Layer families of tight sets of a simple polytope have ridge incidence at most two
ProvedHirsch.polytopal_clf_ridge_incidence_twoLet be a bounded simple H-polytope ( tight rows at every vertex) and let be any layer family whose bases are tight sets of vertices of (for instance the distance layers Hirsch.distance_layers_clf). Then every -set of rows is contained in at most two bases of the whole family.
At a simple vertex the tight normals are linearly independent, so any of them cut out a face of dimension one, an edge with exactly two vertices; a -set of rows tight at a vertex therefore lies in at most two tight sets. This is the polytopal incidence axiom ; with saturation it would give a linear bound (Hirsch.clf_saturated_ridge_incidence_linear_bound), but polytopal distance layers are not saturated (already the -cube fails), which locates the gap between the abstraction and the conjecture.
import Mathlib import Definitions.Def_Hirsch_model import Definitions.Def_Hirsch_walk import Definitions.Def_Hirsch_clf open scoped RealInnerProductSpace
namespace Hirsch
theorem polytopal_clf_ridge_incidence_two (d n : ℕ)
(a : Fin n → EuclideanSpace ℝ (Fin d)) (b : Fin n → ℝ)
(hbd : Bornology.IsBounded (Hpoly a b))
(hsimple : ∀ x ∈ Set.extremePoints ℝ (Hpoly a b),
(Finset.univ.filter (fun i : Fin n => ⟪a i, x⟫ = b i)).card = d)
(F : CLF (Fin n) d)
(hF : ∀ t, ∀ B ∈ F.layer t, ∃ x ∈ Set.extremePoints ℝ (Hpoly a b),
B = Finset.univ.filter (fun i => ⟪a i, x⟫ = b i)) :
CLF.RidgeIncidenceLE F 2 := by sorry
end Hirsch