Documentation

TauCeti.FieldTheory.QuadraticForm.StiefelWhitney.Basic

The Stiefel–Whitney classes of a diagonal form in degrees one and two #

Let K be a field in which 2 is invertible, and write (a) ∈ H¹(G_K, 𝔽₂) for the Kummer class of a unit a (TauCeti.kummerClass). The Stiefel–Whitney classes of a diagonal quadratic form ⟨a₁, …, aₙ⟩ in degrees one and two are

w₁⟨a₁, …, aₙ⟩ = ∑ᵢ (aᵢ) ∈ H¹(G_K, 𝔽₂),     w₂⟨a₁, …, aₙ⟩ = ∑_{i<j} (aᵢ) ∪ (aⱼ) ∈ H²(G_K, 𝔽₂),

the degree-one and degree-two parts of Delzant's total class ∏ᵢ (1 + (aᵢ)). The cup product is the mod-two Kummer cup pairing TauCeti.kummerCup on square classes, so w₂ is visibly a function of the square classes of the coefficients.

This file defines both classes on tuples of units, computes them in low rank and on squares, and expands them on a concatenation of tuples, that is on an orthogonal sum of diagonal forms. Both classes are unchanged by permuting the coefficients, w₂ because the cup pairing is symmetric (TauCeti.kummerCup_comm). The class w₁ is the Kummer class of the plain discriminant a₁ ⋯ aₙ, so it is an invariant of the isometry class of the diagonal form. The invariance of w₂ under isometry, which needs the cup-norm theorem, and the descent of both classes to isometry classes of regular forms are in TauCeti.FieldTheory.QuadraticForm.StiefelWhitney.Class.

Main definitions #

Main results #

References #

noncomputable def TauCeti.sw1 {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) :

The first Stiefel–Whitney class of the diagonal form ⟨a₁, …, aₙ⟩: the sum w₁ = ∑ᵢ (aᵢ) of the Kummer classes of its coefficients, in H¹(G_K, 𝔽₂).

Instances For
    noncomputable def TauCeti.sw2 {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) :

    The second Stiefel–Whitney class of the diagonal form ⟨a₁, …, aₙ⟩: the sum w₂ = ∑_{i<j} (aᵢ) ∪ (aⱼ) of the cup products of the Kummer classes of pairs of distinct coefficients, in H²(G_K, 𝔽₂).

    Instances For
      theorem TauCeti.sw1_def {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) :
      sw1 w = ∑ i : Fin n, kummerClass (w i)

      Unfolding rule for TauCeti.sw1.

      theorem TauCeti.sw2_def {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) :
      sw2 w = ∑ i : Fin n, ∑ j > i, ((kummerCup K) (squareClass (w i))) (squareClass (w j))

      Unfolding rule for TauCeti.sw2.

      The first class is the Kummer class of the discriminant #

      theorem TauCeti.sw1_eq_kummerSquareClassEquiv {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) :
      sw1 w = (kummerSquareClassEquiv K) (squareClass (∏ i : Fin n, w i))

      w₁ read through the Kummer isomorphism on square classes: it is the image of the square class of the discriminant a₁ ⋯ aₙ.

      theorem TauCeti.sw1_eq_kummerClass_prod {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) :
      sw1 w = kummerClass (∏ i : Fin n, w i)

      w₁ is the Kummer class of the discriminant: w₁⟨a₁, …, aₙ⟩ = (a₁ ⋯ aₙ), with the plain discriminant a₁ ⋯ aₙ rather than the signed one.

      w₁ is an isometry invariant: isometric diagonal forms ⟨a₁, …, aₙ⟩ ≅ ⟨b₁, …, bₘ⟩ have the same first Stiefel–Whitney class, because their discriminants agree modulo squares.

      Invariance under permuting the coefficients #

      @[simp]
      theorem TauCeti.sw1_comp_perm {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) (σ : Equiv.Perm (Fin n)) :
      sw1 (w ∘ ⇑σ) = sw1 w

      w₁ does not depend on the order of the coefficients.

      @[simp]
      theorem TauCeti.sw2_comp_perm {K : Type u} [Field K] [Invertible 2] {n : ℕ} (w : Fin n → Kˣ) (σ : Equiv.Perm (Fin n)) :
      sw2 (w ∘ ⇑σ) = sw2 w

      w₂ does not depend on the order of the coefficients, because the cup pairing is symmetric.

      theorem TauCeti.PermutationStep.sw1_eq {K : Type u} [Field K] [Invertible 2] {n : ℕ} {w w' : Fin n → Kˣ} (h : PermutationStep w w') :
      sw1 w = sw1 w'

      A permutation step of a diagonal chain does not change w₁.

      theorem TauCeti.PermutationStep.sw2_eq {K : Type u} [Field K] [Invertible 2] {n : ℕ} {w w' : Fin n → Kˣ} (h : PermutationStep w w') :
      sw2 w = sw2 w'

      A permutation step of a diagonal chain does not change w₂.

      theorem TauCeti.BinaryStep.sw1_eq {K : Type u} [Field K] [Invertible 2] {n : ℕ} {w w' : Fin n → Kˣ} (h : BinaryStep w w') :
      sw1 w = sw1 w'

      A binary step of a diagonal chain does not change w₁, since its endpoints are isometric.

      Low rank #

      @[simp]
      theorem TauCeti.sw1_fin_zero {K : Type u} [Field K] [Invertible 2] (w : Fin 0 → Kˣ) :
      sw1 w = 0

      The first Stiefel–Whitney class of the rank-zero form vanishes.

      @[simp]
      theorem TauCeti.sw2_fin_zero {K : Type u} [Field K] [Invertible 2] (w : Fin 0 → Kˣ) :
      sw2 w = 0

      The second Stiefel–Whitney class of the rank-zero form vanishes.

      theorem TauCeti.sw2_eq_zero_of_le_one {K : Type u} [Field K] [Invertible 2] {n : ℕ} (hn : n ≤ 1) (w : Fin n → Kˣ) :
      sw2 w = 0

      w₂ vanishes in ranks 0 and 1: there are no pairs of distinct coefficients.

      @[simp]
      theorem TauCeti.sw1_fin_one {K : Type u} [Field K] [Invertible 2] (w : Fin 1 → Kˣ) :
      sw1 w = kummerClass (w 0)

      w₁⟨a⟩ = (a).

      @[simp]
      theorem TauCeti.sw2_fin_one {K : Type u} [Field K] [Invertible 2] (w : Fin 1 → Kˣ) :
      sw2 w = 0

      w₂⟨a⟩ = 0.

      @[simp]
      theorem TauCeti.sw1_fin_two {K : Type u} [Field K] [Invertible 2] (w : Fin 2 → Kˣ) :
      sw1 w = kummerClass (w 0) + kummerClass (w 1)

      w₁⟨a, b⟩ = (a) + (b).

      @[simp]
      theorem TauCeti.sw2_fin_two {K : Type u} [Field K] [Invertible 2] (w : Fin 2 → Kˣ) :

      w₂⟨a, b⟩ = (a) ∪ (b): the second Stiefel–Whitney class of a binary form is the cup product of the Kummer classes of its two coefficients.

      Orthogonal sums #

      @[simp]
      theorem TauCeti.sw1_append {K : Type u} [Field K] [Invertible 2] {n m : ℕ} (v : Fin m → Kˣ) (w : Fin n → Kˣ) :
      sw1 (Fin.append v w) = sw1 v + sw1 w

      w₁ is additive over orthogonal sums: w₁⟨a₁, …, aₘ, b₁, …, bₙ⟩ = w₁⟨a₁, …, aₘ⟩ + w₁⟨b₁, …, bₙ⟩.

      @[simp]
      theorem TauCeti.sw2_append {K : Type u} [Field K] [Invertible 2] {n m : ℕ} (v : Fin m → Kˣ) (w : Fin n → Kˣ) :

      The orthogonal-sum formula for w₂: w₂(q ⊥ r) = w₂(q) + w₂(r) + w₁(q) ∪ w₁(r) for the diagonal forms q = ⟨a₁, …, aₘ⟩ and r = ⟨b₁, …, bₙ⟩. The cross term collects the cups (aᵢ) ∪ (bⱼ) of the pairs with one coefficient in each form.

      Tuples of squares #

      theorem TauCeti.sw1_eq_zero_of_isSquare {K : Type u} [Field K] [Invertible 2] {n : ℕ} {w : Fin n → Kˣ} (hw : ∀ (i : Fin n), IsSquare (w i)) :
      sw1 w = 0

      w₁ vanishes on a tuple of squares.

      theorem TauCeti.sw2_eq_zero_of_isSquare {K : Type u} [Field K] [Invertible 2] {n : ℕ} {w : Fin n → Kˣ} (hw : ∀ (i : Fin n), IsSquare (w i)) :
      sw2 w = 0

      w₂ vanishes on a tuple of squares.

      @[simp]
      theorem TauCeti.sw1_one {K : Type u} [Field K] [Invertible 2] {n : ℕ} :
      sw1 1 = 0

      w₁⟨1, …, 1⟩ = 0.

      @[simp]
      theorem TauCeti.sw2_one {K : Type u} [Field K] [Invertible 2] {n : ℕ} :
      sw2 1 = 0

      w₂⟨1, …, 1⟩ = 0.