Normed space

1 Definition and axioms of a norm

A norm on a vector space \(V\) is a function \(\|\cdot\|:V\to[0,\infty)\) that measures the size of vectors in a way compatible with the vector-space structure. When such a function is specified, \(V\) becomes a normed space, providing a framework for talking about distances, convergence, and continuity without requiring an inner product.

1.1 Vector space prerequisites

A normed space starts with a vector space over a field (typically \(\mathbb{R}\) or \(\mathbb{C}\)). The norm must be compatible with scalar multiplication and addition, reflecting the linear structure of \(V\).

1.2 Norm properties (positivity, homogeneity, triangle inequality

A function \(\|\cdot\|\) is a norm if for all \(x,y\in V\) and all scalars \(a\):
1. Positivity: \(\|x\|\ge 0\), and \(\|x\|=0\) implies \(x=0\).
2. Homogeneity: \(\|ax\|=	a	\|x\|\).
3. Triangle inequality: \(\|x+y\|\le \|x\|+\|y\|\).

These axioms formalize the intuitive idea of length and ensure that “size” behaves predictably under scaling and addition.

1.3 Induced metric from a norm

Given a norm \(\|\cdot\|\), a metric \(d\) can be defined by

\[

d(x,y)=\|x-y\|.

\]

This satisfies the metric axioms because the norm guarantees symmetry, nonnegativity, and the triangle inequality (via \(\|(x-z)\|=\|(x-y)+(y-z)\|\)).

1.4 Basic consequences and examples of norms

Many norms arise from concrete measurements or from standard algebraic structures.

A key example is the absolute value norm on \(\mathbb{R}\) or \(\mathbb{C}\), where \(\|x\|=	x	\). More generally, norms on \(\mathbb{R}^n\) can be built from coordinate expressions that satisfy the three norm axioms.

1.4.1 Common inequalities derived from the norm

From the triangle inequality one also obtains the reverse triangle inequality: \[

\big	\|x\|-\|y\|\big	\le \|x-y\|.

\] This follows by applying the triangle inequality to \(x=(x-y)+y\) and similarly to \(y=(y-x)+x\).

1.5 Norm equivalence and rescaling

Two norms \(\|\cdot\|_1\) and \(\|\cdot\|_2\) on the same vector space are equivalent if there exist constants \(c,C>0\) such that

\[

c\|x\|_1\le \|x\|_2\le C\|x\|_1 \quad \text{for all } x\in V.

\] Equivalent norms define the same notion of convergence and continuity. Rescaling a norm by a positive constant preserves equivalence and keeps the induced metric structure uniformly comparable.

2 Convergence and continuity in normed spaces

Once a norm induces a metric, the standard language of limits and continuity applies. Much of analysis in normed spaces can be recast into statements about distances \(\|x-y\|\).

2.1 Sequences and limits

A sequence \((x_n)\) in a normed space converges to \(x\) if \[

\|x_n-x\|\to 0.

\] This convergence notion depends on the metric induced by the norm.

2.1.1 Cauchy sequences

A sequence \((x_n)\) is Cauchy if for every \(\varepsilon>0\) there exists \(N\) such that for all \(m,n\ge N\), \[

\|x_n-x_m\|<\varepsilon.

\] Cauchy sequences capture “internal” convergence behavior, regardless of whether the limiting point exists within the space.

2.2 Continuity of maps

A map \(T:V\to W\) between normed spaces is continuous at a point \(x\) if small changes in input (in norm) lead to small changes in output (in norm).

A standard characterization uses \(\varepsilon\)-\(\delta\) definitions in terms of induced metrics: \[ x_n\to x \implies T(x_n)\to T(x). \]

2.2.1 Lipschitz and Hölder-type estimates

If there is \(L\ge 0\) such that \[

\|T(x)-T(y)\|\le L\|x-y\| \quad \forall x,y,

\] then \(T\) is Lipschitz, hence continuous. More generally, Hölder-type bounds \[

\|T(x)-T(y)\|\le K\|x-y\|^\alpha \quad (0<\alpha\le 1)

\] also imply continuity, with the exponent controlling how rapidly errors shrink.

2.3 Boundedness and bounded sets

A function \(T\) can be bounded on subsets. A set \(A\subset V\) is bounded if there exists \(M\) such that \(\|x\|\le M\) for all \(x\in A\).

Boundedness is metric in nature: in normed spaces it is equivalent to boundedness under the induced distance.

2.4 Interior, closure, and boundaries (norm topology)

The norm topology consists of open sets determined by norms. A point \(x\) is in the interior of a set \(A\) if some ball around \(x\) stays inside \(A\). The closure of \(A\) is the smallest closed set containing it, equivalently the set of all points that can be approached by elements of \(A\). The boundary is the set of points where every neighborhood intersects both \(A\) and its complement.

3 Topological and geometric structure

Normed spaces have a rich geometric interpretation because distances come directly from vector magnitudes. Balls, neighborhoods, and convexity can be described in explicit geometric terms.

3.1 Open and closed balls

For \(x\in V\) and \(r>0\), the open ball and closed ball are \[

B(x,r)=\{y:\|y-x\|<r\},\qquad \overline{B}(x,r)=\{y:\|y-x\|\le r\}.

\] They provide a concrete basis for the topology induced by the norm.

3.2 Norm topology

The collection of all open balls forms a base for a topology. Convergence and continuity can therefore be expressed using topological language, yet computations can still rely on the explicit form \(\|x-y\|\).

3.3 Neighborhoods and local properties

A neighborhood of \(x\) is any set containing an open ball around \(x\). Local properties—such as being open or closed at a point—are formulated through whether appropriate balls lie inside or intersect the set.

3.4 Convexity and star-shaped sets

A subset \(C\subset V\) is convex if for any \(x,y\in C\) and \(t\in[0,1]\), the combination \((1-t)x+ty\) is in \(C\). Convexity interacts well with norms because balls are convex in many common norms, and convex sets support geometric arguments about approximations and projections.

A set is star-shaped with center \(x_0\) if it contains every segment starting at \(x_0\) and reaching any point of the set.

3.5 Compactness in normed spaces

In general metric spaces, compactness is equivalent to every sequence having a convergent subsequence. In normed spaces one often uses criteria tied to total boundedness.

3.5.1 Total boundedness and related characterizations

A set \(A\) is totally bounded if for every \(\varepsilon>0\) it can be covered by finitely many balls of radius \(\varepsilon\). In metric spaces, compactness is equivalent to complete and totally bounded behavior. This viewpoint is particularly useful in analysis and approximation theory.

3.6 Distance to a set

For a point \(x\) and subset \(A\), the distance from \(x\) to \(A\) is \[

\operatorname{dist}(x,A)=\inf\{\|x-a\|:a\in A\}.

\] This quantity underlies approximation: it measures the best possible error when approximating \(x\) by elements of \(A\).

4 Linear operators on normed spaces

Linear operators preserve vector addition and scalar multiplication. When they interact nicely with the norms, they provide a controlled notion of “linear behavior” between normed spaces.

4.1 Bounded linear operators

A linear operator \(T:V\to W\) is bounded if it does not stretch vectors without control. Formally, there exists \(M\ge 0\) such that \[

\|Tx\|\le M\|x\| \quad \forall x\in V.

\] Boundedness ensures that \(T\) is continuous.

4.2 Operator norm

The operator norm of a bounded linear operator \(T\) is \[

\|T\|=\sup_{\|x\|\le 1}\|Tx\|.

\]

Equivalently, \(\|T\|\) is the smallest constant \(M\) satisfying \(\|Tx\|\le M\|x\|\).

4.2.1 Basic operator norm estimates

Operator norms satisfy fundamental inequalities:

- \(\|Tx\|\le \|T\|\|x\|\).
- For composition, \(\|S\circ T\|\le \|S\|\|T\|\).
- For sums, \(\|S+T\|\le \|S\|+\|T\|\).

These estimates enable error tracking and stability analysis.

4.3 Continuity vs. boundedness for linear maps

For linear maps between normed spaces, continuity and boundedness coincide: a linear operator is continuous if and only if it is bounded. This equivalence is a central simplification compared to arbitrary (nonlinear) maps.

4.4 Algebra of bounded operators

Bounded linear operators form an algebra under addition and composition. With the operator norm, the space of bounded operators becomes a normed vector space, and in many settings it supports stronger completeness properties.

4.5 Isometries and contractions

An isometry is a map \(T\) preserving norms: \[

\|Tx\|=\|x\|.

\]

A contraction satisfies \(\|Tx\|\le \|x\|\) for all \(x\). These maps are important because they control distances and, in iterative contexts, can enforce convergence.

5 Subspaces and quotient spaces

Subspaces inherit structure, while quotient spaces compress a normed space by identifying vectors that differ by an element of a chosen subspace.

5.1 Closed subspaces and their role

If \(Y\subset V\) is a closed subspace, then sequences in \(Y\) converging in \(V\) converge to points still in \(Y\). This closure property ensures that \(Y\) itself is a normed space under the restricted norm.

5.2 Quotient norm construction

Given a closed subspace \(Y\subset V\), the quotient space \(V/Y\) consists of cosets \(x+Y\). The quotient norm is defined by \[

\|x+Y\|=\inf_{y\in Y}\|x-y\|.

\] This measures the “distance” from \(x\) to the subspace \(Y\).

5.2.1 The quotient map and its properties

The quotient map \(\pi:V\to V/Y\) given by \(\pi(x)=x+Y\) is linear and norm-decreasing in the sense that \[

\|\pi(x)\|=\|x+Y\| \le \|x\|.

\] It is designed so that norm computations in the quotient reflect best approximations by elements of \(Y\).

5.3 Isomorphisms and norm preservation concepts

A linear bijection \(T:V\to W\) that preserves norms is an isometry. More generally, norm-preserving behavior may be expressed through equivalence or through inequalities bounding \(\|Tx\|\) above and below by multiples of \(\|x\|\). Such properties clarify when two normed spaces are effectively similar from the standpoint of convergence and continuity.

6 Completeness and special cases (Banach spaces)

Completeness addresses whether Cauchy sequences converge within the space. This turns normed spaces into stable analytical environments.

6.1 Definition of completeness

A normed space \(V\) is complete if every Cauchy sequence \((x_n)\) in \(V\) converges to some \(x\in V\).

6.2 Banach spaces as normed spaces

A Banach space is a complete normed vector space. Completeness enables many analytic arguments, such as extending limit processes and applying fixed-point principles.

6.3 Examples and typical constructions

Many standard function spaces and sequence spaces are Banach spaces under familiar norms. Constructions such as taking closures of subspaces and completing spaces via limit points naturally produce Banach spaces.

6.4 Completion of a normed space

If a normed space is not complete, it can be embedded into a larger complete space that reflects its Cauchy behavior.

6.4.1 Extension of sequences and limits to the completion

In the completion \(\widehat{V}\), Cauchy sequences from \(V\) acquire limits. Elements of \(\widehat{V}\) can be modeled as equivalence classes of Cauchy sequences, where two sequences are identified if their mutual distance tends to zero.

7 Dual spaces and functionals

Dual spaces collect continuous linear functionals. They are crucial for understanding norms, separation properties, and topologies associated with linear structure.

7.1 Continuous linear functionals

A continuous linear functional on \(V\) is a linear map \(f:V\to \mathbb{F}\) (with \(\mathbb{F}=\mathbb{R}\) or \(\mathbb{C}\)) that is continuous in the norm topology. Continuity is equivalent to boundedness for linear maps, so such functionals satisfy \(	f(x)	\le C\|x\|\).

7.2 The dual norm and dual space notation

The dual space \(V^*\) is the set of all continuous linear functionals on \(V\). It becomes a normed space with the dual norm \[

\|f\|=\sup_{\|x\|\le 1}	f(x)	.

\] This norm measures the largest possible value of \(f\) on unit vectors.

7.3 Separating points: Hahn–Banach (conceptual role)

A key principle in duality is that continuous linear functionals can separate points from sets in many situations. The Hahn–Banach theorem provides a mechanism to extend functionals while preserving bounds, enabling the construction of dual objects tailored to geometric features of the space.

7.4 Weak* topology (high-level perspective)

When studying dual spaces, one often considers the weak* topology, in which convergence is defined by testing against vectors in the original space. Informally, a sequence (or net) in the dual converges in the weak* sense if it converges pointwise on every vector \(x\in V\).

8 Normed space constructions in practice

Normed spaces commonly appear through standard operations: products, sums, and function spaces. These constructions preserve enough structure for analysis while keeping computational rules explicit.

8.1 Product spaces and induced norms

Given normed spaces \(V\) and \(W\), a product space \(V\times W\) can be endowed with norms such as \[

\|(v,w)\|=\|v\|+\|w\|\quad \text{or}\quad \|(v,w)\|=\max\{\|v\|,\|w\|\}.

\] These choices are typically equivalent and thus lead to the same convergence behavior.

8.2 Direct sums and their norms

A direct sum combines spaces so that vectors are tuples with components drawn from each summand. Norms on direct sums can be defined using component norms, frequently via sums or maxima, with resulting equivalence classes depending on the chosen formula.

8.3 Function spaces (overview of typical norms)

Function spaces become normed spaces when a norm measures the size of a function via its values or its oscillation. Typical norms distinguish between pointwise magnitude, uniform behavior, and integrability.

8.3.1 Supremum norm and related examples

The supremum norm (or uniform norm) on bounded functions is \[

\|f\|_\infty=\sup_x	f(x)	.

\] It controls the largest deviation over the domain and is especially suited to uniform approximation and continuity questions.

8.4 Sequence spaces (overview of common norms)

Sequence spaces assign norms to infinite vectors \((x_n)\). The \(\ell^p\) norms for \(1\le p<\infty\), \[

\|(x_n)\|_p=\left(\sum_{n=1}^\infty	x_n	^p\right)^{1/p},

\] and the \(\ell^\infty\) norm, \[

\|(x_n)\|_\infty=\sup_n	x_n	,

\] provide standard settings for studying convergence, summability, and operator behavior.

9 Relationships with inner product spaces

Inner products generate norms and impose additional geometry. Not all norms come from inner products, but when they do, stronger structure becomes available.

9.1 When a norm arises from an inner product

If a vector space \(V\) has an inner product \(\langle\cdot,\cdot\rangle\), it induces a norm via \[

\|x\|=\sqrt{\langle x,x\rangle}.

\] Such norms satisfy special identities reflecting the underlying inner product geometry.

9.2 The parallelogram law (characterization)

A classical characterization states that a norm arises from an inner product if and only if it satisfies the parallelogram law: \[

\|x+y\|^2+\|x-y\|^2=2\|x\|^2+2\|y\|^2.

\] This identity restricts how distances behave with respect to vector addition.

9.3 Hilbert spaces as complete inner product spaces

A Hilbert space is an inner product space that is complete with respect to the induced norm. Completeness plus inner product geometry yield tools such as orthogonality concepts and projection methods.

10 Applications and typical analytical tools

Normed spaces support many practical techniques because they translate “error” and “change” into inequalities involving distances.

10.1 Fixed-point iteration and contractions

Fixed-point methods often rely on contraction mappings. If a self-map \(T\) satisfies \[

\|T(x)-T(y)\|\le q\|x-y\|\quad (0\le q<1),

\] then iterating \(x_{n+1}=T(x_n)\) converges to a unique fixed point in a complete space. The norm quantifies both existence and convergence speed.

10.2 Stability via norm estimates

When analyzing models depending on parameters or inputs, one seeks to show that small perturbations cause proportionally small changes in outcomes. Norm estimates provide exactly this: they bound the difference between results using inequalities like \(\|T(x)-T(y)\|\le \|T\|\|x-y\|\) for linear operators.

10.3 Error bounds and approximation settings

Approximation theory frequently uses distances to sets, best-approximation errors, and continuity properties. If approximations live in a subspace, quotient norms and distance-to-subspace formulations help describe what can be achieved and at what accuracy.

10.4 Convergence of iterative methods (norm-based viewpoint)

Numerical or analytical iterative procedures are commonly analyzed by tracking how successive approximations approach each other and, ideally, the limiting target. Normed spaces provide a universal language for this tracking through convergence of sequences, Cauchy criteria, and operator norms.