← Back to Linear Algebra

Matrix inverses

Some square matrices can be undone by an inverse matrix. Inverses solve systems, reverse transformations, and connect to many equivalent conditions.

Definition

A square matrix A is invertible if there is a matrix \(A^{-1}\) with \(AA^{-1} = A^{-1}A = I\). If it exists, the inverse is unique. A matrix with no inverse is called singular. Only square matrices can have two-sided inverses, and many square matrices, such as any with a zero row, are singular.

The 2 x 2 formula

For \(A = \begin{bmatrix} a & b \\ c & d \end{bmatrix}\), \(A^{-1} = \frac{1}{ad - bc}\begin{bmatrix} d & -b \\ -c & a \end{bmatrix}\), provided \(ad - bc \ne 0\). The quantity \(ad - bc\) is the determinant; when it is zero, A is singular. For \(\begin{bmatrix} 2 & 1 \\ 5 & 3 \end{bmatrix}\), the determinant is 1 and the inverse has rows (3, -1) and (-5, 2).

Solving systems with inverses

If A is invertible, \(A\mathbf{x} = \mathbf{b}\) has the unique solution \(\mathbf{x} = A^{-1}\mathbf{b}\) for every \(\mathbf{b}\). This is conceptually neat, but computing \(A^{-1}\) is usually more work and less accurate than elimination, so numerical software solves systems by factorisation rather than explicit inversion.

Properties

If A and B are invertible n x n matrices: \((A^{-1})^{-1} = A\); \((AB)^{-1} = B^{-1}A^{-1}\), with order reversed, like taking off socks and shoes; \((A^T)^{-1} = (A^{-1})^T\); and \((cA)^{-1} = \frac{1}{c}A^{-1}\) for non-zero c. A product of invertible matrices is invertible.

Finding inverses by row reduction

To invert an n x n matrix, row reduce the augmented matrix \([A \mid I]\). If A reduces to I, the right side becomes \(A^{-1}\): \([I \mid A^{-1}]\). If a row of zeros appears on the left, A is singular. This works because the row operations that turn A into I, applied to I, produce exactly the matrix that undoes A.

The invertible matrix theorem

For a square n x n matrix A, the following are equivalent: A is invertible; A row reduces to I; A has n pivots; \(A\mathbf{x} = \mathbf{0}\) has only the trivial solution; the columns of A are linearly independent; the columns span \(\mathbb{R}^n\); \(A\mathbf{x} = \mathbf{b}\) has a solution for every \(\mathbf{b}\); \(A^T\) is invertible; and, as later units show, \(\det A \ne 0\) and 0 is not an eigenvalue.

Interpretation

An invertible matrix represents a transformation that loses no information: distinct inputs give distinct outputs, and every output is reached. A singular matrix collapses space into a lower dimension, such as squashing the plane onto a line, so information is lost and cannot be recovered.

Worked example

Solve \(2x + y = 4\), \(5x + 3y = 11\) using the inverse found above: \(\mathbf{x} = A^{-1}\mathbf{b} = \begin{bmatrix} 3 & -1 \\ -5 & 2 \end{bmatrix}\begin{bmatrix} 4 \\ 11 \end{bmatrix} = \begin{bmatrix} 12 - 11 \\ -20 + 22 \end{bmatrix} = \begin{bmatrix} 1 \\ 2 \end{bmatrix}\). Check: \(2 + 2 = 4\) and \(5 + 6 = 11\).

Key takeaways

A square matrix is invertible if some A^{-1} gives AA^{-1} = A^{-1}A = I; otherwise it is singular. For 2 x 2 matrices, A^{-1} = (1/(ad - bc)) times [[d, -b], [-c, a]] when ad - bc is not zero. Invertible A gives the unique solution x = A^{-1} b, though software prefers elimination. (AB)^{-1} = B^{-1} A^{-1} and (A^T)^{-1} = (A^{-1})^T. Row reducing [A | I] to [I | A^{-1}] finds inverses. The invertible matrix theorem links invertibility, n pivots, independent columns, spanning columns, trivial null solutions, and non-zero determinant. Singular matrices collapse space and lose information.

Flashcard

Invertible matrix

Has A^{-1} with A A^{-1} = I

Non-singular.

Check your understanding

What is the inverse of [[4, 7], [2, 6]]?

← Previous lesson Next lesson →