This sheet introduces vector spaces and linear transformations—the fundamental structures underlying matrix theory and many parts of advanced algebra. Vector spaces provide the language for discussing linear equations, eigenvalues, and canonical forms. Linear transformations are the structure-preserving maps between vector spaces.

Commentary. A vector space is an abstraction of Euclidean space Rn\mathbb{R}^n. The key idea is that we can add vectors and multiply them by scalars (elements of a field). Linear transformations preserve these operations. This framework is essential for understanding matrices, linear systems, and the structural results (rational and Jordan canonical forms) that follow.

Prerequisites. This sheet uses the following results from earlier sheets:

  • Fields (Sheet 7, Definition 7.42)
  • Rings, fields as commutative rings with unity (Sheet 7)
  • Groups and abelian groups (Sheet 1)

Fields and Scalars#

Definition 9.5.1. A field is a commutative ring with unity 1≠01 \neq 0 in which every nonzero element has a multiplicative inverse.

Remark. We will use FF to denote an arbitrary field. The elements of FF are called scalars.

Example 9.5.2. Q\mathbb{Q}, R\mathbb{R}, C\mathbb{C}, and Zp\mathbb{Z}_p (for prime pp) are fields.


Vector Spaces#

Definition 9.5.3. Let FF be a field. A vector space over FF (or FF-vector space) is a set VV together with two operations:

  • Vector addition: +:V×V→V+ : V \times V \to V, denoted v+wv + w
  • Scalar multiplication: ⋅:F×V→V\cdot : F \times V \to V, denoted ava v

such that for all u,v,w∈Vu, v, w \in V and a,b∈Fa, b \in F:

  1. (V,+)(V, +) is an abelian group.
  2. a(v+w)=av+awa(v + w) = av + aw (distributivity over vector addition)
  3. (a+b)v=av+bv(a + b)v = av + bv (distributivity over scalar addition)
  4. a(bv)=(ab)va(bv) = (ab)v (associativity of scalar multiplication)
  5. 1v=v1v = v (identity for scalar multiplication)

The elements of VV are called vectors. The additive identity of VV is denoted 0\mathbf{0} (or simply 00).

Example 9.5.4. Rn={(x1,…,xn)∣xi∈R}\mathbb{R}^n = \{(x_1, \ldots, x_n) \mid x_i \in \mathbb{R}\} is a vector space over R\mathbb{R} with component-wise addition and scalar multiplication.

Example 9.5.5. FnF^n is a vector space over FF for any field FF.

Example 9.5.6. The set of all polynomials F[x]F[x] is a vector space over FF.

Example 9.5.7. The set of all m×nm \times n matrices over FF, denoted Mm×n(F)M_{m \times n}(F), is a vector space over FF.

Exercise 9.5.8. Verify that R2\mathbb{R}^2 satisfies all vector space axioms.

Exercise 9.5.9. Show that V={0}V = \{0\} is a vector space over any field FF.


Subspaces#

Definition 9.5.10. A subspace of a vector space VV over FF is a nonempty subset W⊆VW \subseteq V that is itself a vector space over FF with the induced operations.

Theorem 9.5.11 (Subspace Criterion). A nonempty subset W⊆VW \subseteq V is a subspace iff for all w1,w2∈Ww_1, w_2 \in W and a∈Fa \in F:w1+w2∈Wandaw1∈W.w_1 + w_2 \in W \quad \text{and} \quad a w_1 \in W.

Example 9.5.12. The set {(x,y)∈R2∣y=2x}\{(x, y) \in \mathbb{R}^2 \mid y = 2x\} is a subspace of R2\mathbb{R}^2.

Example 9.5.13. The set of polynomials of degree ≤n\leq n is a subspace of F[x]F[x].

Exercise 9.5.14. Prove Theorem 9.5.11.

Exercise 9.5.15. Show that the intersection of any collection of subspaces of VV is a subspace.


Linear Combinations, Span, and Linear Independence#

Definition 9.5.16. Let v1,…,vn∈Vv_1, \ldots, v_n \in V. A linear combination of these vectors is an expression:a1v1+a2v2+⋯+anvn,ai∈F.a_1 v_1 + a_2 v_2 + \cdots + a_n v_n, \quad a_i \in F.

Definition 9.5.17. The span of a set S⊆VS \subseteq V is the set of all finite linear combinations of vectors in SS:span⁡(S)={∑i=1kaivi∣vi∈S,ai∈F}.\operatorname{span}(S) = \left\{ \sum_{i=1}^k a_i v_i \mid v_i \in S, a_i \in F \right\}.If span⁡(S)=V\operatorname{span}(S) = V, we say SS spans VV (or generates VV).

Definition 9.5.18. A set of vectors {v1,…,vn}\{v_1, \ldots, v_n\} is linearly independent if:a1v1+⋯+anvn=0  ⟹  a1=⋯=an=0.a_1 v_1 + \cdots + a_n v_n = 0 \implies a_1 = \cdots = a_n = 0.Otherwise, it is linearly dependent.

Example 9.5.19. In R3\mathbb{R}^3, {(1,0,0),(0,1,0),(0,0,1)}\{(1,0,0), (0,1,0), (0,0,1)\} is linearly independent and spans R3\mathbb{R}^3.

Example 9.5.20. In R2\mathbb{R}^2, {(1,2),(2,4)}\{(1,2), (2,4)\} is linearly dependent (since (2,4)=2(1,2)(2,4) = 2(1,2)).

Exercise 9.5.21. Determine whether the following sets are linearly independent:
(a) {(1,0),(0,1)}\{(1,0), (0,1)\} in R2\mathbb{R}^2
(b) {(1,2),(3,6)}\{(1,2), (3,6)\} in R2\mathbb{R}^2
(c) {1,x,x2}\{1, x, x^2\} in F[x]F[x]

Exercise 9.5.22. Prove that a set containing the zero vector is linearly dependent.

Exercise 9.5.23. Prove that span⁡(S)\operatorname{span}(S) is a subspace of VV.


Bases and Dimension#

Definition 9.5.24. A basis of a vector space VV is a linearly independent set that spans VV.

Theorem 9.5.25 (Existence of Basis). Every vector space has a basis. In particular, every finite-dimensional vector space has a finite basis.

Theorem 9.5.26 (Replacement Theorem). If {v1,…,vn}\{v_1, \ldots, v_n\} is a basis of VV, then any linearly independent set has at most nn elements, and any spanning set has at least nn elements.

Corollary 9.5.27. Any two bases of a finite-dimensional vector space have the same number of elements.

Definition 9.5.28. The dimension of a finite-dimensional vector space VV over FF, denoted dim⁡FV\dim_F V, is the number of elements in any basis.

Example 9.5.29. dim⁡FFn=n\dim_F F^n = n.

Example 9.5.30. dim⁡FF[x]\dim_F F[x] is infinite.

Example 9.5.31. dim⁡RC=2\dim_\mathbb{R} \mathbb{C} = 2 (basis {1,i}\{1, i\}).

Exercise 9.5.32. Find a basis for the subspace {(x,y,z)∈R3∣x+y+z=0}\{(x, y, z) \in \mathbb{R}^3 \mid x + y + z = 0\}.

Exercise 9.5.33. Show that dim⁡FMm×n(F)=mn\dim_F M_{m \times n}(F) = mn.

Exercise 9.5.34. Prove the Replacement Theorem.


Linear Transformations#

Definition 9.5.35. Let V,WV, W be vector spaces over the same field FF. A function T:V→WT: V \to W is a linear transformation (or linear map) if for all u,v∈Vu, v \in V and a∈Fa \in F:T(u+v)=T(u)+T(v),T(av)=aT(v).T(u + v) = T(u) + T(v), \qquad T(a v) = a T(v).

Definition 9.5.36. The kernel (or null space) of TT is:ker⁡(T)={v∈V∣T(v)=0W}.\ker(T) = \{v \in V \mid T(v) = 0_W\}.The image (or range) of TT is:im⁡(T)={T(v)∣v∈V}.\operatorname{im}(T) = \{T(v) \mid v \in V\}.

Theorem 9.5.37. ker⁡(T)\ker(T) is a subspace of VV, and im⁡(T)\operatorname{im}(T) is a subspace of WW.

Theorem 9.5.38 (Dimension Theorem). If VV is finite-dimensional and T:V→WT: V \to W is linear, then:dim⁡FV=dim⁡Fker⁡(T)+dim⁡Fim⁡(T).\dim_F V = \dim_F \ker(T) + \dim_F \operatorname{im}(T).

Definition 9.5.39. A linear transformation T:V→WT: V \to W is:

  • injective if ker⁡(T)={0}\ker(T) = \{0\}
  • surjective if im⁡(T)=W\operatorname{im}(T) = W
  • bijective (or an isomorphism) if both injective and surjective

Example 9.5.40. The map T:R2→R2T: \mathbb{R}^2 \to \mathbb{R}^2 defined by T(x,y)=(y,0)T(x, y) = (y, 0) has ker⁡(T)={(x,0)∣x∈R}\ker(T) = \{(x, 0) \mid x \in \mathbb{R}\} and im⁡(T)={(0,y)∣y∈R}\operatorname{im}(T) = \{(0, y) \mid y \in \mathbb{R}\}.

Exercise 9.5.41. Show that the map T:F[x]→F[x]T: F[x] \to F[x] given by T(f)=f′T(f) = f' (derivative) is linear.

Exercise 9.5.42. Prove Theorem 9.5.37.

Exercise 9.5.43. Prove the Dimension Theorem.


The Matrix of a Linear Transformation#

Definition 9.5.44. Let VV be an nn-dimensional vector space over FF with basis B={v1,…,vn}\mathcal{B} = \{v_1, \ldots, v_n\}. The coordinate vector of v=a1v1+⋯+anvnv = a_1 v_1 + \cdots + a_n v_n is:[v]B=(a1⋮an)∈Fn.[v]_{\mathcal{B}} = \begin{pmatrix} a_1 \\ \vdots \\ a_n \end{pmatrix} \in F^n.

Definition 9.5.45. Let T:V→WT: V \to W be linear, with bases B\mathcal{B} of VV and C\mathcal{C} of WW. The matrix of TT with respect to B\mathcal{B} and C\mathcal{C}, denoted [T]C←B[T]_{\mathcal{C} \leftarrow \mathcal{B}}, is the m×nm \times n matrix whose jj-th column is [T(vj)]C[T(v_j)]_{\mathcal{C}}.

Theorem 9.5.46. For all v∈Vv \in V:[T(v)]C=[T]C←B⋅[v]B.[T(v)]_{\mathcal{C}} = [T]_{\mathcal{C} \leftarrow \mathcal{B}} \cdot [v]_{\mathcal{B}}.

Theorem 9.5.47. The map Φ:End⁡F(V)→Mn(F)\Phi: \operatorname{End}_F(V) \to M_n(F) defined by Φ(T)=[T]B\Phi(T) = [T]_{\mathcal{B}} (with respect to a fixed basis B\mathcal{B}) is a ring isomorphism. In particular, composition of linear transformations corresponds to matrix multiplication.

Example 9.5.48. Let T:R2→R2T: \mathbb{R}^2 \to \mathbb{R}^2 be T(x,y)=(y,0)T(x, y) = (y, 0). With the standard basis e1=(1,0),e2=(0,1)e_1 = (1,0), e_2 = (0,1), we have:[T]=(0100).[T] = \begin{pmatrix} 0 & 1 \\ 0 & 0 \end{pmatrix}.

Exercise 9.5.49. Let T:R3→R3T: \mathbb{R}^3 \to \mathbb{R}^3 be T(x,y,z)=(x+y,y+z,z)T(x, y, z) = (x + y, y + z, z). Find the matrix of TT with respect to the standard basis.

Exercise 9.5.50. Prove Theorem 9.5.46.


Change of Basis and Similarity#

Definition 9.5.51. Let B\mathcal{B} and B′\mathcal{B}' be two bases of VV. The change-of-basis matrix from B\mathcal{B} to B′\mathcal{B}', denoted PB′←BP_{\mathcal{B}' \leftarrow \mathcal{B}}, is the matrix whose columns are [vi]B′[v_i]_{\mathcal{B}'} for each vi∈Bv_i \in \mathcal{B}.

Theorem 9.5.52. For any vector v∈Vv \in V:[v]B′=PB′←B⋅[v]B.[v]_{\mathcal{B}'} = P_{\mathcal{B}' \leftarrow \mathcal{B}} \cdot [v]_{\mathcal{B}}.

Theorem 9.5.53. If T:V→VT: V \to V is a linear transformation, and B,B′\mathcal{B}, \mathcal{B}' are two bases of VV, then:[T]B′=P−1[T]BP,[T]_{\mathcal{B}'} = P^{-1} [T]_{\mathcal{B}} P,where P=PB′←BP = P_{\mathcal{B}' \leftarrow \mathcal{B}}.

Definition 9.5.54. Two matrices A,B∈Mn(F)A, B \in M_n(F) are similar if B=P−1APB = P^{-1}AP for some invertible P∈Mn(F)P \in M_n(F).

Corollary 9.5.55. Similar matrices represent the same linear transformation in different bases.

Exercise 9.5.56. Let T:R2→R2T: \mathbb{R}^2 \to \mathbb{R}^2 be T(x,y)=(y,x)T(x, y) = (y, x). Find the matrix of TT in the standard basis and in the basis {(1,1),(1,−1)}\{(1,1), (1,-1)\}. Show they are similar.

Exercise 9.5.57. Prove Theorem 9.5.53.


Direct Sums#

Definition 9.5.58. Let U1,…,UkU_1, \ldots, U_k be subspaces of VV. Their sum is:U1+⋯+Uk={u1+⋯+uk∣ui∈Ui}.U_1 + \cdots + U_k = \{u_1 + \cdots + u_k \mid u_i \in U_i\}.

Definition 9.5.59. The sum V=U1+⋯+UkV = U_1 + \cdots + U_k is a direct sum, denoted V=U1⊕⋯⊕UkV = U_1 \oplus \cdots \oplus U_k, if every vector v∈Vv \in V can be written uniquely as v=u1+⋯+ukv = u_1 + \cdots + u_k with ui∈Uiu_i \in U_i.

Theorem 9.5.60. V=U1⊕⋯⊕UkV = U_1 \oplus \cdots \oplus U_k iff:

  1. V=U1+⋯+UkV = U_1 + \cdots + U_k, and
  2. Ui∩∑j≠iUj={0}U_i \cap \sum_{j \neq i} U_j = \{0\} for each ii.

Theorem 9.5.61. If V=U1⊕⋯⊕UkV = U_1 \oplus \cdots \oplus U_k, then:dim⁡FV=dim⁡FU1+⋯+dim⁡FUk.\dim_F V = \dim_F U_1 + \cdots + \dim_F U_k.

Example 9.5.62. R3={(x,0,0)}⊕{(0,y,0)}⊕{(0,0,z)}\mathbb{R}^3 = \{(x,0,0)\} \oplus \{(0,y,0)\} \oplus \{(0,0,z)\}.

Exercise 9.5.63. Let U={(x,y,0)∣x,y∈R}U = \{(x,y,0) \mid x, y \in \mathbb{R}\} and W={(0,0,z)∣z∈R}W = \{(0,0,z) \mid z \in \mathbb{R}\}. Show that R3=U⊕W\mathbb{R}^3 = U \oplus W.

Exercise 9.5.64. Prove Theorem 9.5.60.


Invariant Subspaces#

Definition 9.5.65. Let T:V→VT: V \to V be a linear transformation. A subspace W⊆VW \subseteq V is TT-invariant if T(W)⊆WT(W) \subseteq W.

Theorem 9.5.66. If WW is TT-invariant, then the restriction T∣W:W→WT|_W: W \to W is a linear transformation.

Theorem 9.5.67. If V=W1⊕⋯⊕WkV = W_1 \oplus \cdots \oplus W_k is a direct sum of TT-invariant subspaces, then the matrix of TT with respect to a basis adapted to this decomposition is block diagonal.

Example 9.5.68. For T:R2→R2T: \mathbb{R}^2 \to \mathbb{R}^2 with T(x,y)=(2x,3y)T(x,y) = (2x, 3y), the xx-axis and yy-axis are invariant subspaces.

Exercise 9.5.69. Show that ker⁡(T)\ker(T) and im⁡(T)\operatorname{im}(T) are TT-invariant subspaces.

Exercise 9.5.70. Let TT be diagonalizable. Show that each eigenspace is TT-invariant.


Determinants (Review of Properties)#

Definition 9.5.71. Let A∈Mn(F)A \in M_n(F). The determinant of AA, denoted det⁡(A)\det(A), is a scalar in FF characterized by the following properties:

  1. Multilinearity: det⁡\det is linear in each row.
  2. Alternating: If two rows are swapped, det⁡\det changes sign.
  3. Normalization: det⁡(In)=1\det(I_n) = 1.

Theorem 9.5.72. For all A,B∈Mn(F)A, B \in M_n(F):det⁡(AB)=det⁡(A)det⁡(B).\det(AB) = \det(A) \det(B).

Theorem 9.5.73. AA is invertible iff det⁡(A)≠0\det(A) \neq 0.

Theorem 9.5.74. For any A∈Mn(F)A \in M_n(F) and scalar λ∈F\lambda \in F:det⁡(λA)=λndet⁡(A).\det(\lambda A) = \lambda^n \det(A).

Example 9.5.75. For A=(abcd)A = \begin{pmatrix} a & b \\ c & d \end{pmatrix}, det⁡(A)=ad−bc\det(A) = ad - bc.

Exercise 9.5.76. Compute det⁡(1234)\det \begin{pmatrix} 1 & 2 \\ 3 & 4 \end{pmatrix}.

Exercise 9.5.77. Show that det⁡(AT)=det⁡(A)\det(A^T) = \det(A).


Summary#

We have introduced:

  • Vector spaces over a field FF
  • Subspaces
  • Linear combinations, span, linear independence
  • Bases and dimension
  • Linear transformations and the Dimension Theorem
  • Matrices of linear transformations
  • Change of basis and similarity
  • Direct sums
  • Invariant subspaces
  • Determinants (properties)