AFTD Auto-Formalizing Theoretical Domains

persuasion_slice_coupling_marginal

lemmaverified

The first marginal of the slice coupling is the slice prior.

Statement

open Finset in
lemma persuasion_slice_coupling_marginal (m : ℕ) (hm : 1 ≤ m) (x : persuasion_slice_coord m → Bool) :
    ∑ y, persuasion_slice_coupling m x y = persuasion_slice_prior m x
Source
step towards persuasion_receiver_ratio_optimal_constant (optimal receiver-utility constant 3/2, open in arXiv:2606.22226 v1, Sec. 5)
Verified
06 Oct 2026
Axioms
Classical.choiceQuot.soundpropext
Built on 5
Used by 2

Lean source view module on GitHub

open Finset in/-- The first marginal of the slice coupling is the slice prior. -/lemma persuasion_slice_coupling_marginal (m : ℕ) (hm : 1 ≤ m) (x : persuasion_slice_coord m → Bool) :    ∑ y, persuasion_slice_coupling m x y = persuasion_slice_prior m x := by  classical  unfold persuasion_slice_coupling persuasion_slice_prior  rw [sum_comm]  apply sum_congr rfl; intro s _  rw [sum_comm]  by_cases hs : persuasion_slice_state m s = x  · rw [if_pos hs]    have e : ∀ t, (∑ y, if s ≠ t ∧ persuasion_slice_state m s = x ∧        persuasion_slice_state m t = y then 1 / ((2 * m + 1 : ℝ) * (2 * m)) else 0) =        if s ≠ t then 1 / ((2 * m + 1 : ℝ) * (2 * m)) else 0 := by      intro t      by_cases ht : s ≠ t      · simp only [ht, hs, true_and, if_true, ne_eq, not_false_eq_true]        rw [sum_ite_eq]        simp      · simp [ht]    rw [sum_congr rfl (fun t _ => e t), ← sum_filter, Finset.sum_const, persuasion_slice_card_ne,      nsmul_eq_mul]    have : (m : ℝ) ≠ 0 := by exact_mod_cast (show m ≠ 0 by omega)    push_cast    field_simp  · rw [if_neg hs]    apply sum_eq_zero; intro t _    apply sum_eq_zero; intro y _    simp [hs]
See it in the graphDiscuss this resultChallenge it