Weighted two-sided comparison between the rectangular Bregman and Mazur-distance objectives
ProvedHlawkaSchatten.rectangularBregmanObjective_two_sidedLet be finite-dimensional complex inner-product spaces and let be any real number (no hypothesis is placed on in this statement). For a complex-linear map , write for its Hermitian dilation, the self-adjoint operator on given by (hermitianDilation). For a self-adjoint operator on and , write for the operator obtained by applying to the eigenvalues of in any orthonormal eigenbasis of ; this operator itself does not depend on which eigenbasis is chosen, since it acts as on every -eigenvector of , not only on the vectors of one particular chosen eigenbasis — which is what lets below be a single well-defined operator even though and generally have different eigenbases. With , , and the scalar Mazur map (powerPotential, powerGradient, scalarMazur; all three are totalized real powers, defined for every real and ), set
(dilatedBregmanTrace, the trace-level Bregman combination of between and ) and
(dilatedMazurDistanceSq, the squared Hilbert–Schmidt distance between the Mazur images of the two dilations). Separately, let be the rectangular Mazur map of — the unique complex-linear map whose dilation is , i.e. (rectangularMazurMap); it satisfies , where is the Schatten- (Hilbert–Schmidt) norm.
For with singular values , the Schatten -power sphere is (schattenPowerSphere; the sum runs over the nonzero singular values, which for is the same as summing over all of them). Let be a finite index type, let with for every , let be a family on the Schatten -power sphere, and let be a further point on that sphere. Define the rectangular Bregman objective and rectangular Mazur-distance objective
(the second equals as well, since ). Assume real numbers satisfy the pointwise two-sided bound
for every pair of complex-linear maps . Then the theorem states
This lifts a pairwise two-sided bound between and , assumed uniformly over rectangular operators, to the weighted finite-family objectives that the variational argument minimizes over the Schatten power sphere, with the same constants .
Formalization Note Both named objectives already carry the same normalization relative to and respectively (one directly, the other through the identity ), so the displayed two-sided bound needs no extra factor on either side. No convexity of is claimed or needed here, since ranges over all of ; and are used purely as totalized algebraic quantities, and the two-sided bound between them is assumed as a hypothesis on this page, not derived from convexity.
import Definitions.Def_HlawkaSchatten_HermitianDilation
import Definitions.Def_HlawkaSchatten_SchattenNorm
import Definitions.Def_HlawkaSchatten_Variational
import Mathlib.Analysis.Calculus.LHopital
import Mathlib.Analysis.Convex.Deriv
import Mathlib.Analysis.Convex.SpecificFunctions.Basic
import Mathlib.Analysis.InnerProductSpace.Adjoint
import Mathlib.Analysis.InnerProductSpace.Basic
import Mathlib.Analysis.InnerProductSpace.Dual
import Mathlib.Analysis.InnerProductSpace.NormPow
import Mathlib.Analysis.InnerProductSpace.PiL2
import Mathlib.Analysis.InnerProductSpace.ProdL2
import Mathlib.Analysis.InnerProductSpace.SingularValues
import Mathlib.Analysis.InnerProductSpace.Trace
import Mathlib.Data.Sign.Basic
import Mathlib.LinearAlgebra.Matrix.Charpoly.Basic
import Mathlib.Topology.Compactification.OnePoint.Basic
import Mathlib.Topology.Instances.Sign
/-
Copyright (c) 2026 Ezzeri Esa. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Ezzeri Esa
-/
/-!
# Variational minima for the Bregman--Mazur argument
This file proves two reusable parts of the variational layer. First, a
pointwise two-sided comparison transports to attained global minima, even
when the two objectives are indexed by different but equivalent spheres.
Second, the weighted squared-distance objective on a Hilbert unit sphere has
the exact minimum used in the Schatten argument.
-/
open scoped InnerProductSpace ComplexConjugate
variable {𝕜 H ι : Type*} [RCLike 𝕜] [Fintype ι]
[NormedAddCommGroup H] [InnerProductSpace 𝕜 H]
section RectangularMazur
variable {E F κ : Type*} [Fintype κ]
[NormedAddCommGroup E] [InnerProductSpace ℂ E] [FiniteDimensional ℂ E]
[NormedAddCommGroup F] [InnerProductSpace ℂ F] [FiniteDimensional ℂ F]
open HlawkaSchatten
theorem HlawkaSchatten.rectangularBregmanObjective_two_sided
(p m M : ℝ) (a : ι → ℝ) (ha : ∀ i, 0 ≤ a i)
(u : ι → schattenPowerSphere (𝕜 := ℂ) (E := E) (F := F) p)
(v : schattenPowerSphere (𝕜 := ℂ) (E := E) (F := F) p)
(hbound : ∀ S T : E →ₗ[ℂ] F,
m * dilatedMazurDistanceSq p S T ≤ dilatedBregmanTrace p S T ∧
dilatedBregmanTrace p S T ≤ M * dilatedMazurDistanceSq p S T) :
m * rectangularMazurDistanceObjective p a u v ≤
rectangularBregmanObjective p a u v ∧
rectangularBregmanObjective p a u v ≤
M * rectangularMazurDistanceObjective p a u v := by sorry