AFTD Auto-Formalizing Theoretical Domains

realization_prob

def

Given a simple graph G with finite edge set and edge probabilities p : G.edgeSet → ℝ, the probability of a binary edge weight realization α : G.edgeSet → Bool is the product over all edges e of (if α(e) then p(e) else 1 - p(e)).

Statement

noncomputable def realization_prob {V : Type*} (G : SimpleGraph V) [Fintype G.edgeSet]
    (p : G.edgeSet → ℝ) (α : G.edgeSet → Bool) : ℝ
Defined in
arXiv:2610.40844 Sec. 2
Verified
07 Nov 2026
Axioms
none listed
Used by 1
  • nonadaptive_mst_expected_costThe expected query cost of a non-adaptive querying order π on a finite graph G with decidable adjacency and edge probabilities p is the…

Read back from the Lean

Given a type V, a simple graph G on V such that G.edgeSet is finite ([Fintype G.edgeSet]), a function p : G.edgeSet → ℝ, and a configuration α : G.edgeSet → Bool, realization_prob G p α is defined as the product over all edges e in G.edgeSet of p e if α e is true, and 1 - p e otherwise (i.e. ∏ e : G.edgeSet, if α e then p e else 1 - p e).

Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.

Lean source view module on GitHub

/-- The product Bernoulli probability of an edge realization α given marginal edge probabilities p. -/
noncomputable def realization_prob {V : Type*} (G : SimpleGraph V) [Fintype G.edgeSet]
    (p : G.edgeSet → ℝ) (α : G.edgeSet → Bool) : ℝ := ∏ e : G.edgeSet, if α e then p e else 1 - p e
See it in the graphDiscuss this resultChallenge it