Eq. (9) — rescaling error and magnitude bounds
ProvedQSPHamSim.Main.rescale_boundhamiltonian-simulationp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1quantum-signal-processingtrigonometric-polynomials
Suppose real functions approximate the unit-modulus target in the sense for every real . Put and . Then
For the squared magnitude is at least . If also , then ; there is an angle with .
The rescaling makes the approximants eligible for the norm constraint in Theorem 1.
Formalization Note The approximation forces . The upper magnitude bound is implicit in the paper's reason for rescaling; the lower bound requires the stated small-error condition.
Preamble
import Mathlib import Definitions.Def_QSPHamSim_Main_Setting set_option autoImplicit false
Formal statement
namespace QSPHamSim.Main
theorem rescale_bound (A C h : ℝ → ℝ) (ε : ℝ)
(happrox : ∀ θ : ℝ,
‖((A θ : ℂ) + (C θ : ℂ) * Complex.I) -
Complex.exp ((h θ : ℂ) * Complex.I)‖ ≤ ε) :
let A₁ := fun θ : ℝ => A θ / (1 + ε)
let C₁ := fun θ : ℝ => C θ / (1 + ε)
(∀ θ : ℝ,
‖((A₁ θ : ℂ) + (C₁ θ : ℂ) * Complex.I) -
Complex.exp ((h θ : ℂ) * Complex.I)‖ ≤ 2 * ε ∧
A₁ θ ^ 2 + C₁ θ ^ 2 ≤ 1) ∧
(ε ≤ 1 → ∀ θ : ℝ,
((1 - ε) / (1 + ε)) ^ 2 ≤ A₁ θ ^ 2 + C₁ θ ^ 2) ∧
(h 0 = 0 → C 0 = 0 →
(1 - ε) / (1 + ε) ≤ A₁ 0 ∧ A₁ 0 ≤ 1 ∧
∃ δ : ℝ, Real.cos δ = A₁ 0 ∧
(1 - ε) / (1 + ε) ≤ Real.cos δ) := by sorry
end QSPHamSim.Main
Source
Low, Chuang, Optimal Hamiltonian Simulation by Quantum Signal Processing, arXiv:1606.02685v2, p. 3, Eq. (9) and following paragraph
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.