(28), p. 11 — HKM and NT: ℰ⁻¹(σµe₁ − T_G(x,z)) = (σµ/γ²(z))(z⁰; −z¹) − x
ProvedSDPT3.ScaleInv.eq_28interior-pointp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1search-directionsecond-order-cone
Let lie in the interior of the second-order cone, let be either the HKM scaling (19) or the NT scaling (21) at , and let , . For all real ,
The left side is , the term needed in the right-hand side of the Schur complement equation (12). The paper states the formula for HKM and then notes that it also holds for NT.
Formalization Note enters only as a product, so and are arbitrary reals here; in (4), is shared by all blocks.
Preamble
import Mathlib import Definitions.Def_SDPT3_ScaleInv_Setting
Formal statement
namespace SDPT3.ScaleInv
open Matrix
theorem eq_28 {k : ℕ} (dir : Dir) (x z : Fin (k + 1) → ℝ) (hx : socInt x) (hz : socInt z)
(σ μ : ℝ) :
(calE (Gdir dir x z) z)⁻¹ *ᵥ ((σ * μ) • e1 k - TG (Gdir dir x z) x z) =
(σ * μ / gammaSq z) • (Jbar k *ᵥ z) - x := by sorry
end SDPT3.ScaleInv
Source
Tütüncü, Toh & Todd, Solving semidefinite-quadratic-linear programs using SDPT3, Math. Program. 95 (2003) 189–217; authors' copy, p. 11, (28), and the sentence before (30)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.