A lower bound for central-binomial probability sums
ProvedAnalyticCombinatorics.Ch8.Partitions.Erdos.centralBinom_prob_sum_ge_sqrtanalytic-combinatoricselementarylean4partitions-remainder
For every natural number , including zero, put . Then
This is a finite-sum lower bound.
The local source models, scales or predicates are defined exactly by:
noncomputable def cbProb (m : ℕ) : ℝ := (Nat.centralBinom m : ℝ) / (4 : ℝ) ^ m
Preamble
import Definitions.Def_AnalyticCombinatorics_Ch8_Partitions_CentralBinomSum
import Mathlib.Data.Nat.Choose.Central
import Mathlib
set_option autoImplicit true
set_option relaxedAutoImplicit false
set_option maxSynthPendingDepth 3
/-!
# Central-binomial Green sum: `√N ≤ ∑_{m<N} C_m / 4^m` (renewal route B, SRW Green helper)
The lazy/simple-random-walk Green local time over a horizon `N` is `∑_{m<N} a_m` with
`a_m = binom(2m,m)/4^m = Nat.centralBinom m / 4^m` the return probability. The block-heat
lower bound `greenT_lower_fixed_window` consumes the fact that this partial sum grows like `√N`;
this file banks the clean, fully elementary lower bound
`√N ≤ ∑_{m<N} Nat.centralBinom m / 4^m`,
with the *exact* constant `1`, via the term bound `1/√(4m+1) ≤ a_m` and the telescope
`√(m+1) − √m ≤ 1/√(4m+1)`. No Stirling, Wallis, generating functions or integral estimates.
-/
open Finset
open AnalyticCombinatorics AnalyticCombinatorics.Ch8 AnalyticCombinatorics.Ch8.Partitions AnalyticCombinatorics.Ch8.Partitions.Erdos
Formal statement
theorem AnalyticCombinatorics.Ch8.Partitions.Erdos.centralBinom_prob_sum_ge_sqrt (N : ℕ) :
Real.sqrt (N : ℝ) ≤ ∑ m ∈ Finset.range N, cbProb m := by sorrySource
Human review
Confirmed by the mission captain (proposal self-audit).