Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Peterfalvi (2.1): the coprime-coset structure lemma

Definition
OddOrder_Peterfalvi_S06_CertainTypeStructure

by burkh4rt · Nov 12, 2026 · Mathlib 0df444a (Lean v4.33.1)

feit-thompsonfinite-groupsgroup-theoryishida-odd-order

Peterfalvi, Character Theory for the Odd Order Theorem, §2 (The Dade Isometry), p. 10, statement (2.1): let g∈Gg\in Gg∈G and H≤GH\le GH≤G such that ggg normalizes HHH and ⟨g⟩\langle g\rangle⟨g⟩, HHH have coprime orders. Then HgHgHg is the disjoint union of ∣H:CH(g)∣|H:C_H(g)|∣H:CH​(g)∣ subsets conjugate to CH(g)gC_H(g)gCH​(g)g in H⟨g⟩H\langle g\rangleH⟨g⟩.

The bundle consists of lemmas only.

The bundle also proves 3 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/S06_CertainTypeStructure.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_S06_DadeIsometryCertain

namespace OddOrder.Peterfalvi.S06

open scoped Pointwise

variable {G : Type*} [Group G]

/-- A uniform exponent `e` with `g^e = g` and `x^e = 1` for all `x ∈ H`, available when
`o(g)` and `|H|` are coprime.  (Chinese remainder: `e ≡ 1 mod o(g)`, `e ≡ 0 mod |H|`.) -/
private theorem exists_uniform_exponent [Finite G] {H : Subgroup G} {g : G}
    (hcop : Nat.Coprime (orderOf g) (Nat.card H)) :
    ∃ e : ℕ, g ^ e = g ∧ ∀ x ∈ H, x ^ e = 1 := by
  obtain ⟨e, he1, he0⟩ := Nat.chineseRemainder hcop 1 0
  refine ⟨e, ?_, ?_⟩
  · -- `g ^ e = g ^ 1` from `e ≡ 1 [MOD o(g)]`.
    have : g ^ e = g ^ 1 := pow_eq_pow_iff_modEq.mpr he1
    rwa [pow_one] at this
  · intro x hx
    -- `|H| ∣ e` from `e ≡ 0 [MOD |H|]`, and `o(x) ∣ |H|`, so `o(x) ∣ e`.
    have hdvd : Nat.card H ∣ e := (Nat.modEq_zero_iff_dvd).mp he0
    have hox : orderOf x ∣ Nat.card H := Subgroup.orderOf_dvd_natCard _ hx
    exact orderOf_dvd_iff_pow_eq_one.mp (hox.trans hdvd)

/-- **Peterfalvi (2.1)** (the form used downstream).  Let `g` normalize `H ≤ G` (i.e.
`g · x · g⁻¹ ∈ H` for all `x ∈ H`) and assume `o(g)` is coprime to `|H|`.  Then every
element of the right coset `Hg` is `H`-conjugate to an element of `C_H(g)·g`. -/
theorem coset_conj_into_centralizer_coset [Finite G] {H : Subgroup G} {g : G}
    (hg : ∀ x ∈ H, g * x * g⁻¹ ∈ H) (hcop : Nat.Coprime (orderOf g) (Nat.card H))
    {z : G} (hz : z * g⁻¹ ∈ H) :
    ∃ c ∈ H, ∃ w ∈ Subgroup.centralizer ({g} : Set G) ⊓ H, c⁻¹ * z * c = w * g := by
  classical
  set C : Subgroup G := Subgroup.centralizer ({g} : Set G) ⊓ H with hC
  have hCH : C ≤ H := by rw [hC]; exact inf_le_right
  obtain ⟨e, hge, hHe⟩ := exists_uniform_exponent (H := H) (g := g) hcop
  -- Collapse identity: for `w ∈ C`, `(w·g)^e = g`.
  have hcollapse : ∀ w : G, w ∈ C → (w * g) ^ e = g := by
    intro w hw
    rw [hC, Subgroup.mem_inf] at hw
    obtain ⟨hwc, hwH⟩ := hw
    have hwcomm : Commute w g := (Subgroup.mem_centralizer_iff.mp hwc g rfl).symm
    rw [hwcomm.mul_pow, hHe w hwH, hge, one_mul]
  -- `e`-power of a conjugate of `w·g` (`w ∈ C`) is `a·g·a⁻¹`.
  have hconj_pow : ∀ (a w : G), w ∈ C → (a * (w * g) * a⁻¹) ^ e = a * g * a⁻¹ := by
    intro a w hw
    have hmap := map_pow (MulAut.conj a) (w * g) e
    simp only [MulAut.conj_apply] at hmap
    rw [hcollapse w hw] at hmap
    exact hmap.symm
  -- The conjugation map `Φ` from `(H ⧸ C.subgroupOf H) × C` to `G`.
  let Φ : (H ⧸ C.subgroupOf H) × C → G :=
    fun p => ((p.1.out : H) : G) * ((p.2 : G) * g) * ((p.1.out : H) : G)⁻¹
  -- `Φ` lands in the coset `R = Hg`.
  have hsubR : Set.range Φ ⊆ {x : G | x * g⁻¹ ∈ H} := by
    rintro _ ⟨⟨q, w⟩, rfl⟩
    have haH : ((q.out : H) : G) ∈ H := (q.out : H).2
    have hwH : (w : G) ∈ H := hCH w.2
    change (((q.out : H) : G) * ((w : G) * g) * ((q.out : H) : G)⁻¹) * g⁻¹ ∈ H
    have hgag : g * ((q.out : H) : G)⁻¹ * g⁻¹ ∈ H := hg _ (H.inv_mem haH)
    have hrw : (((q.out : H) : G) * ((w : G) * g) * ((q.out : H) : G)⁻¹) * g⁻¹
        = (((q.out : H) : G) * (w : G)) * (g * ((q.out : H) : G)⁻¹ * g⁻¹) := by group
    rw [hrw]
    exact H.mul_mem (H.mul_mem haH hwH) hgag
  -- `Φ` is injective.
  have hinj : Function.Injective Φ := by
    rintro ⟨q, w⟩ ⟨q', w'⟩ hEq
    have haH : ((q.out : H) : G) ∈ H := (q.out : H).2
    have ha'H : ((q'.out : H) : G) ∈ H := (q'.out : H).2
    -- `Φ` value equation, then raise to the `e`-th power.
    have hval : ((q.out : H) : G) * ((w : G) * g) * ((q.out : H) : G)⁻¹
        = ((q'.out : H) : G) * ((w' : G) * g) * ((q'.out : H) : G)⁻¹ := hEq
    have hpow : ((q.out : H) : G) * g * ((q.out : H) : G)⁻¹
        = ((q'.out : H) : G) * g * ((q'.out : H) : G)⁻¹ := by
      have := congrArg (· ^ e) hval
      simpa only [hconj_pow _ (w : G) w.2, hconj_pow _ (w' : G) w'.2] using this
    -- Hence `a'⁻¹ * a` commutes with `g`, so lies in `C`.
    set a : G := ((q.out : H) : G) with ha
    set a' : G := ((q'.out : H) : G) with ha'
    have hcomm : a⁻¹ * a' ∈ Subgroup.centralizer ({g} : Set G) := by
      rw [Subgroup.mem_centralizer_iff]
      intro m hm
      rw [Set.mem_singleton_iff] at hm
      subst hm
      calc m * (a⁻¹ * a') = a⁻¹ * (a * m * a⁻¹) * a' := by group
        _ = a⁻¹ * (a' * m * a'⁻¹) * a' := by rw [hpow]
        _ = (a⁻¹ * a') * m := by group
    have haa'C : a⁻¹ * a' ∈ C := by
      rw [hC, Subgroup.mem_inf]
      exact ⟨hcomm, H.mul_mem (H.inv_mem haH) ha'H⟩
    -- So `q = q'` (same `C.subgroupOf H`-coset).
    have hout : ∀ r : H ⧸ C.subgroupOf H, QuotientGroup.mk (r.out : H) = r :=
      fun r => Quotient.out_eq r
    have hquot : q = q' := by
      have hcoe : (((q.out : H)⁻¹ * (q'.out : H) : ↥H) : G) = a⁻¹ * a' := by
        push_cast; rw [← ha, ← ha']
      have key : ((q.out : H)⁻¹ * (q'.out : H)) ∈ C.subgroupOf H :=
        Subgroup.mem_subgroupOf.mpr (by rw [hcoe]; exact haa'C)
      calc q = QuotientGroup.mk (q.out : H) := (hout q).symm
        _ = QuotientGroup.mk (q'.out : H) := QuotientGroup.eq.mpr key
        _ = q' := hout q'
    subst hquot
    -- Then `w = w'`.
    have hww' : (w : G) = (w' : G) := by
      have h2 : a * ((w : G) * g) * a⁻¹ = a * ((w' : G) * g) * a⁻¹ := hval
      exact mul_right_cancel (mul_left_cancel (mul_right_cancel h2))
    have : w = w' := Subtype.ext hww'
    subst this; rfl
  -- Cardinalities: `|domain Φ| = |H| = |Hg|`.
  have hcardC : Nat.card (C.subgroupOf H) = Nat.card C :=
    Nat.card_congr (Subgroup.subgroupOfEquivOfLe hCH).toEquiv
  have hcardDom : Nat.card ((H ⧸ C.subgroupOf H) × C) = Nat.card H := by
    rw [Nat.card_prod, ← hcardC, ← Subgroup.card_eq_card_quotient_mul_card_subgroup (C.subgroupOf
        H)]
  -- `Hg` (here `R`) has `|H|` elements (the bijection `h ↦ h·g`).
  set R : Set G := {x : G | x * g⁻¹ ∈ H} with hR
  have hRimg : R = (fun h => h * g) '' (H : Set G) := by
    ext x
    constructor
    · intro hx; exact ⟨x * g⁻¹, hx, by group⟩
    · rintro ⟨h, hh, rfl⟩; change h * g * g⁻¹ ∈ H; simpa [mul_assoc] using hh
  have hRcard : R.ncard = Nat.card H := by
    rw [hRimg, Set.ncard_image_of_injective _ (mul_left_injective g)]
    exact (Nat.card_coe_set_eq _).symm
  have hrangeCard : (Set.range Φ).ncard = Nat.card H := by
    rw [← Set.image_univ, Set.ncard_image_of_injective _ hinj, Set.ncard_univ]
    exact hcardDom
  have hrangeEqR : Set.range Φ = R :=
    Set.eq_of_subset_of_ncard_le hsubR (by rw [hRcard, hrangeCard]) (Set.toFinite R)
  -- Conclude: `z ∈ R = range Φ`, so `z = a·(w·g)·a⁻¹` and `a⁻¹·z·a = w·g`.
  have hzR : z ∈ Set.range Φ := by rw [hrangeEqR]; exact hz
  obtain ⟨⟨q, w⟩, hΦeq⟩ := hzR
  refine ⟨((q.out : H) : G), (q.out : H).2, (w : G), w.2, ?_⟩
  have hval : ((q.out : H) : G) * ((w : G) * g) * ((q.out : H) : G)⁻¹ = z := hΦeq
  rw [← hval]; group

end OddOrder.Peterfalvi.S06

namespace OddOrder.Peterfalvi.S06.Hypothesis

variable {L : Type*} [Group L] [Finite L] (h : Hypothesis L)

/-- **Peterfalvi (2.1) applied to `L = K ⋊ W₁`.**  Every element `z ∈ L − K` is `L`-conjugate
to an element `x·y` with `x ∈ W₁^#` and `y ∈ W₂`.

`z` decomposes as `k·x` (`k ∈ K`, `x ∈ W₁`) by the complement, with `x ≠ 1` since `z ∉ K`;
then `z ∈ K·x` and (2.1) conjugates `z` into `C_K(x)·x = W₂·x = x·W₂`. -/
theorem mem_compl_conj_into_W {z : L} (hz : z ∉ h.K) :
    ∃ c : L, ∃ x ∈ h.W1, x ≠ 1 ∧ ∃ y ∈ h.W2, c⁻¹ * z * c = x * y := by
  -- Complement decomposition `z = k·x`.
  obtain ⟨⟨⟨kk, hk⟩, ⟨u, hu⟩⟩, hku, -⟩ := Subgroup.IsComplement.existsUnique h.isComplement z
  change kk * u = z at hku
  have hu1 : u ≠ 1 := by
    rintro rfl
    exact hz (by rw [← hku, mul_one]; exact hk)
  -- Apply the abstract (2.1) with `H = K`, `g = u`.
  have hnorm : ∀ x ∈ h.K, u * x * u⁻¹ ∈ h.K := fun x hx => h.K_normal.conj_mem x hx u
  have hcop : Nat.Coprime (orderOf u) (Nat.card h.K) :=
    h.card_coprime.symm.coprime_dvd_left (Subgroup.orderOf_dvd_natCard _ hu)
  have hzc : z * u⁻¹ ∈ h.K := by rw [← hku, mul_assoc, mul_inv_cancel, mul_one]; exact hk
  obtain ⟨c, _, w, hwC, hcw⟩ := coset_conj_into_centralizer_coset hnorm hcop hzc
  -- `C_K(u) = W₂` (4.2.b).
  have hwW2 : w ∈ h.W2 := by rw [← h.centralizer_W2 u hu hu1]; exact hwC
  -- `c⁻¹ z c = w·u = u·w` (`W₂` centralizes `W₁`).
  refine ⟨c, u, hu, hu1, w, hwW2, ?_⟩
  rw [hcw]
  exact (h.commute_of_mem_W1_of_mem_W2 hu hwW2).symm

end OddOrder.Peterfalvi.S06.Hypothesis
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/S06_CertainTypeStructure.lean

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me