Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

The divisor-closure measure counts multiples below the index

Proved
SubnormalCover.muMeasure_divisorClosure_relIndex

by burkh4rt · Sep 14, 2026 · Mathlib 0df444a (Lean v4.33.1)

combinatoricsgroup-theory

Let GGG be a group and H≤GH \le GH≤G a subgroup of finite index, and let G1,…,GkG_1, \dots, G_kG1​,…,Gk​ be subgroups of GGG all containing HHH. Then the totient measure of the divisor closure of the relative indices equals the number of residues below [G:H][G:H][G:H] hit by one of the index-progressions:

μ(D({[Gi:H]:1≤i≤k}))  =  ∣{ 0≤n<[G:H]  :  [G:Gi]∣n for some 1≤i≤k }∣.\mu\bigl(D(\{[G_i : H] : 1 \le i \le k\})\bigr) \;=\; \bigl|\{\,0 \le n < [G:H] \;:\; [G:G_i] \mid n \text{ for some } 1 \le i \le k\,\}\bigr|.μ(D({[Gi​:H]:1≤i≤k}))=​{0≤n<[G:H]:[G:Gi​]∣n for some 1≤i≤k}​.

The proof applies Lemma 3.2 to the family Γ={Xd:d∣N}\Gamma = \{X_d : d \mid N\}Γ={Xd​:d∣N}, N=[G:H]N = [G:H]N=[G:H], Xd={0≤x<N:d∣x}X_d = \{0 \le x < N : d \mid x\}Xd​={0≤x<N:d∣x}, using ∣Xni∣=N/ni=[Gi:H]|X_{n_i}| = N / n_i = [G_i : H]∣Xni​​∣=N/ni​=[Gi​:H]; this converts the right-hand side of Sun's Theorem 3.1 into a purely group-theoretic quantity.

Formalization note. [Gi:H][G_i : H][Gi​:H] is H.relIndex (K i), which is [Ki:H⊓Ki][K_i : H \sqcap K_i][Ki​:H⊓Ki​] in general, so the hypothesis hHK : ∀ i, H ≤ K i is what makes it mean [Gi:H][G_i : H][Gi​:H] (argument order confirmed by Subgroup.relIndex_mul_index : H ≤ K → H.relIndex K * K.index = H.index). Finiteness of [G:H][G:H][G:H] is spelled H.index ≠ 0 rather than H.FiniteIndex, since Subgroup.index returns 0 for infinite index and the two are equivalent by Subgroup.finiteIndex_iff; the counting set is (Finset.range H.index).filter …, whose decidability instance is found automatically from Fin k being a fintype.

Preamble
import Mathlib
import Definitions.Def_BFFPyramidal_arithmetic
Formal statement
namespace SubnormalCover

open BFFPyramidal

theorem muMeasure_divisorClosure_relIndex {G : Type*} [Group G] (H : Subgroup G)
    (hfin : H.index ≠ 0) {k : ℕ} (K : Fin k → Subgroup G) (hHK : ∀ i, H ≤ K i) :
    muMeasure (divisorClosure (Finset.univ.image fun i => H.relIndex (K i)))
      = ((Finset.range H.index).filter fun n => ∃ i, (K i).index ∣ n).card := by
  sorry

end SubnormalCover
Source
Z.-W. Sun, On the Herzog-Schönheim conjecture for uniform covers of groups, J. Algebra 273 (2004) 153-175, https://doi.org/10.1016/S0021-8693(03)00526-X, Lemma 3.3, eq. (3.4)
Read-back

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

Setting. Let GGG be a group (any type in any universe with a group structure), let H≤GH \le GH≤G be a subgroup, let kkk be a natural number (an implicit argument, so it is inferred from KKK but universally quantified), and let K0,…,Kk−1K_0, \dots, K_{k-1}K0​,…,Kk−1​ be a family of subgroups of GGG indexed by i∈{0,…,k−1}i \in \{0,\dots,k-1\}i∈{0,…,k−1}.

Hypotheses.

  • hfinh_{\text{fin}}hfin​: [G:H]≠0[G : H] \ne 0[G:H]=0. Here [G:H][G:H][G:H] is Mathlib's Subgroup.index, defined as the cardinality #(G/H)\#(G/H)#(G/H) of the set of left cosets, with the convention that it equals 000 when that set is infinite; so this hypothesis says exactly that HHH has finite index (it is stated as a numeric disequality, not via the FiniteIndex typeclass), and in particular [G:H]≥1[G:H] \ge 1[G:H]≥1.
  • hHKh_{HK}hHK​: H≤KiH \le K_iH≤Ki​ for every index iii.

No finiteness assumption is imposed on GGG itself, and none is imposed directly on the KiK_iKi​ (whose indices [G:Ki][G:K_i][G:Ki​] would be 000 by the same convention if they were of infinite index).

Conclusion. An equality of natural numbers:

∑m∈Dφ(m)  =  #{ n∈N  :  0≤n<[G:H]  and  ∃ i, [G:Ki]∣n },\sum_{m \in D} \varphi(m) \;=\; \#\big\{\, n \in \mathbb{N} \;:\; 0 \le n < [G:H] \ \text{ and } \ \exists\, i,\ [G:K_i] \mid n \,\big\},m∈D∑​φ(m)=#{n∈N:0≤n<[G:H]  and  ∃i, [G:Ki​]∣n},

where:

  • R:={ [Ki:H]  :  i∈{0,…,k−1} }R := \{\, [K_i : H] \;:\; i \in \{0,\dots,k-1\} \,\}R:={[Ki​:H]:i∈{0,…,k−1}} is the finite set of natural numbers obtained as the image of the index family under the relative index. Mathlib's relIndex\mathrm{relIndex}relIndex of HHH in KiK_iKi​ is the index of H∩KiH \cap K_iH∩Ki​ inside KiK_iKi​, i.e. [Ki:H∩Ki][K_i : H \cap K_i][Ki​:H∩Ki​], again 000 by convention if that is infinite; under hHKh_{HK}hHK​ this is the ordinary index [Ki:H][K_i : H][Ki​:H]. Duplicate values collapse, so RRR is a set, not a multiset.
  • D:=divisorClosure(R)=⋃m∈Rdivisors(m)D := \mathrm{divisorClosure}(R) = \bigcup_{m \in R} \mathrm{divisors}(m)D:=divisorClosure(R)=⋃m∈R​divisors(m), the set of all positive divisors of members of RRR, using Mathlib's divisors\mathrm{divisors}divisors with divisors(0)=∅\mathrm{divisors}(0) = \varnothingdivisors(0)=∅ — so a member 000 of RRR (which would arise from an infinite relative index) contributes nothing to DDD.
  • The left-hand side is μ(D)=∑m∈Dφ(m)\mu(D) = \sum_{m \in D} \varphi(m)μ(D)=∑m∈D​φ(m), the sum of Euler's totient over that divisor-closed set, each element counted once.
  • The right-hand side counts the natural numbers nnn in the range 0,1,…,[G:H]−10, 1, \dots, [G:H]-10,1,…,[G:H]−1 (note: n=0n = 0n=0 is included and n=[G:H]n = [G:H]n=[G:H] is excluded) that are divisible by [G:Ki][G : K_i][G:Ki​] for at least one iii.

Degenerate cases silently allowed. If k=0k = 0k=0 the family is empty: R=∅R = \varnothingR=∅, D=∅D = \varnothingD=∅, the left side is 000; the existential ∃i\exists i∃i is false for every nnn, so the right side is 000 as well, and the statement reads 0=00 = 00=0. If k≥1k \ge 1k≥1, then n=0n = 0n=0 always satisfies the filter (every natural number, including 000, divides 000), so the right side is at least 111. The hypothesis hfinh_{\text{fin}}hfin​ rules out the range being taken over {0,…,−1}\{0,\dots,-1\}{0,…,−1}; with [G:H]=1[G:H] = 1[G:H]=1 (i.e. H=GH = GH=G) the range is {0}\{0\}{0}. Nothing here mentions normality, subnormality, cosets other than through the index, or any covering condition.

Human review
  • Endorsed by Shuze Chen · Sep 14, 2026

  • Endorsed by burkh4rt · Sep 14, 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