§2, p. 5 — σ_i observation: J(L) = {i : σ_i(L) > 0} and σ(L) = min_{i∈J(L)} σ_i(L)
ProvedProjRescale.MaxSupport.sigma_i_observationcondition-measuremaximum-supportp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1
Let be a linear subspace and, for , let . Then
(the second identity whenever , the case in which the minimum is defined).
So the vector encodes both and ; the analysis of the rescaling step (Lemma 1) and of Algorithm 1 (Proposition 1) tracks these coordinates.
Differences from the source: none.
Preamble
import Mathlib import Definitions.Def_ProjRescale_MaxSupport_Setting
Formal statement
namespace ProjRescale.MaxSupport
open Classical in
/-- §2, p. 5: `J(L) = {i : σ_i(L) > 0}` and `σ(L) = min_{i ∈ J(L)} σ_i(L)`. -/
theorem sigma_i_observation {n : ℕ} (L : Submodule ℝ (E n)) :
maxSupp L = Finset.univ.filter (fun i => 0 < sigmaI L i) ∧
∀ h : (maxSupp L).Nonempty, condMeasure L = (maxSupp L).inf' h (sigmaI L) := by sorry
end ProjRescale.MaxSupport
Source
Peña & Soheili, Projection and rescaling algorithm for finding maximum support solutions to polyhedral conic systems, arXiv:2003.08911v3, p. 5, §2 (unnumbered observation after the definition of σ_i(L))
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.