No public endorsements yet.
Loading review…
Sign in with ORCIDConcept map
Evidence
Each proof establishes this claim relative to its assumptions.
Theorem
This theorem specification is proved in . The outer-layer threshold sixteen allows an unoptimized endgame; it is sufficient for the 216-point convex-position application. Chain replacement, all endpoint cases, and the four-layer reduction are proved in Lean. The empty-hexagon proof uses this theorem interface so Lax records the dependency on the proof supplied in this package.
Reference: Pavel Valtr, "On Empty Hexagons", Section 3 (Section 2 in the author's preprint at https://kam.mff.cuni.cz/~valtr/h.ps).
Lean source view on GitHub
| 1 | import Lax56.ConvexLayers |
| 2 | |
| 3 | /-! |
| 4 | --- |
| 5 | title: Valtr's four-layer lemma |
| 6 | type: theorem |
| 7 | --- |
| 8 | This theorem specification is proved in `Lax56Proofs.ValtrFourLayer`. |
| 9 | The outer-layer threshold sixteen allows an unoptimized endgame; it is |
| 10 | sufficient for the 216-point convex-position application. Chain replacement, |
| 11 | all endpoint cases, and the four-layer reduction are proved in Lean. The |
| 12 | empty-hexagon proof uses this theorem interface so Lax records the dependency |
| 13 | on the proof supplied in this package. |
| 14 | |
| 15 | Reference: Pavel Valtr, "On Empty Hexagons", Section 3 (Section 2 in the |
| 16 | author's preprint at https://kam.mff.cuni.cz/~valtr/h.ps). |
| 17 | -/ |
| 18 | |
| 19 | namespace Lax56.ValtrFourLayer |
| 20 | |
| 21 | open Lax56.Geometry Lax56.ConvexLayers |
| 22 | |
| 23 | /-- The weakened four-layer lemma requested for the unoptimized proof. |
| 24 | `layer S 3` is the fourth layer because layer indices start at zero. -/ |
| 25 | axiom exists_emptyHexagon_of_four_layers |
| 26 | (S : Finset Point) (hgeneral : ¬HasThreeCollinear S) |
| 27 | (hminimal : MinimalOuter S) (hlarge : 16 ≤ (extremeLayer S).card) |
| 28 | (hfourth : (layer S 3).Nonempty) : HasEmptyHexagon S |
| 29 | |
| 30 | end Lax56.ValtrFourLayer |
| 31 |
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