Lemma 9.31 — for with
ProvedYeIPA.QPApprox.arcsin_entrywise_dominatesFor a real matrix let be the matrix with entries , and let . If is positive semidefinite and every diagonal entry satisfies , then
that is, is positive semidefinite.
(Every entry of such an satisfies , so is well defined.) This is the matrix inequality that turns the expected value of the sign rounding into a bound in terms of the relaxation in the proof of Theorem 9.33.
Formalization Note The book writes ; this is stated as for every . is Mathlib's PosSemidef (symmetric and positive semidefinite). The book leaves the proof as Exercise 9.12.
import Mathlib import Definitions.Def_YeIPA_QPApprox_Relaxation open Matrix MeasureTheory
namespace YeIPA.QPApprox
variable {m n : ℕ}
/-- **Lemma 9.31** (Ye, *Interior Point Algorithms*, Wiley 1997, p. 328).
Let `X ⪰ 0` and `d(X) ≤ 1`. Then `arcsin[X] ⪰ X`, i.e. `arcsin[X] − X` is positive
semi-definite, where `arcsin[X]` is the matrix with entries `arcsin(x_ij)`. -/
theorem arcsin_entrywise_dominates (X : Matrix (Fin n) (Fin n) ℝ) (hX : X.PosSemidef)
(hd : ∀ j, diagVec X j ≤ 1) :
(entrywise Real.arcsin X - X).PosSemidef := by sorry
end YeIPA.QPApprox
Read-back
What the Lean code literally says, in plain math · claude-opus-5-5
Let be any natural number and a real matrix. The hypotheses are:
- is positive semidefinite: symmetric, with for all ;
- every diagonal entry satisfies .
The statement asserts that the matrix
is positive semidefinite. That means it is symmetric and
Here is the real arcsine, clamped to outside . Under the two hypotheses every entry satisfies , so no clamping actually occurs.
Degenerate cases.
- . The empty matrix satisfies both hypotheses and the conclusion holds trivially.
- . The statement says for every .
- . The conclusion is about the zero matrix.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.