Summation by parts (Abel's lemma)
ProvedFamousTheorems.summation_by_partsSummation by parts (Abel's lemma). Let be scalars in a ring and vectors in an -module, and write . For ,
This is the discrete analogue of integration by parts. It is the key identity behind the Abel and Dirichlet tests for convergence of series, Abel's theorem on power series at the boundary, and many estimates in analytic number theory for sums weighted by arithmetic functions.
Formalization note. Mathlib's Finset.sum_Ico_by_parts. Sums over Finset.Ico m n run over and Finset.range k over . The scalars act on the module by •.
Preamble
import Mathlib
Formal statement
namespace FamousTheorems
theorem summation_by_parts {R M : Type*} [Ring R] [AddCommGroup M] [Module R M] (f : ℕ → R) (g : ℕ → M) {m n : ℕ}
(hmn : m < n) :
∑ i ∈ Finset.Ico m n, f i • g i =
f (n - 1) • ∑ i ∈ Finset.range n, g i - f m • ∑ i ∈ Finset.range m, g i -
∑ i ∈ Finset.Ico m (n - 1), (f (i + 1) - f i) • ∑ j ∈ Finset.range (i + 1), g j := by sorry
end FamousTheoremsSource
Listed in Mathlib's curated theorem manifests (docs/1000.yaml); formalized in Mathlib as `Finset.sum_Ico_by_parts`. Proof here reduces to that Mathlib result.