Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Colour-critical subsets have minimum degree at least `k`.

Proved
Hadwiger.exists_subset_minDegree_of_not_colorable

by raver1975 · Sep 17, 2026 · Mathlib c5ea003 (Lean v4.30.0)

aether-catalogprobability

Colour-critical subsets have minimum degree at least k. If G is not k-colourable then some non-empty vertex subset S is itself not k-colourable and has the property that every vertex of S has at least k neighbours inside S.

theorem Hadwiger.exists_subset_minDegree_of_not_colorable[Fintype V] [DecidableEq V]
    (h : ¬ G.Colorable k) :
    ∃ S : Finset V, S.Nonempty ∧ ¬ ColorableOn G S k ∧
      ∀ v ∈ S, k ≤ (S.filter (fun w => G.Adj v w)).card := by sorry

Formalization Note Transplanted verbatim from the Aether Catalog source Probability/HadwigerCritical.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.

Preamble
-- Thm stub generated from Probability/HadwigerCritical.lean
import Mathlib
import Definitions.Def_Probability_HadwigerCritical
import Definitions.Def_Probability_HadwigerSmallCases
/-
  # Colour-critical subgraphs and the minimum-degree reduction for Hadwiger's conjecture

  This file closes the first structural step of Conjecture 1 of `FUTURE_DIRECTIONS.md`
  ("Dirac's shape"): it reduces Hadwiger's conjecture for a parameter `k` to the
  *minimum-degree* statement

      every finite graph of minimum degree at least `k` has a `K_{k+1}` minor.

  The reduction proceeds through a colour-critical subgraph:

  1. `ColorableOn` — proper `k`-colourability of a vertex subset;
  2. `exists_critical_subset` — every graph that is not `k`-colourable contains a
     *vertex-minimal* non-`k`-colourable subset;
  3. `exists_subset_minDegree_of_not_colorable` — in that minimal subset **every**
     vertex has at least `k` neighbours inside the subset (the classical
     "critical graphs have large minimum degree" lemma, proved by a greedy
     re-colouring of the deleted vertex);
  4. `hadwigerProperty_of_minDegree_forces` — hence the minimum-degree statement
     implies `HadwigerProperty k`, using `isMinor_of_isMinor_induce` from
     `HadwigerCore.lean` to lift the minor from the induced subgraph.

  As applications we obtain

  * `completeMinor_three_of_two_le_degree` — the case `k = 2` of the minimum-degree
    statement (minimum degree `2` forces a `K₃` minor), proved from the sharp
    extremal bound of `HadwigerDensity.lean` together with the handshake lemma;
  * `hadwiger_two_via_min_degree` — an independent, degeneracy-flavoured proof of
    `HadwigerProperty 2`;
  * `hadwiger_three_of_dirac` — the `k = 3` instance of the reduction: Dirac's
    theorem "minimum degree `3` forces a `K₄` minor" implies `HadwigerProperty 3`.

  Note that the minimum-degree statement is *strictly stronger* than Hadwiger's
  conjecture for large `k` (it fails for `k` large, by Kostochka's bound), so the
  reduction is genuinely one-directional; for `k ≤ 3` the stronger statement is
  the classical route.

  -- !-- Lab Notes -- !--
  * The greedy re-colouring in step 3 needs `k ≥ 1`; the degenerate case `k = 0`
    is handled separately (`S = univ` works, the degree condition being vacuous).
  * Passing from the `Finset`-level degree `(S.filter (G.Adj v)).card` to
    `Nat.card ((G.induce ↑S).neighborSet ⟨v, hv⟩)` is done through the injective
    image under `Subtype.val`; this is `card_neighborSet_induce`.
-/

open Hadwiger

open SimpleGraph Finset

variable {V : Type*} {G : SimpleGraph V} [DecidableRel G.Adj] {k : ℕ}

/-! ## 1.  Colourability of a vertex subset -/




/-! ## 2.  A vertex-minimal non-colourable subset -/


/-! ## 3.  Critical subsets have large minimum degree -/
Formal statement
theorem Hadwiger.exists_subset_minDegree_of_not_colorable[Fintype V] [DecidableEq V]
    (h : ¬ G.Colorable k) :
    ∃ S : Finset V, S.Nonempty ∧ ¬ ColorableOn G S k ∧
      ∀ v ∈ S, k ≤ (S.filter (fun w => G.Adj v w)).card := by sorry
Source
https://github.com/paulklemstine/Lean/blob/53c2925a02/Catalog/Probability/HadwigerCritical.lean#L96

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