RobustOptimization.isPsi_unique
ProvedThe value of a setup is unique.
For fixed norms, closed convex cones, and linear map , there is at most one real number satisfying the predicate. In other words, is a well-defined function of the setup: any two purported values must coincide, since each is characterized as the least upper bound of the same set of distances.
This functionality is what turns the duality symmetry (a biconditional about a single value) into an equality between the two a priori distinct values and in Proposition 11.4.1.
Formal statement
import Definitions.Def_RobustOptimization_GRCConic
import Mathlib
open Matrix Finset
namespace RobustOptimization
/-- The Psi value of a setup is unique.
If two real numbers both satisfy the `IsPsi` predicate for the same setup
(norms, cones, linear map), they are equal. This is the functionality of
Psi: the least-upper-bound characterization determines the value. -/
theorem isPsi_unique {p q : ℕ} (NE : (Fin p → ℝ) → ℝ) (NF : (Fin q → ℝ) → ℝ)
(KE : Set (Fin p → ℝ)) (KF : Set (Fin q → ℝ))
(M : Matrix (Fin q) (Fin p) ℝ) (ψ₁ ψ₂ : ℝ)
(h₁ : IsPsi NE NF KE KF M ψ₁) (h₂ : IsPsi NE NF KE KF M ψ₂) :
ψ₁ = ψ₂ := by sorry
end RobustOptimization