Peterfalvi §13: subcoherence inputs for the setup
DefinitionOddOrder_Peterfalvi_S15_SAndT_Setup_SubcoherenceInputsfeit-thompsonfinite-groupsgroup-theoryishida-odd-order
Formalizes subcoherence inputs for the hypothesis of Peterfalvi §13 on the pair , along with general finite-set and cardinality helper definitions.
Main declarations: finNeg (Negation i ↦ -i ≡ (n − i) (mod n) on Fin n, for 0 < n); Hypothesis (Peterfalvi (13.1)); H (Peterfalvi's H = P C in (13.5)--(13.10)); K (Peterfalvi's K = Q D, used symmetrically for T).
The bundle also proves 12 supporting lemmas used by later statements.
Definition code
/-
Copyright (c) 2026 Yawara Ishida. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Yawara Ishida
-/
-- From github.com/yawara/odd-order @ 20a6fccbb63c (Apache-2.0; LICENSE at that commit):
-- OddOrder/Peterfalvi/S15_SAndT_Setup/SubcoherenceInputs.lean
-- Generated for Prove2Me by deleting declarations; the remaining commands are verbatim except:
-- `private` dropped where another file uses the declaration, and a solution's main theorem
-- renamed `_root_.solution` (platform convention).
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_Hypothesis
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_RhoConstancy
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_PairCoherence
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_FrobeniusStructure
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_MinimalCounterexample
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_DadeContradiction
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_RhoMEvaluation
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_NormPackage
import Definitions.Def_OddOrder_Peterfalvi_S14_MaximalI_TypeICovering
import Definitions.Def_OddOrder_Peterfalvi_S07_Subcoherent
import Definitions.Def_OddOrder_Peterfalvi_S10_CoherenceWiring
import Definitions.Def_OddOrder_Peterfalvi_S09_CertificateDischarge
import Definitions.Def_OddOrder_Isaacs_Ch06_FrobeniusActions_OddComplement
import Definitions.Def_OddOrder_GroupTheory_MaximalSubgroupTypeConj
import Definitions.Def_OddOrder_GroupTheory_WielandtFixedPoint
import Definitions.Def_OddOrder_Algebra_GaloisRationalInteger
import Definitions.Def_OddOrder_GroupTheory_TISubset
import Definitions.Def_OddOrder_GroupTheory_ConjClassSet
import Definitions.Def_OddOrder_GroupTheory_RepresentationTheory_InducedCharacter
import Mathlib.GroupTheory.QuotientGroup.Basic
import Mathlib.GroupTheory.Index
import Mathlib.Algebra.BigOperators.GroupWithZero.Action
import Mathlib.Algebra.BigOperators.ModEq
import Mathlib.FieldTheory.Finite.Basic
import Mathlib.RingTheory.Polynomial.Cyclotomic.Roots
namespace OddOrder.Peterfalvi.S15
open OddOrder.GroupTheory
open OddOrder.RepresentationTheory
open OddOrder.Isaacs
open scoped Pointwise
variable {G : Type*} [Group G]
/-- Negation `i ↦ -i ≡ (n − i) (mod n)` on `Fin n`, for `0 < n`. This is the index map
realizing the conjugation pairing `(i,j) ↦ (−i,−j)` of the Dade `η`-grid in Peterfalvi
(3.9.a)/(14.11.3). Defined identically to the (downstream) `S16.finNeg`, so the two are
definitionally equal — the `(3.9.a)` field `eta_pair_of_coprime` below (stated with this
`finNeg`) is thus citeable to fill `S16.EtaGenericData.eta_pair` (stated with `S16.finNeg`). -/
def finNeg {n : ℕ} (hn : 0 < n) (i : Fin n) : Fin n :=
⟨(n - i.val) % n, Nat.mod_lt _ hn⟩
namespace FiniteInduce
noncomputable scoped instance finiteSubFintype [Finite G] (H : Subgroup G) :
Fintype ↥H := Fintype.ofFinite _
noncomputable scoped instance natCardInvC [Finite G] (H : Subgroup G) :
Invertible (Nat.card H : ℂ) :=
invertibleOfNonzero (Nat.cast_ne_zero.mpr Nat.card_pos.ne')
noncomputable scoped instance ambientFintype [Finite G] : Fintype G :=
Fintype.ofFinite _
noncomputable scoped instance ambientNatCardInvC [Finite G] :
Invertible (Nat.card G : ℂ) :=
invertibleOfNonzero (Nat.cast_ne_zero.mpr Nat.card_pos.ne')
end FiniteInduce
open scoped FiniteInduce in
/-- **Peterfalvi (13.1)**: the main setup for the two maximal subgroups `S` and
`T` in case (b) of Theorem (8.8).
`G` is finite (carried as the instance field `finiteG`, so consumers need not
re-assume `[Finite G]`); this finiteness is what makes the canonical inductions
`Ind_W^S`/`Ind_W^T` in `mu_definition`/`nu_definition` well-defined. -/
structure Hypothesis where
[finiteG : Finite G]
S : Subgroup G
T : Subgroup G
W1 : Subgroup G
W2 : Subgroup G
W : Subgroup G
P : Subgroup G
Q : Subgroup G
U : Subgroup G
V : Subgroup G
C : Subgroup G
D : Subgroup G
S_maximal : S ∈ maximalSubgroups G
T_maximal : T ∈ maximalSubgroups G
S_ne_T : S ≠ T
S_nonI : IsTypeNonI S
T_nonI : IsTypeNonI T
one_typeII : IsTypeII S ∨ IsTypeII T
theorem88_caseB :
∀ M : Subgroup G, M ∈ maximalSubgroups G →
IsTypeI M ∨ (∃ g : G, MulAut.conj g • M = S) ∨
(∃ g : G, MulAut.conj g • M = T)
W_eq_inter : W = S ⊓ T
W_eq_join : W = W1 ⊔ W2
W1_inf_W2_eq_bot : W1 ⊓ W2 = ⊥
W1_commutes_W2 : ∀ x ∈ W1, ∀ y ∈ W2, Commute x y
W_cyclic : IsCyclic ↥W
P_eq_SF : P = maxNilpotentNormalHall S
Q_eq_TF : Q = maxNilpotentNormalHall T
S_deriv_eq_PU : derivedInG S = P ⊔ U
T_deriv_eq_QV : derivedInG T = Q ⊔ V
/-- **Peterfalvi (13.1.b)/(13.2), T-side complement disjointness**: `Q ⊓ V = ⊥`. The κ-Hall
invariant complement `V` to `Q = T_F` in `T'` genuinely *complements* `Q` — a fact available
ungated at the §16 construction (`exists_kappaHall_invariant_complement_to_MF` returns
`M_F ⊓ U = ⊥` from `T`'s type-`P` structure, i.e. from `T_nonI`, **not** requiring
(14.9)/`IsTypeP2 T`),
but dropped by the abstract `T_deriv_eq_QV` (`T' = Q ⊔ V`, which alone does not force
disjointness).
Threading it exposes the T-side linchpin used by every V-side coprimality/complement argument
(`isMulCommutative_V`, `coprime_card_V_card_Q_of_disjoint`, …), previously routed circularly
through
the then-sorried `reconciled_typePData_T` (`Q_inf_V_eq_bot_of_reconciled`). -/
Q_inf_V_eq_bot : Q ⊓ V = ⊥
/-- **Peterfalvi (13.1.b), T-side semidirect complement `T = T' ⋊ W₂`**: the cyclic factor `W₂`
complements the derived subgroup `T'` in `T`. Available ungated at the §16 construction from
`typeP_derivedInG_isComplement_kappaHall` (BG 14.7(h), via `T`'s type-`P` structure `T_nonI`,
**not** (14.9)); the abstract Hypothesis otherwise omits it. Discharges the `M_complement` field
of `reconciled_typePData_T` and lets the ungated `coprime_card_Q_card_VW2` build honestly. -/
W2_isComplement_T_deriv :
Subgroup.IsComplement' ((derivedInG T).subgroupOf T) (W2.subgroupOf T)
C_eq : C = U ⊓ Subgroup.centralizer (P : Set G)
D_eq : D = V ⊓ Subgroup.centralizer (Q : Set G)
W1_normalizes_U : W1 ≤ Subgroup.normalizer (U : Set G)
W2_normalizes_V : W2 ≤ Subgroup.normalizer (V : Set G)
q : ℕ
p : ℕ
q_prime : q.Prime
p_prime : p.Prime
q_odd : Odd q
p_odd : Odd p
q_eq_card_W1 : q = Nat.card ↥W1
p_eq_card_W2 : p = Nat.card ↥W2
u : ℕ
v : ℕ
c : ℕ
d : ℕ
c_eq_card_C : c = Nat.card ↥C
d_eq_card_D : d = Nat.card ↥D
card_U_eq_uc : Nat.card ↥U = u * c
card_V_eq_vd : Nat.card ↥V = v * d
Sset : Set (ClassFunction ↥S ℂ)
Tset : Set (ClassFunction ↥T ℂ)
A0S : Set ↥S
A0T : Set ↥T
tauS : OddOrder.Peterfalvi.S07.IntegralCharacterMap ↥S G
tauT : OddOrder.Peterfalvi.S07.IntegralCharacterMap ↥T G
omega : Fin q → Fin p → ClassFunction ↥W ℂ
eta : Fin q → Fin p → ClassFunction G ℂ
mu : Fin q → Fin p → ClassFunction ↥S ℂ
nu : Fin q → Fin p → ClassFunction ↥T ℂ
delta : Fin p → ℤ
deltaPrime : Fin q → ℤ
/-- **Peterfalvi (13.1.e)**: the signs `δ_j`, `δ'_i` are `±1` (the (4.3.b) `sign_eq`). -/
delta_pm_one : (∀ j : Fin p, delta j = 1 ∨ delta j = -1) ∧
(∀ i : Fin q, deltaPrime i = 1 ∨ deltaPrime i = -1)
/-- **Peterfalvi (4.3.d)**: the degree congruence `μ_{ij}(1) ≡ δ_j (mod q)`
(`μ_{ij}(1) = δ_j + q·a` for an integer `a`, the `Res_{W₁}` value identity). -/
mu_degree_modEq_delta : ∀ (i : Fin q) (j : Fin p), ∃ a : ℤ,
mu i j 1 = (delta j : ℂ) + (q : ℂ) * (a : ℂ)
/-- **Peterfalvi (4.4), the `S`-side base sign**: `δ_0 = 1` (`μ_{00} = 1_S`, the trivial
column's `σ`-anchor). -/
delta_zero_eq_one : delta ⟨0, p_prime.pos⟩ = 1
/-- The Peterfalvi (3.2)/(3.3) transfer map `τ`, typed as an integral
(virtual-character) map via the same `IntegralCharacterMap` convention as
`tauS`/`tauT` — faithful to `τ` being defined on the `ℤ`-lattice of virtual
characters. Pinning `tau3` to the *concrete* (3.2) Dade isometry (built from a
`S05.TICyclicHypothesis` for `W` through `S04.dadeIntegralCharacterMap`, with the
`ω`-grid materialized as the (3.3) characters) is a dedicated §3/§5 construction;
`eta_eq_tau_omega` already forces the `η`-grid to be the integral-linear image of
the `ω`-grid. -/
tau3 : OddOrder.Peterfalvi.S07.IntegralCharacterMap ↥W G
/-- **Peterfalvi (13.1.d)**: `η_{ij} = ω_{ij}^τ`. -/
eta_eq_tau_omega : ∀ (i : Fin q) (j : Fin p), eta i j = tau3 (omega i j)
/-- **Peterfalvi (13.1.e)**: `Ind_W^S (ω_{ij} − ω_{0j}) = δ_j (μ_{ij} − μ_{0j})`,
where the induction is the canonical `Ind_W^S = ClassFunction.induce (W.subgroupOf S)`
(transporting the `W`-grid into `S` via `W ≤ S`) and `δ_j = ±1` is `delta`. -/
mu_definition : ∀ (i : Fin q) (j : Fin p),
ClassFunction.induce (W.subgroupOf S)
(ClassFunction.compHom
(Subgroup.subgroupOfEquivOfLe
((le_of_eq W_eq_inter).trans inf_le_left)).toMonoidHom
(omega i j - omega ⟨0, q_prime.pos⟩ j))
= (delta j : ℂ) • (mu i j - mu ⟨0, q_prime.pos⟩ j)
/-- **Peterfalvi (13.1.e), irreducibility**: each `μ_{ij}` is an irreducible character of `S`
(the (1.4)/(4.3.b) column families consist of irreducibles; threaded from the
`SignedIrreducibleDifferenceFamily` producer, issue 2035 μ-linkage). -/
mu_irreducible : ∀ (i : Fin q) (j : Fin p),
OddOrder.RepresentationTheory.IsIrreducibleCharacter (mu i j)
/-- **Peterfalvi (13.1.e), column distinctness**: within a column `j` the `μ_{ij}` are
pairwise distinct (the (1.4) family `injective` field). Makes the column sums
`μ_j = ∑_i μ_{ij}` sums of `q ≥ 2` distinct irreducibles — hence reducible, the (13.3.a)
entry condition. -/
mu_col_injective : ∀ j : Fin p, Function.Injective (fun i : Fin q => mu i j)
/-- **Peterfalvi (4.3.b), full-grid orthonormality** (issues 9076/9014, the `(13.18)`
grounding): the `μ_{ij}` are pairwise-distinct irreducibles across the *whole* grid,
`⟨μ_{ij}, μ_{kl}⟩ = [(i,j) = (k,l)]`. Within a column this refines `mu_irreducible` +
`mu_col_injective`; the cross-column part is the prime-TI residue-grid fact
(S06 `columnFamily_mu_ne` / `PrimeTIResidueData.mu2_orthonormal`), supplied by the
producer from `Section16CharacterData.muS_orthonormal`. Discharges the `(13.18)` pin
`mu_row0_ne` (row-`0` cross-column distinctness, `S15_HonestTypeP2A0`). -/
mu_orthonormal : ∀ (i k : Fin q) (j l : Fin p),
OddOrder.RepresentationTheory.ClassFunction.inner (mu i j) (mu k l)
= if i = k ∧ j = l then 1 else 0
/-- **Peterfalvi (4.5.a) for the `S`-grid** (issue 2035 μ-linkage): each `μ`-column sum is
induced from an irreducible character of `S' = [S,S]` — the §4 certain-type identity
`μ_j = Ind_{S'}^S χ_j` (`induce_restrict_certainType_eq`), the constructive membership
shape behind (13.3.a)'s `μ_j ∈ 𝒮(H₀)`. -/
mu_colSum_eq_induce : ∀ j : Fin p,
∃ ψ : ClassFunction ↥((derivedInG S).subgroupOf S) ℂ,
OddOrder.RepresentationTheory.IsIrreducibleCharacter ψ ∧
(∑ i : Fin q, mu i j) = ClassFunction.induce ((derivedInG S).subgroupOf S) ψ ∧
(j ≠ ⟨0, p_prime.pos⟩ →
¬ (((W2.subgroupOf S).subgroupOf ((derivedInG S).subgroupOf S) :
Set ↥((derivedInG S).subgroupOf S)) ⊆
OddOrder.Peterfalvi.S03.characterKernel ψ))
/-- **Peterfalvi (9.8)/(9.11) reverse dichotomy (S-side R-family gate)** (issue 9092): a
*reducible* member of the general kernel-filter family `S(X)` over `S' = derivedInG S` (any
kernel demand `X`) is a nonzero `μ`-column sum `∑ᵢ μ_{ij}`, `j ≠ 0`. This is the S-side
analogue of the M-side `reducible_mem_inducedKernelFamily_eq_muGrid_columnSum` (`S12_HcBound`);
it is *supplied by the producer* from the §6 certain-type identity
`μ = certainTypeS.columnFamily.mu` (the abstract `Hypothesis` cannot see this identification,
so the reverse dispatch is threaded as a field). It is consumed by the (Clifford-case-agnostic)
per-member `R`-family (`sSet_memberRFamily`): a reducible `η ∈ 𝒮` bridges into `S(⊥)` — the `𝒳`
condition `¬(H ⊆ Ker)` forces `η`'s source nontrivial — and is dispatched to its μ-column `j`,
whose §6 `certainTypeR` image family is the (5.2.d) `R`-datum. -/
mu_reducible_dichotomy : ∀ {X : Subgroup ↥S} {ψ : ClassFunction ↥S ℂ},
ψ ∈ OddOrder.Peterfalvi.S08.inducedKernelFamily ((derivedInG S).subgroupOf S) X →
¬ OddOrder.RepresentationTheory.IsIrreducibleCharacter ψ →
∃ j : Fin p, j ≠ ⟨0, p_prime.pos⟩ ∧ ψ = ∑ i : Fin q, mu i j
/-- **Peterfalvi (13.1.e)**: `Ind_W^T (ω_{ij} − ω_{i0}) = δ'_i (ν_{ij} − ν_{i0})`,
with the canonical `Ind_W^T = ClassFunction.induce (W.subgroupOf T)` and
`δ'_i = ±1` is `deltaPrime`. -/
nu_definition : ∀ (i : Fin q) (j : Fin p),
ClassFunction.induce (W.subgroupOf T)
(ClassFunction.compHom
(Subgroup.subgroupOfEquivOfLe
((le_of_eq W_eq_inter).trans inf_le_right)).toMonoidHom
(omega i j - omega i ⟨0, p_prime.pos⟩))
= (deltaPrime i : ℂ) • (nu i j - nu i ⟨0, p_prime.pos⟩)
m : ℚ
m_eq : m = 1 - 1 / ((q : ℚ) - 1) - ((q : ℚ) - 1) / (q : ℚ) ^ p +
1 / (((q : ℚ) - 1) * (q : ℚ) ^ p)
/-- **Peterfalvi (13.1.b) carrier (S-side)**: the type-`P` structure data of `S`
(`S = (P ⋊ U) ⋊ W₁`), with its complement `U` and cyclic factor `W₁` reconciled to the
hypothesis's `U`/`W1` (`Sdata_U_eq`/`Sdata_W1_eq`). This is the §16-construction witness
(`Section16TypePStructure.Sdata`, built from `mp.S_typeP2` = Pf (13.2.a)) that pins the otherwise
unconstrained relationship between the abstract `Hypothesis` fields `U`/`W1`/`P` and the intrinsic
type-`P` decomposition of `S`. It supplies the U-side structural facts (`U` complements
`M_F = P`, `W₁ ≤ N_G(U)`, `U` nilpotent) that Peterfalvi §15 reads off `S`. -/
Sdata : TypePData S
Sdata_U_eq : Sdata.U = U
Sdata_W1_eq : Sdata.W1 = W1
/-- **Peterfalvi (13.2.a), U-side**: the complement `U` is abelian. Supplied at construction from
BG Lemma 15.1(b) (`typeP_hall_derived_eq_and_abelian`, with `U` the `(κ∪σ)'`-Hall of `S`), so §15
need not re-derive it; discharges the `U_commutative` field of `basic_structure_gated`. -/
S_U_commutative : IsMulCommutative ↥U
/-- **Reconciliation (W₂-side)**: the intrinsic dual factor `Sdata.W2 = C_{S'}(W₁#)` equals the
abstract `W₂` (complement of `W₁` in the cyclic `W = S ∩ T`). Supplied at construction as
`Sdata.W2 = K*` (`typePData_of_kappaHall_hallComplement_W2`); discharges `card_P_eq` and hence
`basic_structure_gated.P_order` (issue 3001/4014). -/
Sdata_W2_eq : Sdata.W2 = W2
/-- **Peterfalvi (4.8), `μ`-column-difference support** (issues 9076/9014, the `(13.18)`
support grounding; Coq `prDade_sub_TIirr_on`): for nontrivial equal-degree columns
`j, k ≠ 0`, the difference `μ_{ij} − μ_{ik}` is supported in `A₀(S) = A(S) ∪ V^S`.
Supplied by the producer from `Section16CharacterData.muS_diff_support` (the §6 support
engine `certainType_diff_supp_subset_A0` at the `muS`-instance `Hypothesis46` `hyp46Smp`).
Discharges the `(13.18)` support pin `tauS_mu_row0_diff_support` (`S15_HonestTypeP2A0`). -/
mu_diff_support : ∀ (i : Fin q) {j k : Fin p},
j ≠ ⟨0, p_prime.pos⟩ → k ≠ ⟨0, p_prime.pos⟩ →
mu i j 1 = mu i k 1 →
(mu i j - mu i k).support ⊆
OddOrder.Peterfalvi.S04.supportInSubgroup (S10.typePACore0 S Sdata) S
/-- **Peterfalvi (4.3.c), value identity** (Coq `prTIirr_id`, `PFsection4.v:403`; issues
9076/9014, the `(13.18)` V-value grounding): on `W ∖ W₂` the `μ`-grid is the signed
`ω`-grid, `μ_{ij}(w) = δ_j·ω_{ij}(w)`. Supplied by the producer from the §6 value
identity `certainType_apply_eq_of_mem_V` (the (1.3)/(4.3.c) value-match of the
certain-type construction — Dade-free, base `S06.Hypothesis` level). With `δ_j = 1`
(`delta_eq_one_S`, Pf (13.3.c)) this becomes the prime-TI restriction identity
(`μ_{0j}|_{W∖W₂} = ω`-value) feeding the `(13.18)` V-value pin
`tauS_mu_row0_vanish_on_V`. -/
mu_apply_of_not_mem_W2 : ∀ (i : Fin q) (j : Fin p) (w : G) (hwW : w ∈ W)
(hwS : w ∈ S), w ∉ (W2 : Set G) →
mu i j ⟨w, hwS⟩ = (delta j : ℂ) * omega i j ⟨w, hwW⟩
/-- **Peterfalvi (4.9)(a), CF-level conjugation symmetry of the `μ`-grid** (Coq
`prTIirr_aut`; the full-CF strengthening of the generic-element `eta_pair_of_coprime`
pattern): `μ̄_{ij} = μ_{−i,−j}`. Supplied by the producer from `muS_conj`
(the §6 conjugation bridge `certainType_mu_conj_eq` with the power-enumeration index
bridges). Feeds the (13.18.c) reality of `Γ` (`gammaGrid_real`). -/
mu_conj : ∀ (i : Fin q) (j : Fin p),
(mu i j).conj = mu (finNeg q_prime.pos i) (finNeg p_prime.pos j)
/-- **Peterfalvi (3.9.a), CF-level conjugation symmetry of the `η`-grid** (Coq
`cfAut_cycTIiso`): `η̄_{ij} = η_{−i,−j}`. Supplied by the producer from
`tau3W_omegaS_conj` (`σ` Galois-intertwining `sigma_mapRingEquiv_comm` + linear-character
inversion `galoisMap_conj_omega`, through the power enumerations). Feeds the (13.18.c)
reality of `Γ` (`gammaGrid_real`). -/
eta_conj : ∀ (i : Fin q) (j : Fin p),
(eta i j).conj = eta (finNeg q_prime.pos i) (finNeg p_prime.pos j)
/- ### Grid property fields (issue 3002)
The (3.2)/(3.3)/(3.4) character-theoretic content of the Dade grid carriers `tau3`/`omega`,
which the §15 norm cascade ((13.5)–(13.10)) consumes. All are supplied at construction from
the honest spine grid (`Section16CharacterData.omegaS`/`tau3W` in `FeitThompson.lean`):
ω-orthonormality via `S05.TICyclicHypothesis.omega_inner`, the τ₃ facts via the (3.2)
σ-isometry package (`S05.TICyclicHypothesis.sigmaIntegral_*`). -/
/-- **Peterfalvi (3.2), isometry part**: `τ₃` preserves the class-function inner product. -/
tau3_isometry : OddOrder.Peterfalvi.S07.IsIntegralIsometry tau3
/-- **Peterfalvi (3.2)**: `τ₃` sends the trivial character to the trivial character. -/
tau3_trivial : tau3 (trivialClassFunction ↥W) = trivialClassFunction G
/-- **Peterfalvi (3.2.c)**: on the regular set `W ∖ (W₁ ∪ W₂)` the map `τ₃` is the
identity: `(α^{τ₃})(w) = α(w)` for regular `w`. -/
tau3_apply_of_regular : ∀ (α : ClassFunction ↥W ℂ) (w : G) (hwW : w ∈ W),
w ∉ (W1 : Set G) ∪ (W2 : Set G) → tau3 α w = α ⟨w, hwW⟩
/-- **Peterfalvi (3.2)**: `τ₃` sends virtual characters to virtual characters. -/
tau3_mem_ZIrr : ∀ z ∈ ZIrr ↥W, tau3 z ∈ ZIrr G
/-- **Peterfalvi (3.3)/(3.4)**: the `ω`-grid is orthonormal. -/
omega_orthonormal : ∀ (i k : Fin q) (j l : Fin p),
ClassFunction.inner (omega i j) (omega k l) = if i = k ∧ j = l then 1 else 0
/-- The `ω_{ij}` are linear characters: `ω_{ij}(1) = 1`. -/
omega_apply_one : ∀ (i : Fin q) (j : Fin p), omega i j 1 = 1
/-- Each `ω_{ij}` is a virtual character (in fact an irreducible character of `W`). -/
omega_mem_ZIrr : ∀ (i : Fin q) (j : Fin p), omega i j ∈ ZIrr ↥W
/-- **Peterfalvi (3.3)** (issue 2033): each `ω_{ij}` is multiplicative — a linear character. -/
omega_mul : ∀ (i : Fin q) (j : Fin p) (w w' : ↥W),
omega i j (w * w') = omega i j w * omega i j w'
/-- **Peterfalvi (3.3)** (issue 2033): the column-`0` characters `ω_{i0}` are trivial on `W₂`. -/
omega_col_zero_apply_of_mem_W2 : ∀ (i : Fin q) (w : ↥W), (w : G) ∈ W2 →
omega i ⟨0, p_prime.pos⟩ w = 1
/-- **Peterfalvi (3.3)** (issue 2033): the row-`0` characters `ω_{0j}` are trivial on `W₁`. -/
omega_row_zero_apply_of_mem_W1 : ∀ (j : Fin p) (w : ↥W), (w : G) ∈ W1 →
omega ⟨0, q_prime.pos⟩ j w = 1
/-- **Peterfalvi (3.3)** (issue 2033): on `W₁` the grid values are `q`-th roots of unity. -/
omega_pow_q_of_mem_W1 : ∀ (i : Fin q) (j : Fin p) (w : ↥W), (w : G) ∈ W1 →
omega i j w ^ q = 1
/-- **Peterfalvi (3.3)** (issue 2033): on `W₂` the grid values are `p`-th roots of unity. -/
omega_pow_p_of_mem_W2 : ∀ (i : Fin q) (j : Fin p) (w : ↥W), (w : G) ∈ W2 →
omega i j w ^ p = 1
/-- **Peterfalvi (3.2.d)** (issue 2034): a class function of `G` orthogonal to the whole
`τ₃ω`-grid vanishes on the regular set `Ŵ = W ∖ (W₁ ∪ W₂)` — the grid enumerates the
`σ`-image, and every irreducible off the image vanishes on `Ŵ`. -/
eta_complete_vanish : ∀ χ : ClassFunction G ℂ,
(∀ (i : Fin q) (j : Fin p), ClassFunction.inner (tau3 (omega i j)) χ = 0) →
∀ w : G, w ∈ W → w ∉ (W1 : Set G) ∪ (W2 : Set G) → χ w = 0
/-- **Peterfalvi (3.4)/(3.5), the four-corner vanishing** (issue 2036): off the conjugacy
saturation of the regular set `Ŵ = W ∖ (W₁ ∪ W₂)`, the `(3.5)` relation collapses to
`1 − (τ₃ω)_{i0}(x) − (τ₃ω)_{0j}(x) + (τ₃ω)_{ij}(x) = 0` for nonzero row/column indices. -/
eta_fourcorner_vanish : ∀ (i : Fin q) (j : Fin p), i ≠ ⟨0, q_prime.pos⟩ →
j ≠ ⟨0, p_prime.pos⟩ → ∀ x : G,
x ∉ OddOrder.GroupTheory.conjClassSet ((W : Set G) \ ((W1 : Set G) ∪ (W2 : Set G))) →
(1 : ℂ) - tau3 (omega i ⟨0, p_prime.pos⟩) x - tau3 (omega ⟨0, q_prime.pos⟩ j) x
+ tau3 (omega i j) x = 0
/-- **Peterfalvi (3.9.b), row-vanishing transport** (issue 2036): if `(τ₃ω)₁₀` vanishes at
`x`, so do all `(τ₃ω)_{i0}` with `i ≠ 0` — the nontrivial row characters are Galois-conjugate
powers of `ω₁₀`, and the twist fixes the vanishing value. -/
eta_row_vanish_of_one_zero : ∀ x : G,
tau3 (omega ⟨1, q_prime.one_lt⟩ ⟨0, p_prime.pos⟩) x = 0 →
∀ i : Fin q, i ≠ ⟨0, q_prime.pos⟩ → tau3 (omega i ⟨0, p_prime.pos⟩) x = 0
/-- **Peterfalvi (3.9.b), full row-axis Galois orbit.** -/
eta_row_galois_orbit : ∀ i : Fin q, i ≠ ⟨0, q_prime.pos⟩ →
∃ u : ℂ ≃+* ℂ,
ClassFunction.mapRingEquiv u (tau3 (omega ⟨1, q_prime.one_lt⟩ ⟨0, p_prime.pos⟩)) =
tau3 (omega i ⟨0, p_prime.pos⟩)
/-- **Peterfalvi (3.9.b), full column-axis Galois orbit.** -/
eta_column_galois_orbit : ∀ j : Fin p, j ≠ ⟨0, p_prime.pos⟩ →
∃ u : ℂ ≃+* ℂ,
ClassFunction.mapRingEquiv u (tau3 (omega ⟨0, q_prime.pos⟩ ⟨1, p_prime.one_lt⟩)) =
tau3 (omega ⟨0, q_prime.pos⟩ j)
/-- **Peterfalvi (3.9.c), the `η`-grid integrality on generic elements** (issue-3002
keystone): for `g` of order prime to `pq`, each grid value `η_{ij}(g) = (τ₃ω)_{ij}(g)` is a
rational integer. The genuine §3/§5 Dade-Galois fact (`σ` intertwines the cyclotomic Galois
action, so a value at `g` of order prime to the character order `|ξ_{ij}| ∣ pq` is fixed by
every ring automorphism of `ℂ`, hence rational; a rational algebraic integer is an integer).
Supplied from the honest spine grid (`Section16CharacterData.omegaS`/`tau3W`, whose value is
`σ(ω(ξ_{ij}))(g)`) through `S05.exists_intCast_sigma_omega_apply`. The c-side (`S16`) cites
this after `MHypothesis.G0_orderOf_coprime` to fill `EtaGenericData.eta_int`. -/
eta_intCast_of_coprime : ∀ (g : G), Nat.Coprime (orderOf g) (p * q) →
∀ (i : Fin q) (j : Fin p), ∃ m : ℤ, eta i j g = (m : ℂ)
/-- **Peterfalvi (3.9.a), the conjugate-pair symmetry of the `η`-grid on generic elements**
(issue-3002 keystone): for `g` of order prime to `pq`, the grid pairs under the index
negation `(i,j) ↦ (−i,−j)` (`finNeg`), `η_{−i,−j}(g) = η_{ij}(g)`. Peterfalvi's (3.9.a):
the complex conjugate of `η_{ij}` is the grid character at the conjugate index, and the value at
`g` (a rational integer by (3.9.c), hence real) equals its conjugate. Stated with
`S15.finNeg` (defeq to `S16.finNeg`) so the c-side (`S16`) cites it to fill
`EtaGenericData.eta_pair`. -/
eta_pair_of_coprime : ∀ (g : G), Nat.Coprime (orderOf g) (p * q) →
∀ (i : Fin q) (j : Fin p),
eta (finNeg q_prime.pos i) (finNeg p_prime.pos j) g = eta i j g
/-- **Peterfalvi (3.9), the principal grid value on generic elements** (issue-3002 keystone):
for `g` of order prime to `pq` (in fact for every `g`), the principal entry is `η₀₀(g) = 1`.
Since `ω₀₀ = 1_W` is the trivial character and `τ₃(1_W) = 1_G` (`tau3_trivial`), the value
is `1`. The c-side (`S16`) cites this to fill `EtaGenericData.eta_principal`. -/
eta_principal_of_coprime : ∀ (g : G), Nat.Coprime (orderOf g) (p * q) →
eta ⟨0, q_prime.pos⟩ ⟨0, p_prime.pos⟩ g = 1
namespace Hypothesis
/-- Peterfalvi's `H = P C` in (13.5)--(13.10). -/
def H (hyp : Hypothesis (G := G)) : Subgroup G :=
hyp.P ⊔ hyp.C
/-- Peterfalvi's `K = Q D`, used symmetrically for `T`. -/
def K (hyp : Hypothesis (G := G)) : Subgroup G :=
hyp.Q ⊔ hyp.D
/-- `H = PC ≤ S`: `P = S_F ≤ S` and `C = U ∩ C_G(P) ≤ U ≤ S' ≤ S`. -/
theorem H_le_S [Finite G] (hyp : Hypothesis (G := G)) : hyp.H ≤ hyp.S := by
have hUS : hyp.U ≤ hyp.S := by
have h1 : hyp.U ≤ derivedInG hyp.S := by rw [hyp.S_deriv_eq_PU]; exact le_sup_right
exact le_trans h1 (Subgroup.map_subtype_le _)
change hyp.P ⊔ hyp.C ≤ hyp.S
refine sup_le ?_ ?_
· rw [hyp.P_eq_SF]; exact OddOrder.BG.Ch4.S15.maxNilpotentNormalHall_le hyp.S
· rw [hyp.C_eq]; exact le_trans inf_le_left hUS
/-- The generic set `G_0 = G# - ((H#)^G union (Q#)^G)` from (13.9). -/
def G0 (hyp : Hypothesis (G := G)) : Set G :=
sharpSubgroup (⊤ : Subgroup G) \
(conjClassSet (sharpSubgroup hyp.H) ∪
conjClassSet (sharpSubgroup hyp.Q))
/-- Under **Peterfalvi (13.1)**, the prime `q` is odd, hence not `2`. -/
theorem q_ne_two (hyp : Hypothesis (G := G)) : hyp.q ≠ 2 := by
intro hq2
have hodd : Odd 2 := by simpa [hq2] using hyp.q_odd
rcases hodd with ⟨k, hk⟩
omega
/-- Under **Peterfalvi (13.1)**, the prime `p` is odd, hence not `2`. -/
theorem p_ne_two (hyp : Hypothesis (G := G)) : hyp.p ≠ 2 := by
intro hp2
have hodd : Odd 2 := by simpa [hp2] using hyp.p_odd
rcases hodd with ⟨k, hk⟩
omega
/-- Under **Peterfalvi (13.1)**, `q` is at least `3`. -/
theorem three_le_q (hyp : Hypothesis (G := G)) : 3 ≤ hyp.q := by
have htwo : 2 ≤ hyp.q := hyp.q_prime.two_le
have hne : hyp.q ≠ 2 := hyp.q_ne_two
omega
/-- Under **Peterfalvi (13.1)**, `p` is at least `3`. -/
theorem three_le_p (hyp : Hypothesis (G := G)) : 3 ≤ hyp.p := by
have htwo : 2 ≤ hyp.p := hyp.p_prime.two_le
have hne : hyp.p ≠ 2 := hyp.p_ne_two
omega
end Hypothesis
/-- Carrier for the basic structural conclusions of Peterfalvi (13.2). -/
structure BasicStructureData (hyp : Hypothesis (G := G)) where
S_typeII_or_typeIII : IsTypeII hyp.S ∨ IsTypeIII hyp.S
q_lt_p_forces_typeII : hyp.q < hyp.p → IsTypeII hyp.S
U_commutative : IsMulCommutative ↥hyp.U
UW1_frobenius : OddOrder.Isaacs.Ch06.IsFrobeniusGroup
↥(hyp.U ⊔ hyp.W1) (hyp.U.subgroupOf (hyp.U ⊔ hyp.W1))
(hyp.W1.subgroupOf (hyp.U ⊔ hyp.W1))
P_elementaryAbelian : IsElementaryAbelian hyp.p ↥hyp.P
P_order : Nat.card ↥hyp.P = hyp.p ^ hyp.q
u_bound : hyp.u ≤ (hyp.p ^ hyp.q - 1) / (hyp.p - 1)
A0S_TI : Prop
A0S_TI_holds : A0S_TI
tauS_eq_induction : Prop
tauS_eq_induction_holds : tauS_eq_induction
/-- **`S` is of type `P`**: the carried `S_nonI` taxonomy alternative lies in the BG type-`P`
class. This is the common structural input for the `S`-side Dade support constructions and is
faithful to the symmetric (13.1) carrier. -/
theorem Hypothesis.S_isTypeP [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G)) :
OddOrder.BG.Ch4.S14.IsTypeP hyp.S :=
OddOrder.BG.Ch4.S16.isTypeP_of_isTypeNonI hG hyp.S_maximal hyp.S_nonI
/-- **The §9 type-II/type-III setup on `S`** (Peterfalvi (13.2.a) → (9.2)): the
`TypesIIIIIIVSetup` carrier for `S`, from the honest taxonomy alternative supplied by `S_nonI`.
`maximal`/`typeP` are the carried `S_maximal`/`Sdata`; the nontrivial core is witness-independent
and transported to `Sdata`; `type_alt` retains the actual type-II/type-III branch. Note the §9
machinery's `H` is `Sdata.H = S_F = P`, so the
§9 inertia subgroup `HC` is `PC = hyp.H` — the (13.3.a) `Ind_{PC}(linear)` shape. Opens the §9
Clifford/degree machinery ((9.7)–(9.9), the `hcPsi`-induction analysis) on `S`. -/
noncomputable def Hypothesis.toTypesIIIIIIVSetupS [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G)) :
OddOrder.Peterfalvi.S11.TypesIIIIIIVSetup hyp.S := by
have htype : IsTypeII hyp.S ∨ IsTypeIII hyp.S :=
OddOrder.Peterfalvi.S13.isTypeII_or_isTypeIII_of_isTypeNonI
hG hyp.S_maximal hyp.S_nonI
exact { maximal := hyp.S_maximal
typeP := hyp.Sdata
nontrivial := by
rcases htype with hII | hIII
· exact hII.some.common.transfer hyp.Sdata
· exact hIII.some.common.transfer hyp.Sdata
type_alt := htype.elim Or.inl (fun hIII => Or.inr (Or.inl hIII)) }
open OddOrder.Isaacs.Ch03.IsAInvariant (quotientMulAutHom) in
/-- **§9 character data on `S`** (S12-mk mirror; degree-only placeholders): `u = |Ū|` is
rfl-pinned to the `U`-action image; `tau := hyp.tauS`, `H0CprimeSupport := ∅` and
`quotientSemidirectFrobenius := True` are the documented count/degree-only placeholders (as in
`S12.Hypothesis.mkSection11CharacterData` — NOT for (9.11)-coherence consumption; the honest
support/tau construction is issue 2035 step 1b). Opens `caseB_degree_qu`, the (9.9) counts and
the (9.9.c) `hcPsi`-induction exhaustion on `S` for the (13.3) μ-column facts. -/
noncomputable def Hypothesis.mkSection11CharacterDataS [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G))
(chief : OddOrder.Peterfalvi.S11.ChiefFactorData (hyp.toTypesIIIIIIVSetupS hG)) :
OddOrder.Peterfalvi.S11.Section11CharacterData (hyp.toTypesIIIIIIVSetupS hG) chief where
u := Nat.card ↥(((quotientMulAutHom (N := chief.N) chief.N_aInvariant).comp
((hyp.toTypesIIIIIIVSetupS hG).typeP.U.subgroupOf
((hyp.toTypesIIIIIIVSetupS hG).typeP.U
⊔ (hyp.toTypesIIIIIIVSetupS hG).typeP.W1)).subtype).range)
u_eq_card_quotient := rfl
H0CprimeSupport := ∅
tau := hyp.tauS
quotientSemidirectFrobenius := True
open scoped FiniteInduce in
/-- **`Ind_S^G` as a `ℂ`-linear map** `CF(S) →ₗ[ℂ] CF(G)` (issue 2035 step 1). The
canonical class-function induction `ClassFunction.induce hyp.S`, bundled as a linear map;
linearity is `ClassFunction.induce_add` / `ClassFunction.induce_smul`. This is the concrete
Peterfalvi (13.2.e)/(7.2) Dade isometry `τ = Ind_S^G` (the `(S, H^#)` isometry coincides with
induction), used as the honest `tau` of the §9 character data on `S`. -/
noncomputable def Hypothesis.indSLinearC [Finite G] (hyp : Hypothesis (G := G)) :
ClassFunction ↥hyp.S ℂ →ₗ[ℂ] ClassFunction G ℂ where
toFun := ClassFunction.induce hyp.S
map_add' := ClassFunction.induce_add hyp.S
map_smul' c θ := ClassFunction.induce_smul hyp.S c θ
open scoped FiniteInduce in
@[simp] theorem Hypothesis.indSLinearC_apply [Finite G] (hyp : Hypothesis (G := G))
(θ : ClassFunction ↥hyp.S ℂ) :
hyp.indSLinearC θ = ClassFunction.induce hyp.S θ := rfl
open scoped FiniteInduce in
/-- **`Ind_S^G` as an `IntegralCharacterMap ↥S G`** (issue 2035 step 1): the `ℤ`-linear
`Ind_S^G`, obtained by restricting scalars of the `ℂ`-linear `indSLinearC`. This is the honest
`τ` value the §9 coherence (`coherent_H0C_commutator`) consumes for the (13.3) `S`-instance —
Peterfalvi's `τ = Ind_S^G` (13.2.e). -/
noncomputable def Hypothesis.indS [Finite G] (hyp : Hypothesis (G := G)) :
OddOrder.Peterfalvi.S07.IntegralCharacterMap ↥hyp.S G :=
(hyp.indSLinearC).restrictScalars ℤ
open scoped FiniteInduce in
@[simp] theorem Hypothesis.indS_apply [Finite G] (hyp : Hypothesis (G := G))
(θ : ClassFunction ↥hyp.S ℂ) :
hyp.indS θ = ClassFunction.induce hyp.S θ := rfl
open scoped FiniteInduce in
/-- **`Ind_T^G` as an `IntegralCharacterMap ↥T G`** — the `T`-side mirror of `Hypothesis.indS`.
Used as the (unconstrained) `tau1T` of the λ-free core until the ν-side (13.2.e)-T coherence
lands (`nuGridSupply`, gated on the a-owned FT-layer carrier threading), and as the re-grounded
target of the caseA-`T` (9.11) base coherence (`sSetIrrDegT_coherent_indT`). -/
noncomputable def Hypothesis.indT [Finite G] (hyp : Hypothesis (G := G)) :
OddOrder.Peterfalvi.S07.IntegralCharacterMap ↥hyp.T G :=
LinearMap.restrictScalars ℤ
({ toFun := ClassFunction.induce hyp.T
map_add' := ClassFunction.induce_add hyp.T
map_smul' := fun c θ => ClassFunction.induce_smul hyp.T c θ } :
ClassFunction ↥hyp.T ℂ →ₗ[ℂ] ClassFunction G ℂ)
open scoped FiniteInduce in
@[simp] theorem Hypothesis.indT_apply [Finite G] (hyp : Hypothesis (G := G))
(θ : ClassFunction ↥hyp.T ℂ) :
hyp.indT θ = ClassFunction.induce hyp.T θ := rfl
/-- **`C' = [C, C]` as a subgroup** (Peterfalvi (9.5), `S`-instance): the derived subgroup of
`C = C_U(P)`. Definitionally matches the §9 `cprimeSub` (`= derivedInG (cSub …)`) once the
`S`-instance kernel `cSub = C` identification (`toTypesIIIIIIVSetupS_cSub_eq_C`) is applied. -/
def Hypothesis.Cprime (hyp : Hypothesis (G := G)) : Subgroup G := derivedInG hyp.C
/-- **(13.2.e) `normedTI` core — no `A(S)`-point escapes** (Coq `FTtypeP_facts` (e) escaping
exclusion, PFsection13.v:224-238): on a type-`P₂` maximal subgroup, every `a ∈ A(M)` has
`C_G(a) ≤ M`.
Suppose `a ∈ A(M)` escapes. Then `a ∈ M_σ^#` ((8.13.b),
`S10.escaping_typePACore_mem_sigmaSharp`), and BG Theorem D(4)
(`exists_RData_escape_structure`) attaches the unique maximal `N ⊇ C_G(a)`, of type `F` or
`P₂`, with `a ∈ Â_σ(N) ∖ N_σ` and `N_F = N_σ`:
* **`N` of type `P₂`** (Coq `Ltype2` branch, `typePF_exclusion`): the `P₂`-escape package
(BG Cor 15.9, `centralizer_escape_final_local`, via the D(4) tail) makes `M` type `F` —
contradicting `M` type `P₂` (`not_isTypeP_and_isTypeF`).
* **`N` of type `F`** (Coq `Ltype1` branch, `FTtype1_Frobenius`): `N` is of Peterfalvi type I
(Prop 16.1, `isTypeI_iff_isTypeF`), so Peterfalvi (12.7) (`typeI_frobenius`) makes `N`
Frobenius with kernel `N_F = N_σ`; `a ∈ Â_σ(N)` centralizes some `1 ≠ z ∈ N_σ`, and
Frobenius-kernel regularity (Isaacs Thm 6.4, `centralizer_kernel_le`) pulls `a` into the
kernel `N_σ` — contradicting `a ∉ N_σ`. -/
theorem escaping_typePACore_eq_empty [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G)
(hnoV : ¬ ∃ M : Subgroup G, M ∈ maximalSubgroups G ∧ OddOrder.GroupTheory.IsTypeV M)
{M : Subgroup G} (hM : M ∈ maximalSubgroups G)
(hTP : OddOrder.BG.Ch4.S14.IsTypeP M) :
OddOrder.GroupTheory.escapingCentralizerSet M (S10.typePACore M) = ∅ := by
classical
rw [Set.eq_empty_iff_forall_notMem]
rintro a ⟨haA, haesc⟩
-- the matched `κ`-Hall / `(κ∪σ)′`-Hall pair, for the `σ`-sharp confinement (8.13.b)
obtain ⟨K₀, U₀, hKM, hUM, hKne, hK, hU⟩ :=
S10.typeP_exists_kappa_hall_pair hG hM hTP
have haσ : a ∈ OddOrder.BG.Ch4.S14.sigmaSharp M :=
S10.escaping_typePACore_mem_sigmaSharp hG hM hKM hUM hKne hK hU ⟨haA, haesc⟩
-- BG Theorem D(4): the unique neighbour `N ⊇ C_G(a)` with the escape structure
obtain ⟨R, -, N, ⟨hNmem, -, hMFN, hxAN, hNtype, -, hP2imp⟩, -⟩ :=
OddOrder.BG.Ch4.S16.exists_RData_escape_structure hG hM haσ haesc
rcases hNtype with hNF | hNP2
· -- `N` type `F`: Frobenius-kernel regularity forces `a ∈ N_σ`, contradiction.
have hNmax : N ∈ maximalSubgroups G := hNmem.1
have hNI : OddOrder.GroupTheory.IsTypeI N :=
(OddOrder.Peterfalvi.S10Interface.isTypeI_iff_isTypeF hG hNmax).mpr hNF
obtain ⟨fdata, -⟩ := OddOrder.Peterfalvi.S14.typeI_frobenius hG hnoV hNmax hNI
-- the Frobenius kernel is `N_F = N_σ` (the D(4) `S15.MF N = Msigma N` conjunct)
have hker : fdata.typeI.typeF.H = OddOrder.BG.Ch3.S10.Msigma N := by
rw [fdata.typeI.typeF.H_eq]; exact hMFN
-- `a ∈ Â_σ(N)`: some `1 ≠ z ∈ N_σ` commutes with `a`
obtain ⟨haN, hne⟩ : a ∈ OddOrder.BG.Ch4.S16.hatMsigma N := hxAN.1.1
obtain ⟨z, hz1⟩ := Subgroup.ne_bot_iff_exists_ne_one.mp hne
obtain ⟨hzMσ, hzC⟩ := Subgroup.mem_inf.mp z.2
have hzN : (z : G) ∈ N := OddOrder.BG.Ch3.S10.Msigma_le N hzMσ
have hzG1 : (z : G) ≠ 1 := fun h => hz1 (Subtype.ext h)
-- `a` centralizes `z` inside `↥N`; kernel regularity pulls `a` into the kernel
have haker : (⟨a, haN⟩ : ↥N) ∈ fdata.typeI.typeF.H.subgroupOf N := by
refine fdata.frobenius.centralizer_kernel_le ⟨(z : G), hzN⟩
(Subgroup.mem_subgroupOf.mpr (by rw [hker]; exact hzMσ))
(fun h => hzG1 (congrArg Subtype.val h)) ?_
rw [Subgroup.mem_centralizer_singleton_iff]
exact Subtype.ext (Subgroup.mem_centralizer_singleton_iff.mp hzC).symm
exact hxAN.2 (SetLike.mem_coe.mpr
(by rw [← hker]; exact Subgroup.mem_subgroupOf.mp haker))
· -- `N` type `P₂`: the escape package makes `M` type `F`, contradicting type `P`.
obtain ⟨hFM, -⟩ := hP2imp hNP2
exact OddOrder.BG.Ch4.S14.not_isTypeP_and_isTypeF ⟨hTP, hFM⟩
/-- **(13.2.e) `S`-instance Dade hypothesis** (issue 1017 update #10, step 1): the
`Hypothesis`-level
instantiation of `S10.dadeSupportHypothesisData_typePACore` at the type-`P` maximal `S`
(via `hyp.S_maximal`/`hyp.S_isTypeP`), packaging the honest §16 support
`A(S) = ⋃_{x∈S_σ#} C_{S'}(x)#` (`S10.typePACore hyp.S`) as an `S04.Hypothesis`. This is the
concrete
`S04` Dade datum for `S` (previously only available as the standalone theorem taking `hM`/`hP2`);
its
`.fullDadeIsometryData` (given the support's `HConjInvariant`) materialises the Dade isometry
`τ = Ind_S^G` on the `ℤ`-lattice of virtual characters — the (13.2.e) foundation the §9 subcoherence
assembly (`S07.irrSubcoherent`) consumes to re-ground `coherent_H0Cprime_S` off the unsound
`sibleyTarget_H0C`. (Sorry-provenance parity with `S10.dadeSupportHypothesisData_typePACore`: the
inherited shared BG §16 Theorem-II pins, at exact parity with the accepted on-path
`dadeSupportHypotheses_typeI`; no lane-`b` sorry introduced.) -/
noncomputable def Hypothesis.dadeHypS [Fintype G] [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G)) :
OddOrder.Peterfalvi.S04.Hypothesis G (S10.typePACore hyp.S) hyp.S :=
(S10.dadeSupportHypothesisData_typePACore hG hyp.S_maximal (hyp.S_isTypeP hG)).some.dade
open scoped FiniteInduce in
/-- **(13.2.e) `S`-instance `dade = Ind` bridge** (issue 1017): for the honest type-`P₂` maximal
`S`,
on a class function `f` supported in a trivial-`H` sub-support `A₁ ⊆ A(S)` (an `S`-invariant subset
on which the `S`-instance Dade stabilizers vanish), the honest (13.2.e) Dade isometry
`τ = dadeIntegralCharacterMap (dadeHypS hG) …` acts as plain induction `Ind_S^G`. This is the exact
`S`-instance analogue of the type-I bridge `S14.typeI_tau_eq_induce_of_supported_trivial_H`,
instantiating the general step-3 bridge `S14.dadeMap_eq_induce_of_supported_on_trivial_H` at the
`S`-instance Dade map `dadeHypS hG` (via `dadeIntegralCharacterMap_apply_of_support`, which rewrites
the lifted integral map as the §4 `dadeMap` on the supported carrier; `inclusion` widens the support
from `A₁` to `A(S)`). The trivial-`H` sub-support facts (`hA₁A`, `hA₁norm`, `hH₁`, `hf`) are taken
as hypotheses, exactly as the type-I version defers them to the caller. This is the `dade = Ind`
identity the (13.2.e)/(7.2) coherence route (`τ = Ind_S^G`, Peterfalvi (13.2.e)) rests on, matching
the honest `indS` value. -/
theorem Hypothesis.sInstance_dade_eq_induce_of_supported_trivial_H [Fintype G] [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G))
{A₁ : Set G} (hA₁A : A₁ ⊆ S10.typePACore hyp.S)
(hA₁norm : ∀ (l : ↥hyp.S) ⦃a : G⦄, a ∈ A₁ → (l : G) * a * (l : G)⁻¹ ∈ A₁)
(hH₁ : ∀ a, ((hyp.dadeHypS hG).restrict hA₁A hA₁norm).H a = ⊥)
{f : ClassFunction ↥hyp.S ℂ}
(hf : f.support ⊆ OddOrder.Peterfalvi.S04.supportInSubgroup A₁ hyp.S) :
OddOrder.Peterfalvi.S07.dadeIntegralCharacterMap (hyp.dadeHypS hG)
((hyp.dadeHypS hG).fullDadeIsometryData) f
= ClassFunction.induce hyp.S f := by
have hfA : f.support ⊆
OddOrder.Peterfalvi.S04.supportInSubgroup (S10.typePACore hyp.S) hyp.S :=
hf.trans (OddOrder.Peterfalvi.S04.supportInSubgroup_mono hA₁A)
have h1 : OddOrder.Peterfalvi.S07.dadeIntegralCharacterMap (hyp.dadeHypS hG)
((hyp.dadeHypS hG).fullDadeIsometryData) f
= (hyp.dadeHypS hG).dadeMap (k := ℂ)
⟨f, (ClassFunction.mem_supportedSubmodule).mpr hfA⟩ :=
OddOrder.Peterfalvi.S07.dadeIntegralCharacterMap_apply_of_support (hyp.dadeHypS hG)
((hyp.dadeHypS hG).fullDadeIsometryData) hfA
rw [h1]
-- `⟨f, hfA⟩` is defeq to `inclusion hA₁A ⟨f, hf⟩` (same carrier `f`), so step 3 applies directly.
exact OddOrder.Peterfalvi.S14.dadeMap_eq_induce_of_supported_on_trivial_H (hyp.dadeHypS hG)
hA₁A hA₁norm hH₁ ⟨f, (ClassFunction.mem_supportedSubmodule).mpr hf⟩
/-- **(Rung B, reduction step 1) `dadeHypS.H a = ftSupportKernel S (A(S)) a`.** The `S`-instance
Dade
stabilizer at a support point `a` is the faithful per-`x` (8.14) signalizer kernel
`R(a) = ftSupportKernel S (A(S)) a`, read off the `H_eq_ftSupportKernel` field of the underlying
`DadeSupportHypothesisData` (the very `.some` witness `dadeHypS` is projected from). This is the
concrete formula `H(a) = R(a)` of Peterfalvi (8.15)/(8.14), specialised to the honest type-`P₂`
support `A(S)`. -/
theorem Hypothesis.dadeHypS_H_eq_ftSupportKernel [Fintype G] [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G))
(a : {a : G // a ∈ S10.typePACore hyp.S}) :
(hyp.dadeHypS hG).H a =
OddOrder.Peterfalvi.S10.ftSupportKernel hyp.S (S10.typePACore hyp.S) a.1 :=
(OddOrder.Peterfalvi.S10.dadeSupportHypothesisData_typePACore hG hyp.S_maximal
(hyp.S_isTypeP hG)).some.H_eq_ftSupportKernel a
/-- **(Rung B, reduction) No escaping `A(S)`-points ⟹ all `S`-instance Dade stabilizers vanish.**
For the honest type-`P₂` support, `dadeHypS.H a = ftSupportKernel S (A(S)) a`
(`dadeHypS_H_eq_ftSupportKernel`), and the faithful kernel is `⊥` off the escaping set
(`ftSupportKernel_eq_bot_of_not_escaping`). So if `A(S)` has no escaping point
(`∀ a ∈ A(S), a ∉ escapingCentralizerSet S (A(S))`, the Coq `FTtypeP_facts` (e) `normedTI` core,
Rung C), then `∀ a, dadeHypS.H a = ⊥` — the exact input `S04.isDadeMap_induce_of_forall_H_eq_bot` /
`sInstance_dade_eq_induce_of_supported_trivial_H` need to run `τ = Ind_S^G` on the *full* `A(S)`. -/
theorem Hypothesis.forall_dadeHypS_H_eq_bot_of_not_escaping [Fintype G] [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G))
(hno : ∀ a ∈ S10.typePACore hyp.S,
a ∉ OddOrder.GroupTheory.escapingCentralizerSet hyp.S (S10.typePACore hyp.S)) :
∀ a : {a : G // a ∈ S10.typePACore hyp.S}, (hyp.dadeHypS hG).H a = ⊥ := by
intro a
rw [hyp.dadeHypS_H_eq_ftSupportKernel hG a]
exact OddOrder.Peterfalvi.S10.ftSupportKernel_eq_bot_of_not_escaping (hno a.1 a.2)
/-- **(Rung C at `Hypothesis` level): no `A(S)`-point escapes `S`.** The general type-`P`
escaping exclusion `escaping_typePACore_eq_empty` instantiated at the carrier's `S`
(`hyp.S_maximal`, `hyp.S_isTypeP`) — the single input Rung B was reduced to. -/
theorem Hypothesis.no_escaping_typePACore [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G)
(hnoV : ¬ ∃ M : Subgroup G, M ∈ maximalSubgroups G ∧ OddOrder.GroupTheory.IsTypeV M)
(hyp : Hypothesis (G := G)) :
∀ a ∈ S10.typePACore hyp.S,
a ∉ OddOrder.GroupTheory.escapingCentralizerSet hyp.S (S10.typePACore hyp.S) := by
intro a _ ha
rw [escaping_typePACore_eq_empty hG hnoV hyp.S_maximal (hyp.S_isTypeP hG)] at ha
exact Set.notMem_empty a ha
/-- **(13.2.e) for the `S`-instance, stabilizer form: every `S`-instance Dade stabilizer is
trivial.** Rung B + Rung C: no `A(S)`-point escapes (`no_escaping_typePACore`), so every
`dadeHypS` stabilizer `H a = R(a)` vanishes (`forall_dadeHypS_H_eq_bot_of_not_escaping`). -/
theorem Hypothesis.forall_dadeHypS_H_eq_bot [Fintype G] [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G)
(hnoV : ¬ ∃ M : Subgroup G, M ∈ maximalSubgroups G ∧ OddOrder.GroupTheory.IsTypeV M)
(hyp : Hypothesis (G := G)) :
∀ a : {a : G // a ∈ S10.typePACore hyp.S}, (hyp.dadeHypS hG).H a = ⊥ :=
hyp.forall_dadeHypS_H_eq_bot_of_not_escaping hG (hyp.no_escaping_typePACore hG hnoV)
open scoped FiniteInduce in
/-- **(13.2.e) `normedTI`, isometry half — `τ = Ind_S^G` on all of `A(S)`** (Coq
`FTtypeP_facts` (e), `{in 'CF(S, 'A0(S)), tau =1 'Ind}`): the `S`-instance Dade isometry agrees
with plain induction on every `A(S)`-supported class function. This is
`sInstance_dade_eq_induce_of_supported_trivial_H` at the full support `A₁ = A(S)`, whose
trivial-stabilizer input is discharged by Rung C (`forall_dadeHypS_H_eq_bot`) — the honest
`dade = Ind` identity the (13.3) `tau1S_apply_induce_sub` route consumes. -/
theorem Hypothesis.sInstance_dade_eq_induce [Fintype G] [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G)
(hnoV : ¬ ∃ M : Subgroup G, M ∈ maximalSubgroups G ∧ OddOrder.GroupTheory.IsTypeV M)
(hyp : Hypothesis (G := G))
{f : ClassFunction ↥hyp.S ℂ}
(hf : f.support ⊆
OddOrder.Peterfalvi.S04.supportInSubgroup (S10.typePACore hyp.S) hyp.S) :
OddOrder.Peterfalvi.S07.dadeIntegralCharacterMap (hyp.dadeHypS hG)
((hyp.dadeHypS hG).fullDadeIsometryData) f
= ClassFunction.induce hyp.S f :=
hyp.sInstance_dade_eq_induce_of_supported_trivial_H hG (subset_refl _)
(fun l _ ha => S10.typePACore_conj_mem l.2 ha)
(fun a => hyp.forall_dadeHypS_H_eq_bot hG hnoV ⟨a.1, a.2⟩) hf
/-- **The honest `(H₀ ⊔ C')^#`-support for the `S`-instance, `= (C')^#`** (issue 2035 step 2).
For the `S`-instance the chief kernel is trivial (`toTypesIIIIIIVSetupS_chief_N_eq_bot`, giving
`H₀ = ⊥`), so the §9 `H₀C'`-support degenerates to `(C')^#` — the non-identity elements of
`C' = [C, C]`, viewed inside `↥S` via `C' ≤ C ≤ U ≤ S`. This is the genuine `H0CprimeSupport`
of the honest §9 character data on `S`, replacing the `∅`-placeholder of
`mkSection11CharacterDataS`. -/
def Hypothesis.cprimeSharpS (hyp : Hypothesis (G := G)) : Set ↥hyp.S :=
OddOrder.Peterfalvi.S04.sharp ((hyp.Cprime).subgroupOf hyp.S : Set ↥hyp.S)
@[simp] theorem Hypothesis.mem_cprimeSharpS (hyp : Hypothesis (G := G)) {x : ↥hyp.S} :
x ∈ hyp.cprimeSharpS ↔ (x : G) ∈ hyp.Cprime ∧ x ≠ 1 := by
simp only [Hypothesis.cprimeSharpS, OddOrder.Peterfalvi.S04.mem_sharp, SetLike.mem_coe,
Subgroup.mem_subgroupOf]
open OddOrder.Peterfalvi.S11 in
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **Peterfalvi (4.7), `S`-instance induced support**: `Supp Ind_{HU}^S ξ ⊆ A(S) ∪ {1}` for every
`ξ ∈ 𝒳` (Peterfalvi (9.5)'s family, `H ⊄ Ker ξ`), where `A(S) = ⋃_{x∈S_σ#} C_{S'}(x)#` is the honest
type-`P₂` Dade support (`S10.typePACore hyp.S`).
This is the honest §9 instance of the (4.7) support fact underlying the (5.3.a) R-datum's
`hdiffsupp`
input (the family differences of `𝒮` must vanish off `A(S)`). The proof **mirrors** the S06 (4.7)
induced-support lemma `S06.induce_apply_eq_zero_of_not_mem_union_of_not_subset_characterKernel`
(the S06 statement is bound to the `Hypothesis46` structure and cannot be instantiated for the §9
`TypesIIIIIIVSetup`, so we replay its three steps directly in the `↥S` ambient):
* `support_induce_subset_conjugatesIntoSet`: a nonvanishing point `x` of `Ind_{HU}^S ξ` is
`S`-conjugate to a point `w ∈ HU` of `Supp ξ`;
* the (4.7) **core** (Peterfalvi (1.2),
`irreducibleCharacter_apply_eq_zero_of_centralizerInSubgroup_eq_bot`
contrapositive) forces a nontrivial `d ∈ H = P = S_F ≤ S_σ` centralizing `w`, and the covering
condition puts `w ∈ A(S)` (via `S10.mem_typePACore`: `w ∈ S' = M'`, `w ≠ 1`, `w` centralizes
`d ∈ S_σ#`);
* `S10.typePACore_conj_mem` ((4.7)'s `L_normalizes_A` replacement): `A(S)` is `S`-conjugation
invariant, so the image of `x` lies in `A(S) ∪ {1}` too.
The inclusion `P = S_F ≤ S_σ` is the general `maxNilpotentNormalHall_le_Msigma` theorem, so this
support argument is uniform across the type-II/type-III alternative. -/
theorem Hypothesis.sSet_member_support_subset_A [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G))
{ξ : OddOrder.RepresentationTheory.IrreducibleCharacter ↥(huSub (hyp.toTypesIIIIIIVSetupS hG))}
(hξ : ξ ∈ xiSet (hyp.toTypesIIIIIIVSetupS hG)) :
(induceHU (hyp.toTypesIIIIIIVSetupS hG)
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ)).support ⊆
OddOrder.Peterfalvi.S04.supportInSubgroup (S10.typePACore hyp.S) hyp.S ∪ {1} := by
classical
-- `H = P = S_F ≤ S_σ`.
have hHP : (hyp.toTypesIIIIIIVSetupS hG).H = hyp.P := by
change hyp.Sdata.H = hyp.P
rw [hyp.Sdata.H_eq, hyp.P_eq_SF]
have hPle : hyp.P ≤ OddOrder.BG.Ch3.S10.Msigma hyp.S := by
rw [hyp.P_eq_SF]
exact OddOrder.BG.Ch4.S15.maxNilpotentNormalHall_le_Msigma hG hyp.S_maximal
-- the (4.7) core: a nonvanishing `w ∈ HU` with `(w:S) ≠ 1` maps into `A(S)`.
have hcore : ∀ w : ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)),
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ) w ≠ 0 →
((w : ↥hyp.S) : G) ≠ 1 →
((w : ↥hyp.S) : G) ∈ S10.typePACore hyp.S := by
intro w hwval hwne
-- (1.2) contrapositive: `C_{hInHu}(w) ≠ ⊥`.
have := hInHu_normal (hyp.toTypesIIIIIIVSetupS hG)
have hCne : OddOrder.Peterfalvi.S03.centralizerInSubgroup
(hInHu (hyp.toTypesIIIIIIVSetupS hG)) w ≠ ⊥ := fun hbot =>
hwval
(OddOrder.Peterfalvi.S03.irreducibleCharacter_apply_eq_zero_of_centralizerInSubgroup_eq_bot
ξ hξ hbot)
-- extract a nontrivial `d ∈ hInHu` centralizing `w`.
obtain ⟨d, hd_mem, hd_ne⟩ := (Subgroup.bot_or_exists_ne_one _).resolve_left hCne
rw [OddOrder.Peterfalvi.S03.mem_centralizerInSubgroup] at hd_mem
obtain ⟨hd_H, hd_comm⟩ := hd_mem
-- `d`'s ambient image lies in `H = P` and is nontrivial; it commutes with `w`.
have hdS_H : (d : ↥hyp.S) ∈ (hyp.toTypesIIIIIIVSetupS hG).H.subgroupOf hyp.S :=
(Subgroup.mem_subgroupOf).mp hd_H
have hdH_G : ((d : ↥hyp.S) : G) ∈ (hyp.toTypesIIIIIIVSetupS hG).H :=
(Subgroup.mem_subgroupOf).mp hdS_H
have hdG_ne : ((d : ↥hyp.S) : G) ≠ 1 := fun he => hd_ne (by
apply Subtype.ext; apply Subtype.ext; exact he)
have hcommG : ((d : ↥hyp.S) : G) * ((w : ↥hyp.S) : G)
= ((w : ↥hyp.S) : G) * ((d : ↥hyp.S) : G) := by
have := congrArg (fun t : ↥hyp.S => (t : G)) (Subtype.ext_iff.mp hd_comm)
simpa using this
-- covering condition: `(w:S) ∈ A(S)` via `S10.mem_typePACore` with witness `d ∈ S_σ#`.
rw [S10.mem_typePACore]
refine ⟨?_, hwne, ((d : ↥hyp.S) : G), ?_, ?_⟩
· -- `w ∈ S' = M'` since `w ∈ HU = (derivedInG S).subgroupOf S`.
have hwHU : (w : ↥hyp.S) ∈ (derivedInG hyp.S).subgroupOf hyp.S := by
rw [← huSub_eq_derivedInG_subgroupOf]; exact w.2
exact (Subgroup.mem_subgroupOf).mp hwHU
· -- `d ∈ S_σ# = (Msigma S)#`.
refine (Set.mem_sdiff _).mpr ⟨?_, fun he => hdG_ne (Set.mem_singleton_iff.mp he)⟩
have hdP : ((d : ↥hyp.S) : G) ∈ hyp.P := hHP ▸ hdH_G
exact SetLike.mem_coe.mpr (hPle hdP)
· -- `w` centralizes `d`.
rw [Subgroup.mem_centralizer_singleton_iff]
exact hcommG.symm
-- assemble via `support_induce_subset_conjugatesIntoSet` + conjugation invariance.
intro x hx
rw [Set.mem_union, Set.mem_singleton_iff]
by_cases hx1 : x = 1
· exact Or.inr hx1
-- `x` is `S`-conjugate into `Supp ξ`.
have hxsupp : (induceHU (hyp.toTypesIIIIIIVSetupS hG)
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ)) x ≠ 0 :=
ClassFunction.mem_support.mp hx
have hx_conj : x ∈ ClassFunction.conjugatesIntoSet (huSub (hyp.toTypesIIIIIIVSetupS hG))
((ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ)).support := by
have hind : induceHU (hyp.toTypesIIIIIIVSetupS hG)
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ)
= ClassFunction.induce (huSub (hyp.toTypesIIIIIIVSetupS hG))
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ) := rfl
refine ClassFunction.support_induce_subset_conjugatesIntoSet (subset_refl _) ?_
rw [← hind]; exact hxsupp
rw [ClassFunction.mem_conjugatesIntoSet] at hx_conj
obtain ⟨c, hc, hcsupp⟩ := hx_conj
set w : ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) := ⟨c⁻¹ * x * c, hc⟩ with hw_def
have hw_val : (ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ) w ≠ 0 :=
ClassFunction.mem_support.mp hcsupp
-- `(w:S) = c⁻¹ x c`; `x = c (w:S) c⁻¹` with `c ∈ S`.
have hwS_eq : ((w : ↥hyp.S) : G) = (c : G)⁻¹ * (x : G) * (c : G) := rfl
have hxeq : (x : G) = (c : G) * ((w : ↥hyp.S) : G) * (c : G)⁻¹ := by
rw [hwS_eq]; group
have hwne : ((w : ↥hyp.S) : G) ≠ 1 := by
intro he
apply hx1
have hxG : (x : G) = 1 := by rw [hxeq, he]; group
exact Subtype.ext hxG
have hwA : ((w : ↥hyp.S) : G) ∈ S10.typePACore hyp.S := hcore w hw_val hwne
-- conjugate back: `A(S)` is `S`-conjugation invariant.
refine Or.inl ?_
rw [OddOrder.Peterfalvi.S04.mem_supportInSubgroup, hxeq]
exact S10.typePACore_conj_mem c.2 hwA
open OddOrder.Peterfalvi.S11 in
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **(5.3.a) per-member difference support** (issue 1017): the Dade-image support hypothesis
`((Ind ξ)̄ − Ind ξ).support ⊆ A(S)` (as an `S`-support) that
`S07.dadeCharacterDifferenceImageOfDiff`
consumes to build the `CharacterDifferenceImage` of an irreducible §9 member. From
`sSet_member_support_subset_A` (support `⊆ A(S) ∪ {1}`), removing the identity: the difference
`(Ind ξ)̄ − Ind ξ` vanishes at `1` (the degree `Ind ξ(1) = q·ξ(1)` is a positive real,
self-conjugate).
This is the `hdiffsupp` half of the R-datum — the S-instance `R1_diffsupp` (Peterfalvi §12/§5). -/
theorem Hypothesis.sSet_member_diffsupp [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hyp : Hypothesis (G := G))
{ξ : OddOrder.RepresentationTheory.IrreducibleCharacter ↥(huSub (hyp.toTypesIIIIIIVSetupS hG))}
(hξ : ξ ∈ xiSet (hyp.toTypesIIIIIIVSetupS hG)) :
((induceHU (hyp.toTypesIIIIIIVSetupS hG)
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ)).conj
- induceHU (hyp.toTypesIIIIIIVSetupS hG)
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ)).support
⊆ OddOrder.Peterfalvi.S04.supportInSubgroup (S10.typePACore hyp.S) hyp.S := by
set φ : ClassFunction ↥hyp.S ℂ :=
induceHU (hyp.toTypesIIIIIIVSetupS hG)
(ξ : ClassFunction ↥(huSub (hyp.toTypesIIIIIIVSetupS hG)) ℂ) with hφ
have hsupp_eq : φ.conj.support = φ.support := by
ext y
simp only [ClassFunction.mem_support, ne_eq, ClassFunction.conj_apply, star_eq_zero]
intro x hx
have hx0 : (φ.conj - φ) x ≠ 0 := hx
have hxsupp : x ∈ φ.support := by
have hxU := ClassFunction.support_sub_subset _ _ hx
rwa [hsupp_eq, Set.union_self] at hxU
rcases hyp.sSet_member_support_subset_A hG hξ (hφ ▸ hxsupp) with h | h
· exact h
· exfalso
rw [Set.mem_singleton_iff] at h
subst h
obtain ⟨d, _, hd⟩ :=
OddOrder.RepresentationTheory.irreducibleCharacter_apply_one_eq_pos_natCast ξ
apply hx0
rw [ClassFunction.sub_apply, ClassFunction.conj_apply, hφ, induceHU_apply_one_eq_q_mul, hd,
star_mul', star_natCast, star_natCast, sub_self]
end OddOrder.Peterfalvi.S15
Source
Yawara Ishida, Lean 4 formalization of the Feit–Thompson (odd order) theorem, github.com/yawara/odd-order @ 20a6fccbb63c (Apache-2.0), https://github.com/yawara/odd-order/blob/20a6fccbb63c0b138acda9f7fafd798c42f9441b/OddOrder/Peterfalvi/S15_SAndT_Setup/SubcoherenceInputs.lean