Lax12.UniformQuasiWideness

Uniform quasi-wideness

concepts/Lax12/UniformQuasiWideness.lean · lax-12

definition

Loading review…

Sign in with ORCID

Community review

Flags

Each flag is tied to a public ORCID identity and explains why this concept may be incorrect.

No flags have been submitted.

    Community review

    Flag this concept

    State precisely what appears incorrect. This explanation will be public under your ORCID name.

    No source line selected.

    Concept map

    Proven claimDefinitionThis conceptRelated conceptA → B: B builds on A

    Definition

    A set A of vertices is distance-r independent in G if any two distinct vertices of A are at distance more than r. A graph class is uniformly quasi-wide if for every radius r there are a threshold function N and a separator bound s such that in every member G, every vertex set A of size at least N(m) contains a distance-r independent subset of size at least m of GS, for some set S of at most s vertices.

    This is Definition 3.1 of Chapter 4 of the source lecture notes (2019/20 edition), where the constants s are called the margins and the functions N the wideness functions.

    Lean source view on GitHub

    1import Lax12.GraphClasses
    2import Mathlib.Combinatorics.SimpleGraph.Walk.Basic
    3import Mathlib.Data.Set.Card
    4
    5/-!
    6---
    7title: Uniform quasi-wideness
    8type: definition
    9---
    10A set *A* of vertices is distance-*r* independent in *G* if any two
    11distinct vertices of *A* are at distance more than *r*. A graph class is
    12uniformly quasi-wide if for every radius *r* there are a threshold
    13function *N* and a separator bound *s* such that in every member *G*,
    14every vertex set *A* of size at least *N*(*m*) contains a distance-*r*
    15independent subset of size at least *m* of *G* − *S*, for some set *S*
    16of at most *s* vertices.
    17
    18This is Definition 3.1 of Chapter 4 of the source lecture notes (2019/20
    19edition), where the constants *s* are called the margins and the
    20functions *N* the wideness functions.
    21
    22# Formalization notes
    23
    24Distance is stated with walks: two vertices are at distance more than
    25`r` exactly when every walk between them is longer than `r`, which needs
    26no metric, connectivity or decidability instance. Deleting a vertex set
    27is modelled by isolating it — `deleteVerts G S` keeps the carrier and
    28drops every edge incident to `S` — so every set in the statement lives
    29in the same vertex type and no subtype plumbing enters the surface.
    30Since the witness satisfies `B ⊆ A \ S`, the isolated vertices are not
    31in `B` and distance-`r` independence in the isolated graph is the same
    32as in the induced subgraph on the complement of `S`.
    33
    34Sets and `Set.ncard` are used throughout, as in the other concepts of
    35this submission. The threshold `N` may depend on the requested size `m`,
    36while the separator bound `s` may not: that uniformity in `s` is the
    37"uniform" of uniform quasi-wideness and is the whole strength of the
    38notion. `DistIndependent` and `deleteVerts` are stated for an arbitrary
    39vertex type, since both are pointwise notions and the proofs consuming
    40them work over intermediate carriers.
    41-/
    42
    43namespace Lax12.UniformQuasiWideness
    44
    45open Lax12.GraphClasses
    46
    47/-- A set of vertices is distance-`r` independent in `G` when every walk
    48between two distinct members is longer than `r`. -/
    49def DistIndependent {V : Type*} (G : SimpleGraph V) (r : ℕ) (A : Set V) : Prop :=
    50 A.Pairwise fun u v => ∀ p : G.Walk u v, r < p.length
    51
    52/-- `G` with the vertices of `S` isolated: every edge incident to `S` is
    53removed and the vertex type is unchanged. This models `G − S`. -/
    54def deleteVerts {V : Type*} (G : SimpleGraph V) (S : Set V) : SimpleGraph V where
    55 Adj u v := G.Adj u v ∧ u ∉ S ∧ v ∉ S
    56 symm _ _ h := ⟨h.1.symm, h.2.2, h.2.1
    57 loopless := ⟨fun v h => G.loopless.irrefl v h.1
    58
    59/-- A graph class is uniformly quasi-wide if for every radius `r` there
    60are a threshold function `N` and a separator bound `s` such that in
    61every member, every vertex set of size at least `N m` contains a
    62distance-`r` independent subset of size at least `m` after deleting at
    63most `s` vertices. -/
    64def UniformlyQuasiWide (C : GraphClass) : Prop :=
    65 ∀ r : ℕ, ∃ (N : ℕ → ℕ) (s : ℕ),
    66 ∀ (m n : ℕ) (G : SimpleGraph (Fin n)), C n G →
    67 ∀ A : Set (Fin n), N m ≤ A.ncard →
    68 ∃ S B : Set (Fin n),
    69 S.ncard ≤ s ∧ B ⊆ A \ S ∧ m ≤ B.ncard ∧
    70 DistIndependent (deleteVerts G S) r B
    71
    72end Lax12.UniformQuasiWideness
    73

    Formalization notes

    Distance is stated with walks: two vertices are at distance more than rr exactly when every walk between them is longer than rr, which needs no metric, connectivity or decidability instance. Deleting a vertex set is modelled by isolating it — deleteVertsGSdeleteVerts G S keeps the carrier and drops every edge incident to SS — so every set in the statement lives in the same vertex type and no subtype plumbing enters the surface. Since the witness satisfies BA SB ⊆ A \ S, the isolated vertices are not in BB and distance-rr independence in the isolated graph is the same as in the induced subgraph on the complement of SS.

    Sets and Set.ncardSet.ncard are used throughout, as in the other concepts of this submission. The threshold NN may depend on the requested size mm, while the separator bound ss may not: that uniformity in ss is the "uniform" of uniform quasi-wideness and is the whole strength of the notion. DistIndependentDistIndependent and deleteVertsdeleteVerts are stated for an arbitrary vertex type, since both are pointwise notions and the proofs consuming them work over intermediate carriers.

    Community review

    Discussion

    Ask a question or add context. Endorsements and structured flags are kept in the review panel above; your ORCID profile must share a public name.

    0 comments

    Loading discussion…