1 Definition and basic properties

An invertible matrix is a square matrix that has a matrix inverse. If a matrix \(A\) is invertible, then there exists another square matrix \(A^{-1}\) such that \[ AA^{-1} = A^{-1}A = I, \] where \(I\) is the identity matrix. Invertibility means that the matrix can be “undone” by another matrix, which makes it a fundamental concept in linear algebra.

Invertible matrices arise naturally in the study of linear systems, linear transformations, and coordinate changes. They are also known as nonsingular matrices. Many important properties of a matrix can be tested by checking whether it is invertible.

1.1 Matrix inverse

The inverse of a matrix is the matrix that reverses its effect under multiplication. If \(A\) sends a vector \(x\) to \(Ax\), then \(A^{-1}\) sends \(Ax\) back to \(x\). Not every matrix has an inverse, and when an inverse exists, it is unique.

1.2 Square matrices

Only square matrices can be invertible. A matrix with different numbers of rows and columns cannot have a two-sided inverse, because multiplication would not produce the identity matrix in both orders. The square shape is therefore necessary for invertibility.

1.3 Identity matrix

The identity matrix acts as the multiplicative neutral element for matrices of a fixed size. Multiplying by the identity leaves a matrix unchanged. It plays the same role for matrices that the number 1 plays for ordinary multiplication.

1.4 Nonsingular matrices

A nonsingular matrix is another name for an invertible matrix. The opposite term is singular matrix, meaning a square matrix that does not have an inverse. Nonsingularity is often used in discussions of determinants and linear systems.

2 Equivalent characterizations

Invertibility can be recognized in several equivalent ways. These criteria connect algebraic, geometric, and computational viewpoints. In practice, one often uses whichever test is easiest in a given setting.

2.1 Determinant criterion

A square matrix is invertible exactly when its determinant is nonzero. If the determinant equals zero, the matrix is singular and has no inverse. This criterion is especially useful for small matrices and theoretical arguments.

2.2 Rank criterion

A square matrix is invertible exactly when it has full rank, meaning its rank equals its size. Full rank indicates that all rows and columns contain independent information. If the rank is smaller than the dimension, the matrix collapses some directions and cannot be reversed.

2.3 Linear independence of rows and columns

A square matrix is invertible when its rows are linearly independent, and likewise when its columns are linearly independent. If one row or column can be written as a combination of the others, the matrix loses information and fails to be invertible. The independence of rows and columns are equivalent conditions for square matrices.

2.4 Existence of unique solutions

A square matrix is invertible exactly when the system \(Ax=b\) has a unique solution for every vector \(b\). This property shows that the matrix defines a one-to-one correspondence between input vectors and output vectors. When invertibility fails, some systems have no solutions or infinitely many.

2.5 Row reduction criterion

A square matrix is invertible exactly when it can be row-reduced to the identity matrix. During Gaussian elimination, this happens when every column contains a pivot. If the reduced form has a zero row, then the matrix is singular.

3 Algebraic properties

Invertible matrices behave well under the standard operations of matrix algebra. These rules make them easy to combine and manipulate in calculations. They also show that invertible matrices form a stable class under many transformations.

3.1 Uniqueness of the inverse

If a matrix has an inverse, that inverse is unique. No matrix can serve as a second distinct two-sided inverse for the same matrix. This follows from the associativity of matrix multiplication and the defining identity relations.

3.2 Inverse of a product

The inverse of a product reverses the order of the factors: \[ (AB)^{-1} = B^{-1}A^{-1}. \] This is a key rule in matrix algebra. The order matters because matrix multiplication is generally not commutative.

3.3 Inverse of a transpose

If a matrix is invertible, then its transpose is also invertible, and \[ (A^T)^{-1} = (A^{-1})^T. \] This relationship is useful when working with symmetric, orthogonal, and other structured matrices.

3.4 Inverse of a scalar multiple

For a nonzero scalar \(c\) and an invertible matrix \(A\), \[ (cA)^{-1} = c^{-1}A^{-1}. \] If the scalar is zero, the product becomes the zero matrix, which is never invertible. Scalar multiplication therefore preserves invertibility only when the scalar itself is nonzero.

3.5 Powers of invertible matrices

If \(A\) is invertible, then any integer power \(A^n\) is also invertible. For positive \(n\), the inverse is \(A^{-n}\). This extends the algebraic behavior of inverse matrices to repeated multiplication.

4 Methods for finding inverses

Several computational techniques can be used to determine a matrix inverse. The best method depends on the size and structure of the matrix. In numerical work, some approaches are preferred because they are more stable or efficient.

4.1 Gauss-Jordan elimination

Gauss-Jordan elimination finds an inverse by augmenting the matrix with the identity and row-reducing the result. If the left side can be reduced to the identity, the right side becomes the inverse. This method is direct and widely taught in introductory linear algebra.

4.2 Adjugate formula

For a square matrix with nonzero determinant, the inverse can be written using the adjugate matrix: \[ A^{-1} = \frac{1}{\det(A)}\operatorname{adj}(A). \] This formula is theoretically important, though it is usually inefficient for large matrices because computing cofactors is costly.

4.3 Decomposition methods

Matrix factorizations can make inversion easier by breaking a matrix into simpler components. Once the factors are known, the inverse can often be built from the inverses of those factors. These methods are especially useful in numerical linear algebra.

4.3.1 LU decomposition

In LU decomposition, a matrix is written as a product of a lower triangular matrix and an upper triangular matrix. If both factors are invertible, then the original matrix is invertible as well. Solving systems through LU factorization often avoids explicitly forming the inverse.

4.3.2 QR decomposition

In QR decomposition, a matrix is expressed as an orthogonal matrix times an upper triangular matrix. This factorization can simplify the process of solving equations and computing inverses indirectly. It is often favored in numerical applications because of its stability.

4.4 Block matrix methods

Matrices with a block structure can sometimes be inverted by using formulas for partitions. These methods exploit the arrangement of submatrices and are useful when one block is simple or already known to be invertible. Block inversion appears frequently in advanced algebra and applied mathematics.

5 Relationships with linear transformations

A matrix represents a linear transformation, so invertibility has a direct geometric meaning. An invertible matrix corresponds to a transformation that can be reversed without loss of information. This links matrix theory with the study of vector spaces.

5.1 One-to-one and onto maps

A linear transformation is invertible exactly when it is both one-to-one and onto. One-to-one means different vectors map to different outputs, while onto means every target vector is reached. For finite-dimensional spaces of the same dimension, these conditions are equivalent.

5.2 Isomorphisms

An invertible linear map is called an isomorphism. Such a map preserves the linear structure of a vector space and shows that two spaces are essentially the same from the standpoint of linear algebra. In matrix terms, an isomorphism is represented by an invertible matrix.

5.3 Change of basis

Invertible matrices describe changes of basis. When coordinates are converted from one basis to another, the transformation is reversible, so the change-of-basis matrix must be invertible. This makes invertible matrices central to coordinate geometry and representation theory.

5.4 Matrix representations of reversible transformations

Any reversible linear transformation is represented by an invertible matrix once bases are chosen for the domain and codomain. The matrix depends on the bases, but invertibility does not. This reflects the intrinsic reversibility of the underlying transformation.

6 Applications

Invertible matrices are used throughout mathematics, science, and engineering. Their ability to reverse linear processes makes them valuable in modeling, computation, and analysis. They also provide a framework for understanding how systems respond to inputs.

6.1 Solving linear systems

A primary use of invertible matrices is solving systems of linear equations. If \(A\) is invertible, then the solution to \(Ax=b\) is given by \(x=A^{-1}b\). This provides a clean theoretical solution, even when practical computations use elimination methods instead of directly computing the inverse.

6.2 Computing determinants

Invertibility and determinants are closely linked. A nonzero determinant signals that a matrix is invertible, while a zero determinant signals singularity. Determinants also reflect how a matrix scales volume and whether it preserves orientation.

6.3 Geometry and coordinate transformations

In geometry, invertible matrices describe transformations such as rotations, reflections, scalings, and shears when they preserve enough information to be reversed. They are used to translate between coordinate systems and to analyze the effect of transformations on shapes and distances.

6.4 Differential equations

Systems of linear differential equations often involve invertible matrices, especially when solving for state variables or diagonalizing coefficient matrices. Matrix inverses may appear in formulas for fundamental solutions and in transformations that simplify the system. They help convert coupled equations into more manageable forms.

7 Special cases and examples

Certain classes of matrices have invertibility criteria that are especially simple. These cases illustrate the general theory in concrete form. They also show how matrix structure affects whether an inverse exists.

7.1 2×2 invertible matrices

A \(2\times 2\) matrix \[ \begin{pmatrix} a & b \\ c & d \end{pmatrix} \] is invertible exactly when \(ad-bc \neq 0\). In that case, its inverse is \[ \frac{1}{ad-bc} \begin{pmatrix} d & -b \\ -c & a \end{pmatrix}. \] This is one of the simplest and most familiar inverse formulas.

7.2 3×3 invertible matrices

A \(3\times 3\) matrix is invertible when its determinant is nonzero and its rows or columns are linearly independent. In practice, the inverse can be found by elimination, cofactors, or decomposition methods. The computations are more involved than in the \(2\times 2\) case, but the same principles apply.

7.3 Diagonal matrices

A diagonal matrix is invertible exactly when all diagonal entries are nonzero. Its inverse is also diagonal, with each diagonal entry replaced by its reciprocal. This makes diagonal matrices among the easiest matrices to invert.

7.4 Triangular matrices

A triangular matrix is invertible exactly when all entries on its main diagonal are nonzero. The inverse of an invertible triangular matrix is again triangular of the same type. Because of their structure, triangular matrices are often efficient to work with in computation.

7.5 Orthogonal matrices

An orthogonal matrix has the property that its transpose equals its inverse. Such matrices are always invertible, and their inverse is especially simple to compute. They arise in rotations, reflections, and other length-preserving transformations.

Invertible matrices are connected to several broader ideas in algebra and numerical analysis. These related notions help describe when matrices fail to be invertible and how near-invertibility affects computation. They also extend the concept of reversal to settings where exact inverses do not exist.

8.1 Singular matrices

A singular matrix is a square matrix that is not invertible. Singular matrices have determinant zero and reduced rank, and they typically lead to systems with no solution or multiple solutions. They represent the failure of a linear transformation to be reversible.

8.2 General linear group

The general linear group is the set of all invertible \(n\times n\) matrices over a given field, with matrix multiplication as the operation. It forms a group because inverses exist, multiplication is associative, and the identity matrix serves as the neutral element. This group is a central object in linear algebra and abstract algebra.

8.3 Condition number

The condition number measures how sensitive a matrix problem is to small changes in the input. Matrices can be invertible yet still be close to singular, making numerical inversion unstable. A large condition number often indicates that computing an inverse or solving a system may be unreliable in floating-point arithmetic.

8.4 Pseudoinverse

The pseudoinverse generalizes the idea of an inverse to matrices that are not invertible or not square. It is used to obtain least-squares solutions and to handle rank-deficient systems. When an ordinary inverse does not exist, the pseudoinverse provides a useful substitute.