introvoyz045's picture
Migrated from GitHub
4a56cd6 verified
Raw History Blame Contribute Delete
5.03 kB
import OAI.AlgebraicGeometry.GroupSchemes.MultiplicativeLevels
import OAI.AlgebraicGeometry.GroupSchemes.PowerCoordinates
import OAI.AlgebraicGeometry.GroupSchemes.NontrivialMultiplication
import OAI.AlgebraicGeometry.GroupSchemes.EtaleCyclic
/-! # Dual Cyclic -/
namespace OAI.AlgebraicGeometry.GroupSchemes.DualCyclic
noncomputable section
/-! The split Cartier-dual levels in height one have cyclic rational-point groups. -/
open _root_.CategoryTheory _root_.AlgebraicGeometry Opposite WithConv _root_.CategoryTheory.MonObj
open OAI.AlgebraicGeometry.GroupSchemes.PowerCoordinates.FiniteEtale
namespace FiniteEtale
universe u
variable (k A : Type u) [Field k] [CommRing A] [HopfAlgebra k A]
[Coalgebra.IsCocomm k A] [Module.Finite k A]
variable (p : ℕ) [Fact p.Prime] [CharP k p] [PerfectRing k p]
local instance : CommRing (AbelianScheme.CartierDuality.DualAlgebra.Construction.DualCoordinate k A) := LinearMap.convCommRing
local instance : Algebra k (AbelianScheme.CartierDuality.DualAlgebra.Construction.DualCoordinate k A) := LinearMap.convAlgebra
/-- If a split dual's rational points are killed by p^n, so is multiplication on the original. -/
theorem constant_of_dual_points [Algebra.IsFiniteSplit k (AbelianScheme.CartierDuality.DualAlgebra.Construction.DualCoordinate k A)] (n : ℕ)
(hpoints : ∀ y : (FiniteEtale.FiniteCommGroupScheme.dualOfHopf (k := k) A).rationalPoints,
y ^ (p ^ n) = 1) (x : A) :
PowerCoordinates.FiniteEtale.powerMap k A (p ^ n) x = algebraMap k A (Coalgebra.counit (R := k) x) := by
have hpow : ∀ f : AbelianScheme.CartierDuality.DualAlgebra.Construction.DualCoordinate k A →ₐ[k] k, (toConv f) ^ (p ^ n) = 1 := by
intro f
let e := @Spec.mapMulEquiv k (AbelianScheme.CartierDuality.DualAlgebra.Construction.DualCoordinate k A) k _ _ _ inferInstance _
apply e.injective
exact (map_pow e (toConv f) (p ^ n)).trans ((hpoints (e (toConv f))).trans e.map_one.symm)
exact PowerCoordinates.FiniteEtale.powerMap_constant_of_dual k A p n
(fun f => PowerCoordinates.FiniteEtale.powerMap_constant_of_points k (AbelianScheme.CartierDuality.DualAlgebra.Construction.DualCoordinate k A) (p ^ n) hpow f) x
end FiniteEtale
namespace FiniteEtale.PDivisibleGroup
universe u
variable {k : Type u} [Field k] [IsAlgClosed k]
variable {p : ℕ} [Fact p.Prime] [CharP k p]
/-- The dual point group contains a point of order p^n when all dual levels split. -/
theorem _root_.OAI.AlgebraicGeometry.GroupSchemes.FiniteEtale.PDivisibleGroup.exists_dual_point_order (G : GroupSchemes.FiniteEtale.PDivisibleGroup k p 1)
(hsplit : ∀ n, Algebra.IsFiniteSplit k (G.level n).cartierDual.algebra) (n : ℕ) :
∃ x : (G.level n).cartierDual.rationalPoints, orderOf x = p ^ n := by
cases n with
| zero => exact ⟨1, by simp⟩
| succ n =>
have := hsplit (n + 1)
have hcard : Nat.card (G.level (n + 1)).cartierDual.rationalPoints = p ^ (n + 1) := by
rw [(G.level (n + 1)).cartierDual.rationalPoints_card,
(G.level (n + 1)).cartierDual_rank, G.rank, Nat.mul_one]
have hex : ∃ x : (G.level (n + 1)).cartierDual.rationalPoints, x ^ (p ^ n) ≠ 1 := by
by_contra h
push Not at h
have hzero := @constant_of_dual_points
k (G.level (n + 1)).algebra _ _ (G.level (n + 1)).coordinateHopf
(G.level (n + 1)).coordinateHopf_cocomm (G.level (n + 1)).finite
p _ _ _ (hsplit (n + 1)) n h
apply G.identity_power_ne_one n
have halg : (G.level (n + 1)).coordinatePower (p ^ n) =
(G.level (n + 1)).coordinateIdentity (G.level (n + 1)).algebra := by
ext x
exact hzero x
have hp := ((G.level (n + 1)).coordinatePoints_pow_eq_one_iff
(G.level (n + 1)).algebra (AlgHom.id k (G.level (n + 1)).algebra) (p ^ n)).mpr
(by simpa only [AlgHom.id_comp] using halg)
have hid : (G.level (n + 1)).coordinatePoints (G.level (n + 1)).algebra
(toConv (AlgHom.id k (G.level (n + 1)).algebra)) = (𝟙 (G.level (n + 1)).scheme) := by
rw [FiniteEtale.FiniteCommGroupScheme.coordinatePoints_apply]
exact (algSpec (CommRingCat.of k)).map_id _
exact (congrArg (fun z => z ^ (p ^ n)) hid).symm.trans hp
obtain ⟨x, hx⟩ := hex
refine ⟨x, orderOf_eq_prime_pow hx ?_⟩
simpa only [hcard] using (pow_card_eq_one' (x := x))
/-- Every split Cartier-dual level is cyclic in height one. -/
theorem _root_.OAI.AlgebraicGeometry.GroupSchemes.FiniteEtale.PDivisibleGroup.dual_rationalPoints_cyclic (G : GroupSchemes.FiniteEtale.PDivisibleGroup k p 1)
(hsplit : ∀ n, Algebra.IsFiniteSplit k (G.level n).cartierDual.algebra) (n : ℕ) :
IsCyclic (G.level n).cartierDual.rationalPoints := by
have := hsplit n
obtain ⟨x, hx⟩ := G.exists_dual_point_order hsplit n
apply isCyclic_of_orderOf_eq_card x
rw [hx, (G.level n).cartierDual.rationalPoints_card, (G.level n).cartierDual_rank,
G.rank, Nat.mul_one]
end FiniteEtale.PDivisibleGroup
end
end OAI.AlgebraicGeometry.GroupSchemes.DualCyclic