Proof of Theorem 4.1, p. 56 — a symmetric matrix is a P-matrix iff it is positive definite
ProvedRobinsonSR.NLP.pmatrix_iff_posdef_of_symmp-matrixp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1positive-definiteprincipal-minorssymmetric-matrix
Let be a symmetric real square matrix indexed by a finite set. Then all principal minors of are positive if and only if is positive definite:
In the proof of Theorem 4.1 this reduces the P-matrix requirement on the Schur complement (4.6) to positive definiteness.
Formalization Note Positive definiteness is stated without a symmetry clause (the paper's convention in §3); symmetry is the separate hypothesis. For an empty index set both sides hold vacuously.
Preamble
import Mathlib import Definitions.Def_RobinsonSR_NLP_Setting open scoped RealInnerProductSpace Matrix
Formal statement
namespace RobinsonSR.NLP
/-- Proof of Theorem 4.1, p. 56: a symmetric real matrix is a P-matrix (all principal minors
positive) if and only if it is positive definite. -/
theorem pmatrix_iff_posdef_of_symm {ι : Type*} [Fintype ι] [DecidableEq ι]
(M : Matrix ι ι ℝ) (hM : M.IsSymm) :
IsPMatrix M ↔ RobinsonSR.Schur.PosDefNS M := by sorry
end RobinsonSR.NLP
Source
Robinson, Strongly regular generalized equations, Math. Oper. Res. 5 (1980), p. 56, proof of Theorem 4.1; P-matrix defined p. 52
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.