ConceptComplete

Vector Spaces

A vector space is the central object of study in linear algebra. It abstracts the familiar properties of arrows in the plane to a far more general setting.


Definition

Definition1.1Vector space over a field F

A vector space over a field FF is a set VV together with two operations:

  • Addition: V×VVV \times V \to V, denoted (u,v)u+v(u, v) \mapsto u + v
  • Scalar multiplication: F×VVF \times V \to V, denoted (a,v)av(a, v) \mapsto av

satisfying the following axioms for all u,v,wVu, v, w \in V and a,bFa, b \in F:

  1. u+v=v+uu + v = v + u (commutativity)
  2. (u+v)+w=u+(v+w)(u + v) + w = u + (v + w) (associativity)
  3. There exists 0V0 \in V such that v+0=vv + 0 = v (zero vector)
  4. For each vv, there exists v-v such that v+(v)=0v + (-v) = 0 (additive inverse)
  5. a(bv)=(ab)va(bv) = (ab)v (compatibility)
  6. 1v=v1v = v (identity)
  7. a(u+v)=au+ava(u + v) = au + av (distributivity over vector addition)
  8. (a+b)v=av+bv(a + b)v = av + bv (distributivity over scalar addition)
RemarkTerminology

The elements of VV are called vectors and the elements of FF are called scalars. Common choices for FF are R\mathbb{R} (real vector spaces) and C\mathbb{C} (complex vector spaces). When the field is clear from context, we simply say "VV is a vector space."


Fundamental examples

ExampleThe space R^n

Rn={(a1,a2,,an)aiR}\mathbb{R}^n = \{(a_1, a_2, \ldots, a_n) \mid a_i \in \mathbb{R}\} with componentwise addition and scalar multiplication:

(a1,,an)+(b1,,bn)=(a1+b1,,an+bn),c(a1,,an)=(ca1,,can).(a_1, \ldots, a_n) + (b_1, \ldots, b_n) = (a_1 + b_1, \ldots, a_n + b_n), \quad c(a_1, \ldots, a_n) = (ca_1, \ldots, ca_n).

The zero vector is (0,0,,0)(0, 0, \ldots, 0). This is the prototypical vector space.

ExampleThe space C^n

Cn\mathbb{C}^n is a vector space over C\mathbb{C} with the same operations as Rn\mathbb{R}^n. Note that Cn\mathbb{C}^n can also be viewed as a vector space over R\mathbb{R} (of dimension 2n2n).

ExampleThe space M_{m x n}(F)

The set of all m×nm \times n matrices with entries in FF, with the usual matrix addition and scalar multiplication, forms a vector space over FF. The zero vector is the zero matrix.

ExamplePolynomial spaces P_n(F) and P(F)

Pn(F)={a0+a1x++anxnaiF}P_n(F) = \{a_0 + a_1 x + \cdots + a_n x^n \mid a_i \in F\} is the space of polynomials of degree at most nn. With the usual addition and scalar multiplication of polynomials, Pn(F)P_n(F) is a vector space over FF.

P(F)=n=0Pn(F)P(F) = \bigcup_{n=0}^{\infty} P_n(F) is the space of all polynomials, which is an infinite-dimensional vector space.

ExampleFunction spaces

Let SS be any nonempty set. The set F(S,R)={f:SR}F(S, \mathbb{R}) = \{f : S \to \mathbb{R}\} of all real-valued functions on SS is a vector space with pointwise operations:

(f+g)(s)=f(s)+g(s),(cf)(s)=cf(s).(f + g)(s) = f(s) + g(s), \quad (cf)(s) = c \cdot f(s).

The zero vector is the zero function f(s)=0f(s) = 0 for all ss.

ExampleContinuous functions C[a, b]

C[a,b]={f:[a,b]Rf is continuous}C[a, b] = \{f : [a, b] \to \mathbb{R} \mid f \text{ is continuous}\} is a vector space over R\mathbb{R}. The sum of continuous functions is continuous, and a scalar multiple of a continuous function is continuous. This is a subspace of F([a,b],R)F([a,b], \mathbb{R}).

ExampleThe zero vector space

V={0}V = \{0\} consisting of just the zero vector is a vector space over any field FF. It is the smallest possible vector space, with dimV=0\dim V = 0.

ExampleF as a vector space over F

Any field FF is a vector space over itself, with scalar multiplication being the field multiplication. More generally, if E/FE/F is a field extension, then EE is a vector space over FF. For instance, C\mathbb{C} is a 2-dimensional vector space over R\mathbb{R}, with basis {1,i}\{1, i\}.

ExampleSequence spaces

The set F={(a1,a2,a3,)aiF}F^\infty = \{(a_1, a_2, a_3, \ldots) \mid a_i \in F\} of all infinite sequences is a vector space over FF with componentwise operations. Notable subspaces include:

  • 2\ell^2: sequences with ai2<\sum |a_i|^2 < \infty (a Hilbert space)
  • c0c_0: sequences converging to 00
  • c00c_{00}: sequences with only finitely many nonzero terms
ExampleSolution space of a homogeneous system

The set of all solutions to a homogeneous linear system Ax=0Ax = 0 (where AA is an m×nm \times n matrix) is a vector space over FF: if Au=0Au = 0 and Av=0Av = 0, then A(u+v)=0A(u + v) = 0 and A(cu)=0A(cu) = 0. This is a subspace of FnF^n.

ExampleC as a real vector space

C\mathbb{C} is a vector space over R\mathbb{R} with basis {1,i}\{1, i\}. Every complex number z=a+biz = a + bi is a unique linear combination of 11 and ii with real coefficients. Thus dimRC=2\dim_{\mathbb{R}} \mathbb{C} = 2.

ExampleR as a vector space over Q

R\mathbb{R} is a vector space over Q\mathbb{Q}. However, this vector space is infinite-dimensional (in fact, uncountably so). A basis is called a Hamel basis, and its existence requires the axiom of choice.


Basic properties

Theorem1.1Uniqueness and cancellation

In any vector space VV over FF:

  1. The zero vector 00 is unique.
  2. The additive inverse v-v is unique for each vVv \in V.
  3. 0v=00v = 0 for all vVv \in V (scalar zero gives vector zero).
  4. a0=0a0 = 0 for all aFa \in F.
  5. (1)v=v(-1)v = -v for all vVv \in V.
  6. If av=0av = 0, then a=0a = 0 or v=0v = 0.
Proof

For (3): 0v=(0+0)v=0v+0v0v = (0 + 0)v = 0v + 0v. Adding (0v)-(0v) to both sides gives 0=0v0 = 0v.

For (5): v+(1)v=1v+(1)v=(1+(1))v=0v=0v + (-1)v = 1v + (-1)v = (1 + (-1))v = 0v = 0, so (1)v(-1)v is the additive inverse of vv.

For (6): Suppose av=0av = 0 and a0a \neq 0. Then a1a^{-1} exists in FF, and v=1v=(a1a)v=a1(av)=a10=0v = 1v = (a^{-1}a)v = a^{-1}(av) = a^{-1}0 = 0.

RemarkLooking ahead

A vector space is an abelian group (V,+)(V, +) equipped with a compatible scalar multiplication by a field FF. More generally, replacing the field by a ring RR gives the notion of an RR-module. Modules over rings are central in commutative algebra and algebraic geometry.