
- δ * ∑ x, ∑ y, (if Good x y then Umat P rnk W i j t x y else 0) := by sorrySource
Human review
Confirmed by the mission captain (proposal self-audit).