Theorem 10.4 — Separation Theorem for polyhedra
OpenVanderbeiLP.StrictComp.separation_polyhedraconvex-analysisp2o-batch-b23ap2o-gran-per-chapterp2o-plan-bookp2o-v1polyhedraseparation
Let and be two polyhedra in , i.e. sets of the form and . If and are both nonempty and disjoint, then there exist halfspaces and with
Here a halfspace is a set with , so the two separating sets are genuine halfspaces with nonzero (and necessarily opposite-pointing) normals, not all of or the empty set.
The theorem is the polyhedral case of the separating hyperplane theorem, obtained without any topology.
Formalization Note The book's "" is inclusion (not necessarily proper). Both nonemptiness hypotheses are kept: they are what forces the normals to be nonzero.
Preamble
import Mathlib import Definitions.Def_VanderbeiLP_StrictComp_Polyhedron
Formal statement
namespace VanderbeiLP.StrictComp
/-- **Vanderbei, Theorem 10.4 (p. 145), Separation Theorem for polyhedra.** Two disjoint
nonempty polyhedra `P`, `P̃` of `ℝⁿ` lie in disjoint halfspaces `H ⊇ P`, `H̃ ⊇ P̃`
(halfspaces in the sense of (10.3): nonzero normal). -/
theorem separation_polyhedra {n : ℕ} (P Ptil : Set (Fin n → ℝ))
(hP : IsPolyhedron P) (hPtil : IsPolyhedron Ptil)
(hPne : P.Nonempty) (hPtilne : Ptil.Nonempty) (hdisj : Disjoint P Ptil) :
∃ H Htil : Set (Fin n → ℝ), IsHalfspace H ∧ IsHalfspace Htil ∧ Disjoint H Htil ∧
P ⊆ H ∧ Ptil ⊆ Htil := by sorry
end VanderbeiLP.StrictComp
Source
Vanderbei, Linear Programming: Foundations and Extensions, 4th ed., Springer 2014, p. 145, Theorem 10.4 (PDF p. 158); halfspace Eq. (10.3), p. 144
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.