Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Pyramidal groups

Definition
BFFPyramidal_pyramidal

by burkh4rt · Sep 13, 2026 · Mathlib 0df444a (Lean v4.33.1)

combinatoricscoset-partitionsgroup-theoryherzog-schonheim

A finite group GGG is pyramidal if it admits a chain of subgroups

{1}=Gn⊆Gn−1⊆⋯⊆G1⊆G0=G\{1\} = G_n \subseteq G_{n-1} \subseteq \cdots \subseteq G_1 \subseteq G_0 = G{1}=Gn​⊆Gn−1​⊆⋯⊆G1​⊆G0​=G

in which each step has index equal to the least prime factor of the order of the preceding term:

[Gk−1:Gk]=p ⁣(∣Gk−1∣),1≤k≤n,[G_{k-1} : G_k] = p\!\left(|G_{k-1}|\right), \qquad 1 \le k \le n,[Gk−1​:Gk​]=p(∣Gk−1​∣),1≤k≤n,

where p(m)p(m)p(m) denotes the least prime factor of mmm.

It is classical that a subgroup whose index is the smallest prime dividing the group order is normal; such a chain is therefore a composition series, so every pyramidal group is solvable, and every supersolvable group is pyramidal. Pyramidality is thus a chain condition lying between supersolvability and solvability. It is the hypothesis under which Berger, Felzenbaum and Fraenkel establish their multiplicity bound for coset partitions.

Formalization note. The chain is a function c : ℕ → Subgroup G together with a length nnn, with c 0 = ⊤, c n = ⊥, and the step condition imposed exactly for k < n; values of c beyond nnn are unconstrained. Each step asserts both c (k+1) ≤ c k and that Subgroup.relIndex (c (k+1)) (c k) — the index of c (k+1) ⊓ c k in c k — equals Nat.minFac (Nat.card (c k)); the containment is what makes the relative index the ordinary index. Descent is not required to be strict: once the chain reaches ⊥\bot⊥, Nat.minFac 1 = 1 forces index 111, so ⊥\bot⊥ may repeat and nnn is not determined by GGG. None of this changes the class of groups described. Normality of the steps is a consequence, not a hypothesis, and solvability is stated separately as BFFPyramidal.isSolvable_of_isPyramidal.

Definition code
import Mathlib

namespace BFFPyramidal

/-- A finite group `G` is **pyramidal** if it admits a chain of subgroups
`⊤ = c 0 ≥ c 1 ≥ ⋯ ≥ c n = ⊥` in which every step has relative index equal to the *least*
prime factor of the order of the preceding term:
`[c k : c (k+1)] = Nat.minFac |c k|` for every `k < n`.

This is conditions (1)–(2) of Berger–Felzenbaum–Fraenkel.

Three points about the formalization, none of which changes the class of groups described.
The chain is a total function on `ℕ`; only its values at indices `≤ n` are constrained, and the
step condition is imposed exactly for `k < n`. The containment `c (k+1) ≤ c k` is stated
alongside the index condition, and is what makes `Subgroup.relIndex (c (k+1)) (c k)` — the index
of `c (k+1) ⊓ c k` in `c k` — equal to the ordinary index `[c k : c (k+1)]`. Descent is not
required to be strict: once a chain reaches `⊥`, `Nat.minFac 1 = 1` forces index `1`, so `⊥` may
repeat, and `n` is therefore not determined by `G`.

It is classical that a subgroup whose index is the least prime dividing the order is normal, so
such a chain is a composition series and a pyramidal group is solvable; that consequence is *not*
part of this definition and is stated separately as `BFFPyramidal.isSolvable_of_isPyramidal`.
Every supersolvable group is pyramidal. -/
def IsPyramidal (G : Type*) [Group G] [Finite G] : Prop :=
  ∃ (n : ℕ) (c : ℕ → Subgroup G),
    c 0 = ⊤ ∧ c n = ⊥ ∧
    ∀ k < n, c (k + 1) ≤ c k ∧
      Subgroup.relIndex (c (k + 1)) (c k) = Nat.minFac (Nat.card (c k))

end BFFPyramidal
Source
Berger--Felzenbaum--Fraenkel 1987, Remark on the multiplicity of a partition of a group into cosets, Fund. Math. 128 (1987) 139-144, https://doi.org/10.4064/fm-128-3-139-144, p. 139, eqs. (1)-(2)
Read-back

What the Lean code literally says, in plain math · claude-opus-5

Read-back: BFFPyramidal.IsPyramidal

The artifact introduces a definition, not a theorem: it defines a predicate (a Prop-valued property) named IsPyramidal\mathrm{IsPyramidal}IsPyramidal on types. Its arguments are: an explicit type variable GGG (in an arbitrary universe), and two instance-implicit typeclass arguments — a group structure on GGG, and a proof that GGG is finite (Finite G, i.e. GGG is in bijection with some Fin m; note this is the propositional finiteness class, not a chosen enumeration). Nothing else is assumed; in particular GGG is not assumed nontrivial, and no commutativity or solvability assumption appears.

The predicate asserts the existence of the following data and properties:

∃ n∈N, ∃ c:N→Sub(G),\exists\, n \in \mathbb{N},\ \exists\, c : \mathbb{N} \to \mathrm{Sub}(G),∃n∈N, ∃c:N→Sub(G),

where Sub(G)\mathrm{Sub}(G)Sub(G) is the lattice of all subgroups of GGG, such that

  1. c0=⊤c_0 = \topc0​=⊤, i.e. c0c_0c0​ is the whole group GGG;
  2. cn=⊥c_n = \botcn​=⊥, i.e. cnc_ncn​ is the trivial subgroup {1}\{1\}{1};
  3. for every natural number kkk with k<nk < nk<n (strict inequality; so this is the set of indices k∈{0,1,…,n−1}k \in \{0, 1, \dots, n-1\}k∈{0,1,…,n−1}, empty when n=0n = 0n=0), both of the following hold:
    • ck+1≤ckc_{k+1} \le c_kck+1​≤ck​ as subgroups, i.e. ck+1c_{k+1}ck+1​ is contained in ckc_kck​ (non-strict containment: equality is permitted);
[ ck:ck+1∩ck ]  =  minFac( ∣ck∣ ).\bigl[\, c_k : c_{k+1} \cap c_k \,\bigr] \;=\; \mathrm{minFac}\bigl(\,|c_k|\,\bigr).[ck​:ck+1​∩ck​]=minFac(∣ck​∣).

Here the left-hand side is Mathlib's relative index relIndex(ck+1,ck)\mathrm{relIndex}(c_{k+1}, c_k)relIndex(ck+1​,ck​), which by definition is the index, inside the group ckc_kck​, of the subgroup ck+1∩ckc_{k+1} \cap c_kck+1​∩ck​ of ckc_kck​. It is stated in that intersected form regardless of the containment hypothesis; given the containment ck+1≤ckc_{k+1} \le c_kck+1​≤ck​ imposed in the same conjunct, this intersection is just ck+1c_{k+1}ck+1​, so the quantity is the ordinary index [ck:ck+1][c_k : c_{k+1}][ck​:ck+1​]. (Indices here are Mathlib's natural-number index, which is defined to be 000 when the index is infinite; since GGG is finite this degenerate value cannot occur, and the index is a positive natural number.)

The right-hand side is minFac\mathrm{minFac}minFac, the least prime factor function on N\mathbb{N}N, applied to ∣ck∣=Nat.card(ck)|c_k| = \mathrm{Nat.card}(c_k)∣ck​∣=Nat.card(ck​), the cardinality of the underlying set of the subgroup ckc_kck​ viewed as a type. Because GGG is finite and a subgroup is nonempty, ∣ck∣≥1|c_k| \ge 1∣ck​∣≥1, so the junk value minFac(0)=2\mathrm{minFac}(0) = 2minFac(0)=2 is unreachable. The genuinely relevant boundary value is ∣ck∣=1|c_k| = 1∣ck​∣=1: Mathlib defines minFac(1)=1\mathrm{minFac}(1) = 1minFac(1)=1, so at a step where ckc_kck​ is the trivial subgroup the condition reads [ck:ck+1]=1[c_k : c_{k+1}] = 1[ck​:ck+1​]=1, which (with ck+1≤ck=⊥c_{k+1} \le c_k = \botck+1​≤ck​=⊥) is automatic. Consequently the chain is not required to descend strictly: once it reaches ⊥\bot⊥ it may stall at ⊥\bot⊥ for arbitrarily many further steps, and more generally the containments ck+1≤ckc_{k+1} \le c_kck+1​≤ck​ are non-strict wherever minFac(∣ck∣)=1\mathrm{minFac}(|c_k|) = 1minFac(∣ck​∣)=1, which happens exactly when ∣ck∣=1|c_k| = 1∣ck​∣=1. For any ∣ck∣>1|c_k| > 1∣ck​∣>1 the required index is the least prime dividing ∣ck∣|c_k|∣ck​∣, hence ≥2\ge 2≥2, forcing strict descent there.

Three further points about what the quantifiers silently include:

  • ccc is a total function on all of N\mathbb{N}N; its values at indices >n> n>n are completely unconstrained, and the conditions in (3) are imposed only for k<nk < nk<n. So the existential over ccc is really an existential over the finite tuple (c0,…,cn)(c_0, \dots, c_n)(c0​,…,cn​).
  • The case n=0n = 0n=0 is allowed. Then conditions (1) and (2) together say ⊤=c0=⊥\top = c_0 = \bot⊤=c0​=⊥, i.e. GGG is the trivial group, and condition (3) is vacuous. Hence the trivial group satisfies the predicate via n=0n = 0n=0; for a nontrivial GGG the witness nnn must be at least 111. The number nnn is not pinned down by GGG (it can always be increased by repeating ⊥\bot⊥).
  • No normality is required of any ckc_kck​, no assertion that [ck:ck+1][c_k : c_{k+1}][ck​:ck+1​] is prime as such is made beyond its being the least prime factor of ∣ck∣|c_k|∣ck​∣ (which for ∣ck∣=1|c_k| = 1∣ck​∣=1 is the non-prime value 111), and nothing about solvability, composition series, or supersolvability is asserted by this definition.

Finally, the declaration carries a long documentation comment describing the intended meaning, attributing conditions to Berger–Felzenbaum–Fraenkel, and asserting that such groups are solvable and that every supersolvable group is pyramidal. Those remarks are prose only: they are not part of the logical content of the definition and nothing in the artifact proves them.

Human review
  • Endorsed by Shuze Chen · Sep 13, 2026

  • Endorsed by burkh4rt · Sep 13, 2026

    Confirmed by the mission captain (proposal self-audit).

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