Finite Newton prefixes give a characteristic-polynomial obstruction
ProvedCouret.CayleyDoublet.abelianThree_charpoly_ne_of_newton_prefixcouret-programmecouret-unification-20261108
Let be a finite abelian group, , , and . Assume generate , is odd and equals , and the image of in has exact order three. Write for the rational directed Cayley matrices of supports and . Assume , . For each matrix let be its reversed characteristic coefficient (zero beyond dimension) and its power trace. Assume explicitly for both matrices and every that
Then .
Preamble
import Definitions.Def_Couret_CayleyTwoColourNewton_1eb4b626b6a5
import Mathlib
import Theorems.Thm_Couret_CayleyDoublet_abelianThree_canonical_odd_trace_difference
import Theorems.Thm_Couret_CayleyDoublet_abelianThree_trace_eq_before_third
import Theorems.Thm_Couret_CayleyDoublet_quotient_alpha_order_of_pair_generation
import Theorems.Thm_Couret_CayleyNewton_newton_first_coefficient_ne_of_trace_prefix
import Theorems.Thm_Couret_CayleyNewton_reverseCharpolyCoeff_congr
import Definitions.Def_Couret_CayleyTwoColourNewton_1eb4b626b6a5
/-!
COU-20 -- canonical order-three return in the odd cyclic quotient.
Research source, **UNCOMPILED** with pinned Lean 4.29.1.
No existing certified theorem has been modified.
The first lemma needs only the order of alpha and a bounded coordinate,
and is independent of any representation of a directed graph. The last
lemma translates its canonical witness back to the quotient G/<h>.
-/
set_option autoImplicit false
noncomputable section
namespace Couret.CayleyCanonical
end Couret.CayleyCanonical
namespace Couret.CayleyDoublet
open Couret.CayleyCanonical
variable {G : Type*} [AddCommGroup G] [DecidableEq G] [Fintype G]
Formal statement
theorem abelianThree_charpoly_ne_of_newton_prefix
(a d h : G) (ell : ℕ)
(hhalf : h + h = 0) (hne : h ≠ 0)
(hpair : AddSubgroup.closure ({a, h} : Set G) = ⊤)
(hodd : Odd (Nat.card (G ⧸ AddSubgroup.zmultiples h)))
(hell : 0 < ell)
(hsize : Nat.card (G ⧸ AddSubgroup.zmultiples h) = 3 * ell)
(hdorder : addOrderOf (d : G ⧸ AddSubgroup.zmultiples h) = 3)
(hS : ∀ k : ℕ, 0 < k → k ≤ ell →
Couret.CayleyNewton.NewtonTraceEquation
(Couret.CayleyNewton.reverseCharpolyCoeff
(abelianCayleyMatrix (AbelianSource a d h)))
(fun m => Matrix.trace
(abelianCayleyMatrix (AbelianSource a d h) ^ m)) k)
(hT : ∀ k : ℕ, 0 < k → k ≤ ell →
Couret.CayleyNewton.NewtonTraceEquation
(Couret.CayleyNewton.reverseCharpolyCoeff
(abelianCayleyMatrix (AbelianTarget a d h)))
(fun m => Matrix.trace
(abelianCayleyMatrix (AbelianTarget a d h) ^ m)) k) :
(abelianCayleyMatrix (AbelianSource a d h)).charpoly ≠
(abelianCayleyMatrix (AbelianTarget a d h)).charpoly := by sorry
end Couret.CayleyDoublet
Source
Alexandre Couret, https://github.com/alexcour/couret-finite-verifications/blob/1483e8ae3e5491ee176364fe944c90c4341d40ff/research/correspondants_20261013/lean/CayleyTwoColourCanonicalReturn.lean#L195-L245, lines 195–245; source file SHA-256 d42d32ebf171bc2a72aa60523719dbdaaf3f0930bd30560960b5a8d6dae8cfa9; theorem Couret.CayleyDoublet.abelianThree_charpoly_ne_of_newton_prefix