Lebesgue's lemma

An approximation bound stating that a bounded linear projection's error is at most one plus its operator norm times the best attainable error from the target subspace.

Version

v2 · 2026-09-19 · History

Domain-specific #

5290

Domain group

Formal Sciences

Origin domain

Mathematics

Subdomain

Approximation Theory → Mathematics

Core Idea

For a normed space X, subspace U and projection P onto U, the inequality norm(v-Pv) no greater than (1+norm P) times inf over u in U of norm(v-u) establishes quasi-optimality. Subtracting an arbitrary u in the range, using P(u)=u and applying the triangle and operator-norm inequalities relates projection error to any competitor, then infimization gives the bound. The abstraction is therefore identified by a declared carrier, a transformation or constraint over that carrier, and an invariant that tells an analyst whether the named structure is genuinely present.

How would you explain it like I'm…

Helper Never Too Far

You want to get close to a treasure, but you have to stand on a path. A helper rule picks a spot on the path for you. Lebesgue's lemma promises the helper's spot is never more than a certain number of times farther from the treasure than the very best spot. That number depends on how much the helper's rule can stretch things.

Nearly-Best Approximation Promise

Sometimes we want to replace a complicated thing with the closest thing from a simpler collection. Finding the exact best choice can be hard, so we often use a fixed rule, called a projection, that sends things into the simpler collection. A projection leaves things already in the collection unchanged. Lebesgue's lemma says the error of the rule's choice is at most (1 + how much the rule can stretch things) times the error of the best possible choice. So a rule that doesn't stretch much is guaranteed to be nearly as good as the best.

Projection Quasi-Optimality Bound

Lebesgue's lemma works in a normed space X, a space with a way to measure lengths, together with a subspace U of allowed approximations. P is a projection onto U, a linear map that lands in U and leaves points of U unchanged. The lemma states that the error ‖v − Pv‖ is at most (1 + ‖P‖) times the best possible error, the infimum of ‖v − u‖ over all u in U. Here ‖P‖ measures the most P can stretch any vector. The proof is short: for any u in U, v − Pv = (v − u) − P(v − u) because Pu = u, so the triangle inequality gives the bound. Taking the best u gives the result. This shows that projection-based approximations are 'quasi-optimal': within a fixed factor of the best.

 

Let X be a normed space, U ⊆ X a subspace, and P: X → X a bounded linear projection onto U, so that P² = P, range P = U and Pu = u for u ∈ U. Lebesgue's lemma states that ‖v − Pv‖ ≤ (1 + ‖P‖) inf_{u∈U} ‖v − u‖ for every v ∈ X. Proof: for any u ∈ U, v − Pv = (v − u) − P(v − u). So ‖v − Pv‖ ≤ ‖v − u‖ + ‖P‖‖v − u‖, and taking the infimum over u gives the bound. This is a quasi-optimality estimate: the projection's error is within the constant 1 + ‖P‖ of the best-approximation error. It is the standard tool for converting operator-norm bounds on interpolation or projection operators into approximation-error bounds. Applying it requires making each ingredient explicit: the norm, the subspace, boundedness and linearity of P, the operator norm, and the infimum.

Scope of Application

Lebesgue's lemma belongs to approximation theory and is useful where the analyst can specify the typed approximation theory carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets, then evaluate the normed space, approximation subspace, bounded linear projection onto it, operator norm, target vector and best-approximation infimum are explicit and the stated inequality follows. The scope is broad within that domain but bounded by the need for the normed space, approximation subspace, bounded linear projection onto it, operator norm, target vector and best-approximation infimum are explicit and the stated inequality follows. The entry records a descriptive analytical identity; practical use requires the governing domain's evidence, standards, and safety obligations.

Clarity

The abstraction clarifies a crowded vocabulary by making the normed space, approximation subspace, bounded linear projection onto it, operator norm, target vector and best-approximation infimum are explicit and the stated inequality follows the center of the account. A claim should name the carrier, the governing operation or relation, the applicable assumptions, and the recognition test. A bare label is insufficient because the name Lebesgue's lemma can be used for a formal identity, an implementation, or a neighboring result unless carrier and convention are stated.

Manages Complexity

Without the abstraction, an analyst must reason directly over many local details: the carrier roles, admissibility assumptions, competing conventions, derived invariants, boundary cases, and proof or validation obligations specific to Lebesgue's lemma. Lebesgue's lemma compresses them into the roles in the structural signature. That compression permits comparison across instances without erasing the variables that determine validity. It also exposes which details may be varied safely and which are constitutive.

Abstract Reasoning

1. Identify the carrier. State what the elements, states, objects, or observations are: the typed approximation theory carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets. Reject examples whose alleged carrier belongs to a different problem.
2. Lock the constitutive rule. Express the normed space, approximation subspace, bounded linear projection onto it, operator norm, target vector and best-approximation infimum are explicit and the stated inequality follows independently of one notation or implementation.

Knowledge Transfer

Knowledge transfers strongly among subfields of approximation theory because they reuse the typed approximation theory carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets, Subtracting an arbitrary u in the range, using P(u)=u and applying the triangle and operator-norm inequalities relates projection error to any competitor, then infimization gives the bound., and type the carrier, state every parameter and convention in the definition, test that the normed space, approximation subspace, bounded linear projection onto it, operator norm, target vector and best-approximation infimum are explicit and the stated inequality follows, compare the nearest accepted identity, and report counterexamples, uncertainty, and limiting cases.

Relationships to Other Abstractions

Current abstraction Lebesgue's lemma Domain-specific

Parents (1) — outgoing typed relations

• Lebesgue's lemma is a kind of Approximation Prime

  What makes it its own entry: the domain-specific identity determined by the normed space, approximation subspace, bounded linear projection onto it, operator norm, target vector and best-approximation infimum are explicit and the stated inequality follows.

Hierarchy path (1) — routes to 1 parentless root

• Lebesgue's lemma → is a kind of Approximation → is a decomposition of Representation → presupposes Abstraction

Neighborhood in Abstraction Space

Lebesgue's lemma sits in a crowded region of the domain-specific corpus (22^nd percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.

Family — Banach & Topological Vector Spaces (30 abstractions)

Nearest neighbors

• Unisolvent functions — 0.90
• Lomonosov's invariant subspace theorem — 0.90
• Ba space — 0.89
• Bounded operator — 0.89
• Banach–Mazur compactum — 0.89

Computed from structural-signature embeddings · 2026-10-25