1 Definition

A linear isomorphism is a linear map between vector spaces that is both injective and surjective. It preserves the algebraic operations of addition and scalar multiplication, so it carries the structure of one vector space to another without loss or distortion. When such a map exists, the two spaces are called isomorphic.

1.1 Linear map

A linear map is a function between vector spaces that respects linear combinations. If \(T:V \to W\) is linear, then for vectors \(u,v \in V\) and scalar \(c\), it satisfies \[ T(u+v)=T(u)+T(v), \qquad T(cు)=cT(u). \] Because of these rules, the image of any combination of vectors depends only on the images of the individual vectors.

1.2 Bijective linear map

A linear isomorphism is, by definition, a linear map that is bijective. Injectivity means distinct vectors in the domain map to distinct vectors in the codomain. Surjectivity means every vector in the codomain is reached by the map. Together, these properties ensure that the map sets up a one-to-one correspondence between the spaces.

1.3 Invertibility

A bijective linear map has an inverse function, and that inverse is also linear. This makes the map reversible in an algebraically compatible way. In many contexts, invertibility is the practical feature that identifies an isomorphism.

1.3.1 Inverse linear map

If \(T:V \to W\) is a linear isomorphism, then there exists a function \(T^{-1}:W \to V\) such that \[ T^{-1}(T(v))=v \quad \text{and} \quad T(T^{-1}(w))=w. \] This inverse reverses the action of \(T\) while preserving vector space operations.

1.3.2 Uniqueness of the inverse

The inverse of a linear isomorphism is unique. If two functions both undo the same bijective linear map, they must agree on every vector. This uniqueness follows from the defining equations of inverse functions.

2 Basic properties

Linear isomorphisms preserve the essential structure of vector spaces. They transport linear dependence, spans, subspaces, and linear relations from one space to another. As a result, many properties can be studied in either space without changing the outcome.

2.1 Preservation of vector space structure

Because an isomorphism respects addition and scalar multiplication, it sends subspaces to subspaces and linear combinations to linear combinations. Bases are carried to bases, and linearly independent sets remain linearly independent. In this way, the map preserves the full algebraic organization of the space.

2.2 Kernel and image

The kernel and image of a linear map measure how far it is from being an isomorphism. For an isomorphism, both are as small or as large as possible in the appropriate sense. These two subspaces provide a convenient test for whether a linear map is reversible.

2.2.1 Trivial kernel

A linear map is injective exactly when its kernel contains only the zero vector. For an isomorphism, this is always the case. If a nonzero vector were sent to zero, distinct vectors would collapse to the same output.

2.2.2 Full image

A linear map is surjective exactly when its image equals the entire codomain. For an isomorphism, every target vector must be attained. Thus the map covers the whole codomain without leaving gaps.

2.3 Composition of isomorphisms

The composition of two linear isomorphisms is again a linear isomorphism. Likewise, the inverse of an isomorphism is an isomorphism. These closure properties make isomorphisms behave like reversible transformations that can be chained together.

3 Characterizations

Several equivalent criteria identify linear isomorphisms. In finite dimensions, these criteria are especially simple and can be checked using dimensions or matrices. The most useful characterizations depend on whether the spaces have finite bases.

3.1 Finite-dimensional vector spaces

For finite-dimensional vector spaces, dimension plays a decisive role in determining whether a linear map can be an isomorphism. A linear map between spaces of equal finite dimension is an isomorphism exactly when it is either injective or surjective. This symmetry is a consequence of the rank-nullity theorem.

3.1.1 Equal dimension criterion

If \(V\) and \(W\) are finite-dimensional and \(\dim V=\dim W\), then a linear map \(T:V\to W\) is an isomorphism if and only if it is injective, and also if and only if it is surjective. Equal dimension removes the possibility that one condition holds without the other.

3.1.2 Rank-nullity interpretation

The rank-nullity theorem states that \[ \dim V = \dim(\ker T) + \dim(\operatorname{im} T). \] For an isomorphism, the kernel has dimension zero and the image has full dimension. Thus rank and nullity reach their extreme values, making the map both one-to-one and onto.

3.2 Matrix representation

With chosen bases, a linear map between finite-dimensional vector spaces can be represented by a matrix. Under this representation, isomorphisms correspond to invertible matrices. This connection makes linear isomorphisms accessible through ordinary matrix algebra.

3.2.1 Invertible matrices

A matrix represents a linear isomorphism exactly when it is invertible. Such a matrix has a two-sided inverse matrix that reverses its action on coordinate vectors. Invertible matrices are also called nonsingular matrices.

3.2.2 Determinant criterion

For square matrices over a field, invertibility is equivalent to having nonzero determinant. A zero determinant indicates that the transformation collapses dimension in some direction and cannot be reversed. A nonzero determinant guarantees that the associated linear map is an isomorphism.

4 Relation to bases

Bases provide a concrete way to describe linear isomorphisms. By examining what happens to basis vectors, one can determine the entire map. Changes of basis are themselves examples of isomorphisms and are central to coordinate computations.

4.1 Action on basis vectors

A linear map is completely determined by its values on a basis. If the images of basis vectors form a basis of the codomain, then the map is an isomorphism. This principle makes basis data enough to reconstruct the whole transformation.

4.2 Change of basis

A change of basis replaces one coordinate description of a vector space with another. The underlying vectors do not change, but their coordinates do. Such transformations are linear isomorphisms between coordinate representations of the same space.

4.2.1 Transition matrices

Transition matrices convert coordinates from one basis to another. They are invertible because the change can be reversed. These matrices encode the same vector with respect to different coordinate systems.

4.2.2 Coordinate isomorphisms

A coordinate isomorphism identifies a vector space with a coordinate space such as \(F^n\) after a basis has been chosen. This identification turns abstract vectors into tuples of scalars. It is one of the main reasons finite-dimensional spaces can be studied through matrices.

5 Examples

Examples help distinguish genuine isomorphisms from maps that merely resemble them. The simplest cases are identity and scaling maps, while more elaborate examples arise from standard coordinate identifications. Non-examples show what fails when a map loses injectivity or surjectivity.

5.1 Identity map

The identity map sends each vector to itself. It is always a linear isomorphism, since it is both injective and surjective. It serves as the most basic example of a reversible linear transformation.

5.2 Scaling maps

Multiplication by a nonzero scalar is a linear isomorphism on any vector space. The inverse is multiplication by the reciprocal scalar. By contrast, scaling by zero is not an isomorphism because it sends every vector to zero.

5.3 Standard isomorphisms between coordinate spaces

The space \(F^n\) is linearly isomorphic to any \(n\)-dimensional vector space after a basis is chosen. The map sending a vector to its coordinate tuple is a linear isomorphism. This standard identification underlies much of elementary linear algebra.

5.4 Non-examples

A map that projects vectors onto a lower-dimensional subspace is not an isomorphism because it is not injective. Likewise, an inclusion map into a larger space is not surjective. Any linear map with a nontrivial kernel or incomplete image fails to be reversible.

6 Applications

Linear isomorphisms are used to classify spaces, simplify calculations, and translate problems into more convenient forms. Their role is foundational in linear algebra because they allow one to replace a complicated setting with an equivalent simpler one. Many practical computations depend on this flexibility.

6.1 Classification of vector spaces

Finite-dimensional vector spaces are classified up to isomorphism by their dimension. Two spaces with the same dimension are isomorphic, and spaces with different dimensions are not. This gives a complete structural description in the finite-dimensional case.

6.2 Simplifying linear transformations

By choosing suitable bases, a linear transformation can often be represented in a simpler matrix form. Isomorphisms help transfer the problem to a coordinate system where calculations are easier. This is especially useful for understanding eigenvalues, canonical forms, and diagonalization.

6.3 Solving linear systems

Linear systems can be rewritten using matrix language, where invertible transformations preserve the solution set structure. If a coefficient matrix is invertible, the system has a unique solution. Isomorphisms therefore provide the theoretical basis for solving systems by algebraic manipulation.

6.4 Geometry and coordinate changes

In geometry, linear isomorphisms describe transformations that preserve the linear structure of space. They appear in rotations, reflections, shears, and coordinate conversions, though not every geometric transformation is an isomorphism. When the map is invertible, it gives a new description of the same geometry in different coordinates.

Linear isomorphisms are connected to several broader notions in algebra. Some of these describe special cases, while others generalize the idea of equivalence between algebraic objects. Together, they place linear isomorphisms within the wider theory of structure-preserving maps.

7.1 Linear automorphism

A linear automorphism is an isomorphism from a vector space to itself. It is a reversible linear transformation of one space rather than a correspondence between two different spaces. The set of all such maps forms a group under composition.

7.2 Vector space isomorphism

A vector space isomorphism is the same concept as a linear isomorphism. The term emphasizes the equivalence relation between spaces rather than the map itself. If two vector spaces are isomorphic, they share the same linear structure.

7.3 Isomorphism theorem

Isomorphism theorems describe relationships among quotient spaces, kernels, images, and homomorphisms. They are not limited to vector spaces, but they play an important role in linear algebra. These theorems explain how a map can factor through simpler structures.

7.4 Endomorphism and automorphism groups

An endomorphism is a linear map from a vector space to itself, whether or not it is invertible. The invertible endomorphisms form the automorphism group. This group captures all reversible linear symmetries of the space.

</INTERNAL_LINK_CANDIDATES> Linear map (a function preserving addition and scalar multiplication) Injective map (a function with distinct inputs giving distinct outputs) Surjective map (a function whose outputs cover the entire codomain) Bijective map (a function that is both injective and surjective) Kernel (the set of vectors sent to zero by a linear map) Image (the set of vectors attained by a linear map) Rank-nullity theorem (the dimension formula relating kernel and image) Invertible matrix (a square matrix with a matrix inverse) Determinant (a scalar criterion for matrix invertibility) Basis (a linearly independent spanning set) Transition matrix (a matrix converting coordinates between bases) Coordinate space (a standard vector space of tuples) Identity map (the map sending each vector to itself) Scalar multiplication (multiplying vectors by field elements) Vector space dimension (the number of vectors in a basis) Linear system (a set of simultaneous linear equations) Eigenvalue (a scalar associated with a linear transformation) Diagonalization (representing a transformation in diagonal form) Linear automorphism (an invertible linear map from a space to itself) Quotient space (a vector space formed by modding out a subspace)