OAI/Combinatorics/DiscreteConvexity/CoveringResidualProbability.lean

GitHub source file

5 declarations, 0 contain sorry, 0 declared as axioms

Exact survival probability of a tuple of distinct edges

The sampling coordinates are ordered pairs of edges, not their unions. Every ordered pair of positions contributes a distinct coordinate, including each diagonal position. The product Bernoulli law therefore assigns exactly (1 - p) ^ (m * m) to the event that all these coordinates are absent.

  1. theorem OAI.TalagrandDiscreteConvexity.Covering.orderedPairMap_injective line 18

    Source documentation

    Applying an injection in both coordinates preserves ordered pairs.

    theorem orderedPairMap_injective {I E : Type*} {e : I → E}
        (he : Function.Injective e) :
        Function.Injective (fun ij : I × I => (e ij.1, e ij.2)) := by
    …
  2. theorem OAI.TalagrandDiscreteConvexity.Covering.tuplePairs_card line 31

    Source documentation

    The tuple uses all m * m ordered pair coordinates, including the diagonal.

    theorem tuplePairs_card {H : Family X} {m : ℕ} (f : EdgeTuple H m) :
        (tuplePairs f).card = m * m := by
    …

    Mentions: OAI.TalagrandDiscreteConvexity.Family, OAI.TalagrandDiscreteConvexity.Covering.EdgeTuple, OAI.TalagrandDiscreteConvexity.Covering.tuplePairs

  3. theorem OAI.TalagrandDiscreteConvexity.Covering.diagonal_mem_tuplePairs line 39

    Source documentation

    A diagonal edge pair is one of the sampled coordinates used by the tuple.

    theorem diagonal_mem_tuplePairs {H : Family X} {m : ℕ}
        (f : EdgeTuple H m) (i : Fin m) : (f.val i, f.val i) ∈ tuplePairs f := by
    …

    Mentions: OAI.TalagrandDiscreteConvexity.Family, OAI.TalagrandDiscreteConvexity.Covering.EdgeTuple, OAI.TalagrandDiscreteConvexity.Covering.tuplePairs

  4. theorem OAI.TalagrandDiscreteConvexity.Covering.disjoint_tuplePairs_iff line 44

    Source documentation

    Avoiding the pair set means that none of the ordered tuple pairs was sampled.

    theorem disjoint_tuplePairs_iff {H : Family X} {m : ℕ}
        (f : EdgeTuple H m) (sample : Sample H) :
        Disjoint sample (tuplePairs f) ↔ ∀ i j, (f.val i, f.val j) ∉ sample := by
    …

    Mentions: OAI.TalagrandDiscreteConvexity.Family, OAI.TalagrandDiscreteConvexity.Covering.EdgeTuple, OAI.TalagrandDiscreteConvexity.Covering.Sample, OAI.TalagrandDiscreteConvexity.Covering.tuplePairs

  5. theorem OAI.TalagrandDiscreteConvexity.Covering.tuple_survival_probability line 57

    Source documentation

    Exact residual probability under the actual finite product Bernoulli law.

    theorem tuple_survival_probability {H : Family X} {m : ℕ}
        (f : EdgeTuple H m) (p : ℝ) (hp0 : 0 ≤ p) (hp1 : p ≤ 1) :
        (bernoulliLaw (α := ↥H × ↥H) p hp0 hp1).prob
    …

    Mentions: OAI.TalagrandDiscreteConvexity.Family, OAI.TalagrandDiscreteConvexity.Covering.EdgeTuple, OAI.TalagrandDiscreteConvexity.bernoulliLaw