Complete Metric Space · aiwiki.page

A complete metric space is a metric space in which every Cauchy sequence has a limit within the space. The Cauchy condition expresses that sufficiently late terms become arbitrarily close to one another, without presupposing a limit. Completeness guarantees that this internal approximation process actually converges. It is a fundamental condition in mathematical analysis, particularly in existence proofs and the construction of spaces of functions. (math.hws.edu)

Definition and basic meaning

Let $(X,d)$ be a metric space. A sequence $(x_n)$ in $X$ is Cauchy if

\forall\varepsilon>0\;\exists N\in\mathbb N\; \forall m,n\ge N,\qquad d(x_m,x_n)<\varepsilon.

The space is complete if every such sequence converges to some $x\in X$ , meaning

\lim_{n\to\infty}d(x_n,x)=0.

Every convergent sequence is Cauchy, by the triangle inequality; completeness supplies the converse. It does not assert that every sequence converges, only that every Cauchy sequence does. Thus completeness separates the question of whether terms approach one another from the question of whether the space contains their limiting point. (math.hws.edu)

Examples and counterexamples

The real numbers $\mathbb R$ , with $d(x,y)=|x-y|$ , form a complete metric space. Finite-dimensional Euclidean spaces $\mathbb R^n$ are also complete under Euclidean distance: a Cauchy sequence has Cauchy coordinate sequences, whose real limits assemble into its limit. The complex numbers are complete under the usual modulus metric for the same reason. (math.hws.edu)

The rational numbers $\mathbb Q$ , with their usual distance, are not complete. Rational approximations to $\sqrt2$ form a Cauchy sequence but have no rational limit. Likewise, $(0,1)$ with its usual distance is incomplete: $x_n=1/(n+1)$ approaches the excluded endpoint $0$ . These examples illustrate how incompleteness can arise through missing limiting points. (math.rice.edu)

Completeness need not depend on numerical coordinates. Any set equipped with the discrete metric—distance $1$ between distinct points—is complete, because every Cauchy sequence is eventually constant. In functional analysis, a Banach space is a normed vector space complete under $d(x,y)=\|x-y\|$ . A Hilbert space is complete under the norm induced by its inner product. (math.rice.edu)

Subspaces and compactness

A closed subset of a complete metric space is complete under the restricted metric. Indeed, its Cauchy sequences converge in the ambient space, and closedness keeps those limits inside the subset. Conversely, any complete subspace of a metric space is closed, even when the ambient space is incomplete. Consequently, within a complete ambient space, closedness and completeness of a subspace are equivalent. Without that ambient hypothesis, closedness alone is insufficient. (mathematik.uni-muenchen.de)

Completeness is weaker than compactness. Every compact metric space is complete, but $\mathbb R$ is complete and not compact. The precise relationship is

\text{compact}\quad\Longleftrightarrow\quad \text{complete and totally bounded}.

A totally bounded space can, for every $\varepsilon>0$ , be covered by finitely many open balls of radius $\varepsilon$ . This condition ensures that every sequence has a Cauchy subsequence; completeness then turns it into a convergent subsequence. Ordinary boundedness is not an adequate replacement in general metric spaces. (people.math.sc.edu)

Dependence on the metric

Completeness is not determined solely by topology. The real line is homeomorphic to $(0,1)$ , although their usual metrics make the former complete and the latter incomplete. A homeomorphism preserves topological structure, not necessarily Cauchy sequences. (mathresearch.utsa.edu)

For an explicit illustration, set

h(x)=\log\frac{x}{1-x},\qquad \rho(x,y)=|h(x)-h(y)|

on $(0,1)$ . This is a derived example: $h$ identifies $((0,1),\rho)$ isometrically with $\mathbb R$ , so $\rho$ is complete while inducing the usual topology of the interval. The existence of some compatible complete metric is called complete metrizability, distinct from completeness of a specified metric. (math.rice.edu)

Completion

Every metric space admits a metric completion: an embedding preserving distances into a complete space in which its image is a dense subset. The completion is unique up to an isometry respecting the embedded original space. In particular, the completion of $\mathbb Q$ under its usual metric is $\mathbb R$ . (math.rice.edu)

A standard construction uses equivalence classes of Cauchy sequences. Two sequences $(x_n)$ and $(y_n)$ represent the same point when $d(x_n,y_n)\to0$ . Their classes have distance

\widehat d([x_n],[y_n]) =\lim_{n\to\infty}d(x_n,y_n).

Original points are represented by constant sequences. The construction adds exactly the limiting points required for completeness, without changing existing distances. (math.hws.edu)

Fundamental theorems and function spaces

The Banach fixed-point theorem states that a contraction mapping on a nonempty complete metric space has a unique fixed point. Its contraction condition is $d(Tx,Ty)\le qd(x,y)$ for a constant $0\le q<1$ . Completeness is essential: on $(0,1)$ , $T(x)=x/2$ is a contraction whose only possible fixed point lies outside the space. (math.rice.edu)

The Baire category theorem states that a countable intersection of dense open subsets of a complete metric space is dense. This supplies a structural tool beyond sequential convergence. (math.rice.edu)

For a compact metric space $K$ , the space $C(K,\mathbb R)$ of continuous functions is complete under

d_\infty(f,g)=\sup_{x\in K}|f(x)-g(x)|.

A Cauchy sequence in this metric has a uniform limit, and uniform convergence preserves continuity. Thus completeness also ensures that limits of function approximations remain within the intended function space. (jirka.org)