Hypersphere activity sets: sup_{aⱼ∈Kⱼ} eᵢ·aⱼ = aᵢⱼ + ρⱼ, so āⱼ = aⱼ + ρⱼ·e
ProvedSoysterInexactLP.Auxiliary.hyperBall_supportp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1robust-optimizationsupport-function
Let be an matrix and . For the Euclidean balls ,
that is, with the vector of all ones, so .
This computes the auxiliary linear program explicitly in the paper's inexact linear programming application.
Formalization Note The radii are assumed nonnegative (a "magnitude"); a negative radius would give an empty set. The norm is Euclidean.
Preamble
import Mathlib import Definitions.Def_SoysterInexactLP_Auxiliary_Problems open Pointwise Matrix
Formal statement
namespace SoysterInexactLP.Auxiliary
theorem hyperBall_support {m n : ℕ} (A₀ : Matrix (Fin m) (Fin n) ℝ)
(ρ : Fin n → ℝ) (hρ : ∀ j, 0 ≤ ρ j) :
(∀ i j, supportFun (hyperBall A₀ ρ j) (Pi.single i 1) = ((A₀ i j + ρ j : ℝ) : EReal)) ∧
Abar (hyperBall A₀ ρ) = A₀ + Matrix.of (fun _ j => ρ j) := by sorry
end SoysterInexactLP.Auxiliary
Source
Soyster, Convex Programming with Set-Inclusive Constraints and Applications to Inexact Linear Programming, Operations Research 21 (1973), p. 1157 (PDF p. 5), §Inexact Linear Programming, second paragraph
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.