YukonModule.ProximityPrize.SubmissionLower.WeightedPlaceOrder6807.part0
DefinitionYukon_7ac04eb8d9104a96b1b2bba0better-codes
Source module ProximityPrize.SubmissionLower.WeightedPlaceOrder6807.
Definition code
import Definitions.Def_Yukon_9d89c768274d20327591ec1d
/-
UNCOMPILED. Ideal-power multiplicities at the original affine-model places.
No equality of multiplicities after base change is required. A point valuation
is nonpositive on the affine coordinate ring and strictly positive in order on
the point's maximal ideal. Ideal powers then give the full lower bound mu.
-/
set_option backward.isDefEq.respectTransparency.types false
namespace ProximityPrize.SubmissionLower.WeightedPlaceOrder6807
open scoped Classical BigOperators WithZero
open IsDedekindDomain
noncomputable section
set_option autoImplicit false
set_option maxRecDepth 20000
set_option maxHeartbeats 3000000
section ValuationFiltration
variable {A L : Type*} [CommRing A] [Field L]
(iota : A →+* L) (v : Valuation L (WithZero (Multiplicative ℤ)))
(hbound : ∀ a : A, v (iota a) ≤ 1)
/-- Functions having order at least n. All affine functions have order >= 0,
so this subset really is an ideal, not just an additive subgroup. -/
def orderIdeal (n : ℕ) : Ideal A where
carrier := {a | v (iota a) ≤ WithZero.exp (-(n : ℤ))}
zero_mem' := by simp
add_mem' := by
intro a b ha hb
change v (iota (a+b)) ≤ _
rw [map_add]
exact (v.map_add _ _).trans (max_le ha hb)
smul_mem' := by
intro r a ha
change v (iota (r*a)) ≤ _
rw [map_mul, map_mul]
simpa only [one_mul] using mul_le_mul' (hbound r) ha
theorem ideal_pow_value_le (J : Ideal A)
(hvanish : ∀ a ∈ J, v (iota a) < 1) (n : ℕ) :
J^n ≤ orderIdeal iota v hbound n := by
induction n with
| zero =>
intro a _
change v (iota a) ≤ WithZero.exp (-(0 : ℤ))
simpa only [neg_zero, WithZero.exp_zero] using hbound a
| succ n ih =>
rw [pow_succ]
apply Ideal.mul_le.mpr
intro a ha b hb
have h1 : v (iota b) ≤ WithZero.exp (-1 : ℤ) :=
RCN359.value_le_exp_neg_one (hvanish b hb)
have hn : v (iota a) ≤ WithZero.exp (-(n : ℤ)) := ih ha
change v (iota (a*b)) ≤ WithZero.exp (-((n+1 : ℕ) : ℤ))
rw [map_mul, map_mul]
have hm := mul_le_mul' hn h1
have he : WithZero.exp (-(n : ℤ)) * WithZero.exp (-1 : ℤ) =
WithZero.exp (-((n+1 : ℕ) : ℤ)) := by
rw [← WithZero.exp_add]
congr 1
push_cast
ring
exact hm.trans_eq he
theorem order_ge_of_value_le (x : L) (hx : x ≠ 0) (n : ℕ)
(hn : v x ≤ WithZero.exp (-(n : ℤ))) :
(n : ℤ) ≤ -(v x).log := by
have hv0 : v x ≠ 0 := (Valuation.ne_zero_iff v).mpr hx
have hlog : (v x).log ≤ -(n : ℤ) := by
have h := (WithZero.log_le_log hv0 (by simp)).2 hn
simpa only [WithZero.log_exp] using h
omega
include hbound in
/-- A denominator nonvanishing at the point has valuation exactly one.
It need not already have an inverse inside the affine coordinate ring. -/
theorem normalized_order_ge_pow {K : Type*} [Field K] [Algebra K A]
(phi : A →ₐ[K] K)
(hpoint : ∀ a, v (iota a) < 1 ↔ phi a = 0)
(n : ℕ) (a b : A) (ha : a ∈ (RingHom.ker phi.toRingHom)^n)
(hb : phi b ≠ 0) (hx : iota a / iota b ≠ 0) :
(n : ℤ) ≤ -(v (iota a / iota b)).log := by
have hbval : v (iota b) = 1 := by
apply le_antisymm (hbound b)
apply le_of_not_gt
intro h
exact hb ((hpoint b).mp h)
have han : v (iota a) ≤ WithZero.exp (-(n : ℤ)) :=
ideal_pow_value_le iota v hbound (RingHom.ker phi.toRingHom)
(fun a ha => (hpoint a).mpr ha) n ha
apply order_ge_of_value_le v _ hx n
simpa only [map_div₀, hbval, div_one] using han
end ValuationFiltration
section ActualAffinePlaces
variable (K A L : Type*) [Field K] [IsAlgClosed K]
[CommRing A] [IsDomain A] [Field L]
[Algebra K A] [Algebra K L] [Algebra A L] [IsFractionRing A L]
[Algebra (Polynomial K) A] [Algebra (Polynomial K) L] [Algebra (RatFunc K) L]
[IsScalarTower K (Polynomial K) A] [IsScalarTower K (Polynomial K) L]
[IsScalarTower K A L] [IsScalarTower (Polynomial K) A L]
[IsScalarTower (Polynomial K) (RatFunc K) L]
[FiniteDimensional (RatFunc K) L] [Algebra.IsSeparable (RatFunc K) L]
local instance _root_.ProximityPrize.SubmissionLower.WeightedPlaceOrder6807.instIsScalarTowerRatFunc_proximityPrize : IsScalarTower K (RatFunc K) L :=
IsScalarTower.of_algebraMap_eq fun a => by
rw [IsScalarTower.algebraMap_apply K (Polynomial K) (RatFunc K),
← IsScalarTower.algebraMap_apply (Polynomial K) (RatFunc K) L,
← IsScalarTower.algebraMap_apply K (Polynomial K) L]
/-- Multiplicity at the same actualPointPlace used by the pinned zero-counting
infrastructure. This does not posit a new low-cost abstract place. -/
theorem actual_normalized_order_ge_pow (phi : A →ₐ[K] K)
(n : ℕ) (a b : A) (ha : a ∈ (RingHom.ker phi.toRingHom)^n)
(hb : phi b ≠ 0)
(hx : algebraMap A L a / algebraMap A L b ≠ 0) :
(n : ℤ) ≤ RCN026.order K L (RCN344.modelPlace K L A phi)
(algebraMap A L a / algebraMap A L b) := by
exact normalized_order_ge_pow (algebraMap A L)
((RCN000.actualPointPlace K A L phi).valuation L)
(RCN000.actual_model_value_le_one K A L phi) phi
(RCN000.actual_model_value_lt_one_iff K A L phi) n a b ha hb hx
end ActualAffinePlaces
end
end ProximityPrize.SubmissionLower.WeightedPlaceOrder6807
Source