Error bound for polynomial interpolation
ProvedMetodosNumericos.lagrange_error_boundinterpolationnumerical-analysis
Under the hypotheses of the goal theorem, if in addition for every , then
This is Proposição 7.3.2.
Preamble
import Mathlib import Definitions.Def_MetodosNumericos_interpolacaoDefs
Formal statement
namespace MetodosNumericos
theorem lagrange_error_bound {n : ℕ} (f : ℝ → ℝ) (a b M : ℝ) (xs : Fin (n + 1) → ℝ)
(hxs : StrictMono xs) (hmem : ∀ i, xs i ∈ Set.Ioo a b)
(hf : ContDiffOn ℝ (n + 1 : ℕ) f (Set.Ioo a b))
(hM : ∀ y ∈ Set.Ioo a b, |iteratedDerivWithin (n + 1) f (Set.Ioo a b) y| ≤ M)
(x : ℝ) (hx : x ∈ Set.Ioo a b) (hne : ∀ i, x ≠ xs i) :
|f x - lagrangeInterp xs (fun i => f (xs i)) x| ≤
M / (Nat.factorial (n + 1) : ℝ) * ∏ k : Fin (n + 1), |x - xs k| := by sorry
end MetodosNumericosSource
S. R. Freitas, Métodos Numéricos (UFMS, 2000), Cap. 7, Proposição 7.3.2, pp. 141–142.
Read-back
What the Lean code literally says, in plain math · self-authored-by-drafting-agent (non-blind)
Disclosure: this read-back is not blind. It was written by the same agent that drafted the Lean statement, at the explicit instruction of the mission's human owner, and not by an independent auditor with fresh context.
For a natural number , a function , reals , , and a family of reals, the hypotheses are:
- is strictly increasing;
- every lies in the open interval ;
- is continuously differentiable of order on ;
- for every , the absolute value of the -st derivative of computed within at is at most ;
- the point lies in and differs from every node.
The conclusion is
The values interpolated are the values of itself at the nodes. The bound is assumed only on , and the inequality is non-strict.
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.