OAI/Combinatorics/StableGraphs/ExteriorBounds.lean

GitHub source file

66 declarations, 0 contain sorry, 0 declared as axioms

Exterior monomials, color counts and finite stable-graph bounds

The color-incidence set indexes slots in triples of colors. Its lexicographic enumeration defines the color word and is used to count occurrences and word-preserving permutations. The exterior-algebra constructions prove nonvanishing of a factorial-scaled monomial, not a detector evaluation.

The strata class Y, its realizations, Pixton relation span, detector pairing and geometric vanishing are not defined here. These combinatorial results therefore do not assert the geometric main theorem or cohomological corollary.

  1. def OAI.PixtonChow.genus line 26
  2. def OAI.PixtonChow.entryCount line 30

    Source documentation

    Computed by the falling-factorial formula to avoid expanding Pascal's recursion at 10^60; entryCount_eq_choose proves the specified value.

    def entryCount : ℕ := genus.descFactorial 3 / 6
  3. theorem OAI.PixtonChow.entryCount_eq_choose line 32
  4. def OAI.PixtonChow.markingCount line 36
  5. theorem OAI.PixtonChow.parameter_bounds line 39

    Source documentation

    The elementary numerical bounds needed for the parameters.

    theorem parameter_bounds (g d : ℕ) (hg : 2 ≤ g) :
        2 < 2 * g + 3 * d ∧ 2 * d < 3 * g - 3 + 3 * d := by
    …
  6. theorem OAI.PixtonChow.parameters_stable line 44

    Source documentation

    Stability of the specific pair in the source, in the form 2g + n > 2.

    theorem parameters_stable : 2 < 2 * genus + markingCount :=
    …

    Mentions: OAI.PixtonChow.genus, OAI.PixtonChow.markingCount

  7. theorem OAI.PixtonChow.proposed_codimension_lt_dimension line 49

    Source documentation

    The proposed codimension is below the moduli dimension. This says nothing about the existence or the degree of a strata-algebra element.

    theorem proposed_codimension_lt_dimension :
        2 * entryCount < 3 * genus - 3 + markingCount :=
    …

    Mentions: OAI.PixtonChow.entryCount, OAI.PixtonChow.genus, OAI.PixtonChow.markingCount

  8. abbrev OAI.PixtonChow.Triple line 56

    Source documentation

    A set of three distinct colors; sorting it gives its unique increasing triple.

    abbrev Triple (g : ℕ) := Set.powersetCard (Fin g) 3
  9. abbrev OAI.PixtonChow.ColorSlot line 59

    Source documentation

    Slots indexed by a triple together with one of its colors.

    abbrev ColorSlot (g : ℕ) := (s : Triple g) × {c : Fin g // c ∈ s.val}

    Mentions: OAI.PixtonChow.Triple

  10. def OAI.PixtonChow.incidenceColor line 62

    Source documentation

    The color of a slot.

    def incidenceColor {g : ℕ} (p : ColorSlot g) : Fin g := p.2.val

    Mentions: OAI.PixtonChow.ColorSlot

  11. theorem OAI.PixtonChow.triple_card line 64
  12. theorem OAI.PixtonChow.block_card line 69

    Source documentation

    Each block contains exactly three slots.

    theorem block_card {g : ℕ} (s : Triple g) :
        Fintype.card {c : Fin g // c ∈ s.val} = 3 := by
    …

    Mentions: OAI.PixtonChow.Triple

  13. theorem OAI.PixtonChow.colorSlot_card line 73
  14. def OAI.PixtonChow.colorFiberEquiv line 80

    Source documentation

    Erasing the choice of a distinguished color in a fiber of incidenceColor leaves precisely a triple containing that color.

    def colorFiberEquiv {g : ℕ} (c : Fin g) :
        {p : ColorSlot g // incidenceColor p = c} ≃ {s : Triple g // c ∈ s.val} where
    …

    Mentions: OAI.PixtonChow.ColorSlot, OAI.PixtonChow.incidenceColor, OAI.PixtonChow.Triple

  15. def OAI.PixtonChow.containingTripleEquiv line 94

    Source documentation

    Repackaging as the finite set of three-element supersets of {c}.

    def containingTripleEquiv {g : ℕ} (c : Fin g) :
        {s : Triple g // c ∈ s.val} ≃
          {s // s ∈ ((Finset.univ : Finset (Fin g)).powersetCard 3).filter
            ({c} ⊆ ·)} where
    …

    Mentions: OAI.PixtonChow.Triple

  16. theorem OAI.PixtonChow.containingTriple_card line 105

    Source documentation

    Every color belongs to choose (g-1) 2 of the triples.

    theorem containingTriple_card {g : ℕ} (c : Fin g) :
        Fintype.card {s : Triple g // c ∈ s.val} = (g - 1).choose 2 := by
    …

    Mentions: OAI.PixtonChow.Triple

  17. theorem OAI.PixtonChow.colorFiber_card line 111
  18. abbrev OAI.PixtonChow.ColorStabilizer line 116

    Source documentation

    A permutation of slots preserving their colors.

    abbrev ColorStabilizer (g : ℕ) :=
    …
  19. theorem OAI.PixtonChow.colorStabilizer_card line 121

    Source documentation

    The factorial multiplicity appearing in eq:detector-candidate-value. This counts permutations only: no equality involving the detector is asserted.

    theorem colorStabilizer_card (g : ℕ) :
        Fintype.card (ColorStabilizer g) = ((g - 1).choose 2).factorial ^ g := by
    …

    Mentions: OAI.PixtonChow.ColorStabilizer

  20. theorem OAI.PixtonChow.reindexed_stabilizer_card line 129

    Source documentation

    Counts are unchanged by any enumeration of the slots, including the lexicographic enumeration used in Definition def:candidate.

    theorem reindexed_stabilizer_card (g : ℕ) {A : Type*} [Fintype A] [DecidableEq A]
        (e : A ≃ ColorSlot g) :
        Fintype.card {p : Equiv.Perm A // (incidenceColor ∘ e) ∘ p = incidenceColor ∘ e} =
          ((g - 1).choose 2).factorial ^ g := by
    …

    Mentions: OAI.PixtonChow.ColorSlot, OAI.PixtonChow.incidenceColor

  21. def OAI.PixtonChow.tripleLexOrder line 146

    Source documentation

    Lexicographic order on the sorted lists representing the triples.

    @[instance_reducible] def tripleLexOrder (g : ℕ) : LinearOrder (Triple g) :=
    …

    Mentions: OAI.PixtonChow.Triple

  22. def OAI.PixtonChow.tripleEnumeration line 157

    Source documentation

    Increasing enumeration of all three-color blocks in lexicographic order.

    def tripleEnumeration (g : ℕ) : Fin (g.choose 3) ≃ Triple g :=
    …

    Mentions: OAI.PixtonChow.Triple

  23. def OAI.PixtonChow.blockSlotEquiv line 161

    Source documentation

    Each block is enumerated in increasing color order.

    def blockSlotEquiv (g : ℕ) : Fin (g.choose 3) × Fin 3 ≃ ColorSlot g :=
    …

    Mentions: OAI.PixtonChow.ColorSlot

  24. def OAI.PixtonChow.orderedSlotEquiv line 170

    Source documentation

    The first three positions are the first lexicographic block, etc.

    def orderedSlotEquiv (g : ℕ) : Fin (3 * g.choose 3) ≃ ColorSlot g :=
    …

    Mentions: OAI.PixtonChow.ColorSlot

  25. def OAI.PixtonChow.colorWord line 176

    Source documentation

    The concatenation of the increasing triples, with colors numbered 0,...,g-1 instead of 1,...,g.

    def colorWord (g : ℕ) : Fin (3 * g.choose 3) → Fin g :=
    …
  26. theorem OAI.PixtonChow.colorWord_fiber_card line 179
  27. theorem OAI.PixtonChow.colorWord_stabilizer_card line 187
  28. abbrev OAI.PixtonChow.ExteriorCoefficients line 197

    Source documentation

    The exterior algebra on a freely specified set of generators.

    abbrev ExteriorCoefficients (I : Type*) := ExteriorAlgebra ℚ (I →₀ ℚ)
  29. def OAI.PixtonChow.theta line 200

    Source documentation

    The generator indexed by i.

    def theta {I : Type*} (i : I) : ExteriorCoefficients I :=
    …

    Mentions: OAI.PixtonChow.ExteriorCoefficients

  30. def OAI.PixtonChow.topExteriorProduct line 204

    Source documentation

    The product of all generators, in the given increasing order.

    def topExteriorProduct (I : Type*) [Fintype I] [LinearOrder I] :
        ExteriorCoefficients I :=
    …

    Mentions: OAI.PixtonChow.ExteriorCoefficients

  31. theorem OAI.PixtonChow.topExteriorProduct_eq_ordered_product line 210
  32. theorem OAI.PixtonChow.topExteriorProduct_eq_basis line 218

    Source documentation

    Identifying the ordered monomial with an actual exterior-algebra basis vector, not positing its nonvanishing.

    theorem topExteriorProduct_eq_basis (I : Type*) [Fintype I] [LinearOrder I] :
        topExteriorProduct I =
          (Finsupp.basisSingleOne : Module.Basis I ℚ (I →₀ ℚ)).ExteriorAlgebra Finset.univ := by
    …

    Mentions: OAI.PixtonChow.topExteriorProduct

  33. theorem OAI.PixtonChow.topExteriorProduct_ne_zero line 224
  34. theorem OAI.PixtonChow.factorial_multiple_topExteriorProduct_ne_zero line 231

    Source documentation

    The proposed detector's final scalar-multiple monomial is nonzero over ℚ. The crucial claim that it equals a pairing with t_n(Y) is not formalized.

    theorem factorial_multiple_topExteriorProduct_ne_zero (g : ℕ) (I : Type*)
        [Fintype I] [LinearOrder I] :
        (((((g - 1).choose 2).factorial : ℕ) : ℚ) ^ g) • topExteriorProduct I ≠ 0 := by
    …

    Mentions: OAI.PixtonChow.topExteriorProduct

  35. def OAI.PixtonChow.candidateTheta line 240

    Source documentation

    The monomial Θ indexed by all triples in lexicographic order, in the exterior algebra from Section 3.

    def candidateTheta (g : ℕ) : ExteriorCoefficients (Triple g) :=
    …

    Mentions: OAI.PixtonChow.ExteriorCoefficients, OAI.PixtonChow.Triple

  36. theorem OAI.PixtonChow.candidateTheta_ne_zero line 244
  37. theorem OAI.PixtonChow.candidateTheta_multiple_ne_zero line 250

    Source documentation

    The unsigned right side of eq:detector-candidate-value is nonzero. No equality with a value of the candidate class is claimed here.

    theorem candidateTheta_multiple_ne_zero (g : ℕ) :
        (((((g - 1).choose 2).factorial : ℕ) : ℚ) ^ g) • candidateTheta g ≠ 0 := by
    …

    Mentions: OAI.PixtonChow.candidateTheta

  38. abbrev OAI.PixtonChow.GraphFlag line 263

    Source documentation

    The two flags at each edge, together with the labeled legs.

    abbrev GraphFlag (E : Type*) (n : ℕ) := (E × Fin 2) ⊕ Fin n
  39. def OAI.PixtonChow.EdgeAdjacent line 266

    Source documentation

    Undirected adjacency in a graph presentation.

    def EdgeAdjacent {V E : Type*} {n : ℕ} (atVertex : GraphFlag E n → V)
        (u v : V) : Prop :=
    …

    Mentions: OAI.PixtonChow.GraphFlag

  40. structure OAI.PixtonChow.StableGraph line 274

    Source documentation

    A finite connected stable graph of type (g,n), given by vertices, edges, flag incidences, and nonnegative vertex genera. The integer genus equation avoids truncated subtraction in |E| - |V| + 1.

    structure StableGraph (g n : ℕ) (V E : Type*) [Fintype V] [Fintype E]
        [DecidableEq V] where
    …
  41. def OAI.PixtonChow.StableGraph.valence line 288

    Source documentation

    Number of flags, including legs, incident to a vertex.

    def valence (Γ : StableGraph g n V E) (v : V) : ℕ :=
    …

    Mentions: OAI.PixtonChow.StableGraph

  42. theorem OAI.PixtonChow.StableGraph.sum_valence line 291
  43. theorem OAI.PixtonChow.StableGraph.sum_stability line 298

    Source documentation

    Summing the stability weights cancels the individual vertex genera.

    theorem sum_stability (Γ : StableGraph g n V E) :
        ∑ v, (2 * (Γ.vertexGenus v : ℤ) - 2 + Γ.valence v) = 2 * g - 2 + n := by
    …

    Mentions: OAI.PixtonChow.StableGraph

  44. theorem OAI.PixtonChow.StableGraph.vertex_bound line 307

    Source documentation

    The usual bound on the number of vertices, including marked graphs.

    theorem vertex_bound (Γ : StableGraph g n V E) :
        (Fintype.card V : ℤ) ≤ 2 * g - 2 + n := by
    …

    Mentions: OAI.PixtonChow.StableGraph

  45. theorem OAI.PixtonChow.StableGraph.edge_bound line 319
  46. def OAI.PixtonChow.StableGraph.noiseCount line 326

    Source documentation

    R_Γ from lem:elementary-entry-tests.

    def noiseCount (_Γ : StableGraph g n V E) : ℕ :=
    …

    Mentions: OAI.PixtonChow.StableGraph

  47. theorem OAI.PixtonChow.StableGraph.noiseCount_bound line 330

    Source documentation

    The unmarked global noise bound eq:global-noise-bound.

    theorem noiseCount_bound (Γ : StableGraph g 0 V E) :
        (Γ.noiseCount : ℤ) ≤ 10 * g := by
    …

    Mentions: OAI.PixtonChow.StableGraph

  48. theorem OAI.PixtonChow.StableGraph.other_vertices_bound line 340

    Source documentation

    For any distinguished vertex, the first small-deficit bound. It does not require that this vertex have maximum genus.

    theorem other_vertices_bound (Γ : StableGraph g 0 V E) (v : V) :
        (Fintype.card V : ℤ) - 1 ≤ 2 * ((g : ℤ) - Γ.vertexGenus v) := by
    …

    Mentions: OAI.PixtonChow.StableGraph

  49. theorem OAI.PixtonChow.StableGraph.small_deficit_edge_bound line 360

    Source documentation

    The edge part of eq:small-graph-size.

    theorem small_deficit_edge_bound (Γ : StableGraph g 0 V E) (v : V) :
        (Fintype.card E : ℤ) ≤ 3 * ((g : ℤ) - Γ.vertexGenus v) := by
    …

    Mentions: OAI.PixtonChow.StableGraph

  50. theorem OAI.PixtonChow.StableGraph.small_deficit_noise_bound line 367

    Source documentation

    The noise budget when the genus deficit at a distinguished vertex is small.

    theorem small_deficit_noise_bound (Γ : StableGraph g 0 V E) (v : V) :
        (Γ.noiseCount : ℤ) ≤ 10 * ((g : ℤ) - Γ.vertexGenus v) + 2 := by
    …

    Mentions: OAI.PixtonChow.StableGraph

  51. def OAI.PixtonChow.ColorBudget.K line 382
  52. def OAI.PixtonChow.ColorBudget.blockSize line 384
  53. def OAI.PixtonChow.ColorBudget.blockCount line 386
  54. def OAI.PixtonChow.ColorBudget.C line 389

    Source documentation

    C_K, computed without recursion in the upper argument of choose.

    def C : ℕ := (2 * K).descFactorial 3 / 6

    Mentions: OAI.PixtonChow.ColorBudget.K

  55. def OAI.PixtonChow.ColorBudget.D₀ line 392

    Source documentation

    D_0, computed by the same falling-factorial formula.

    def D₀ : ℕ := blockSize.descFactorial 3 / 6
  56. theorem OAI.PixtonChow.ColorBudget.C_eq_choose line 394
  57. theorem OAI.PixtonChow.ColorBudget.D₀_eq_choose line 398
  58. theorem OAI.PixtonChow.ColorBudget.block_scale line 403

    Source documentation

    10^40 is the integer value required for g^(2/3) at this genus.

    theorem block_scale : (10 ^ 40 : ℕ) ^ 3 = genus ^ 2 := by
    …

    Mentions: OAI.PixtonChow.genus

  59. theorem OAI.PixtonChow.ColorBudget.reservation_bound line 407

    Source documentation

    Fewer than one tenth of the colors are needed for the reserved blocks.

    theorem reservation_bound : 10 * (blockCount * blockSize) < genus := by
    …

    Mentions: OAI.PixtonChow.ColorBudget.blockCount, OAI.PixtonChow.ColorBudget.blockSize, OAI.PixtonChow.genus

  60. theorem OAI.PixtonChow.ColorBudget.remaining_bound line 411

    Source documentation

    The remaining number of colors is strictly greater than 0.9 g.

    theorem remaining_bound : 9 * genus < 10 * (genus - blockCount * blockSize) := by
    …

    Mentions: OAI.PixtonChow.genus, OAI.PixtonChow.ColorBudget.blockCount, OAI.PixtonChow.ColorBudget.blockSize

  61. theorem OAI.PixtonChow.ColorBudget.C_bound line 415
  62. theorem OAI.PixtonChow.ColorBudget.block_entry_bound line 418
  63. theorem OAI.PixtonChow.ColorBudget.ordinary_small_cost line 421
  64. theorem OAI.PixtonChow.ColorBudget.small_loss_bounds line 426

    Source documentation

    The two numerical inequalities in eq:small-loss, with the second one multiplied by 100 to stay in natural numbers.

    theorem small_loss_bounds :
        3 * genus * (C + 1) + genus * (10 * K + 2) + 2 * K * C < 10 ^ 17 * genus ∧
          100 * (10 ^ 17 * genus) < D₀ := by
    …

    Mentions: OAI.PixtonChow.genus, OAI.PixtonChow.ColorBudget.C, OAI.PixtonChow.ColorBudget.K, OAI.PixtonChow.ColorBudget.D₀

  65. theorem OAI.PixtonChow.ColorBudget.expensive_count_bounds line 433

    Source documentation

    Bounds on the number of reserved blocks; no existence of an advancement is assumed or derived from these numerical comparisons.

    theorem expensive_count_bounds :
        4 * (genus - K - 2) > 3 * genus - 3 ∧ 3 * K + 5 < blockCount := by
    …

    Mentions: OAI.PixtonChow.genus, OAI.PixtonChow.ColorBudget.K, OAI.PixtonChow.ColorBudget.blockCount

  66. theorem OAI.PixtonChow.ColorBudget.ordinary_cost_bounds line 437