Which matrices can be diagonalized?

A matrix is diagonalizable if it has a complete set of linearly independent eigenvectors. Every real symmetric matrix is orthogonally diagonalizable.

Why is diagonalization useful?

Diagonalization simplifies matrix powers, linear transformations, differential equations, and many problems in physics, engineering and data science.

Can a matrix with repeated eigenvalues be diagonalized?

Yes. Repeated eigenvalues do not prevent diagonalization, provided there are enough linearly independent eigenvectors.

What is orthogonal diagonalization?

Orthogonal diagonalization is a decomposition A = QDQ transpose, where Q is an orthogonal matrix whose columns are orthonormal eigenvectors. Every real symmetric matrix has this property.

Matrix Diagonalization Simulator - Calculate P, D and P Inverse

Interactive simulator

Diagonalize a Matrix

Enter a real symmetric matrix. The calculator finds an orthonormal eigenvector basis and constructs its diagonal decomposition.

Free to use

Matrix A (one row per line)

Separate entries with commas or spaces. Each row must have the same number of values. Enter a symmetric square matrix with dimensions from 2×2 to 5×5.

Load example

Matrix power (k)

What this simulator calculates

Eigenvalues of the input matrix.
Orthonormal eigenvectors.
The eigenvector matrix P.
The diagonal eigenvalue matrix D.
The inverse of P, which equals P transpose here.
The reconstructed matrix PDP inverse.
Matrix powers using the diagonal decomposition.

\[ A=PDP^{-1} \] \[ P^{-1}=P^T \]

The orthogonal form is valid because the input matrix is real and symmetric.

Numerical results

Diagonalization Results

Diagonalized

Matrix size

3×3

Eigenvalues

Reconstruction error

Orthogonality error

Diagonalization verification

\[ A=PDP^{-1} \]

Matrix decomposition

Eigenvalues and eigenvectors

Index	Eigenvalue	Corresponding eigenvector	Residual

Reconstruction check

Compare the original matrix A with the reconstructed matrix PDP inverse.

Matrix powers

When A = PDP inverse, its integer powers can be computed using A^k = PD^kP inverse.

Numerical method

Understanding the mathematics

What Is Matrix Diagonalization?

Matrix diagonalization is the process of representing a square matrix using a diagonal matrix and a change-of-basis matrix. Instead of working directly with the original matrix, we express its action in a coordinate system formed by its eigenvectors.

\[ \boxed{A=PDP^{-1}} \]

Here, P is an invertible matrix whose columns are linearly independent eigenvectors of A. D is a diagonal matrix containing the corresponding eigenvalues in the same order.

Matrix A

The original square matrix that we want to diagonalize.

Matrix P

The eigenvectors of A arranged as columns.

Matrix D

The eigenvalues arranged along the main diagonal.

The mathematical process

How to Diagonalize a Matrix

01.Find the eigenvalues

Calculate the characteristic polynomial by subtracting λI from A and taking its determinant.

\[ \det(A-\lambda I)=0 \]

The roots of this equation are the eigenvalues of A.

02.Find the eigenvectors

For each eigenvalue, solve the homogeneous system:

\[ (A-\lambda I)v=0 \]

The eigenvectors form the basis used to construct the matrix P.

03.Construct P and D

Place the corresponding eigenvectors in the columns of P and arrange their eigenvalues along the diagonal of D.

\[ P=[v_1\ v_2\ \cdots\ v_n] \] \[ D=\operatorname{diag}(\lambda_1,\lambda_2,\ldots,\lambda_n) \]

04.Verify the decomposition

If P is invertible and the eigenvectors are paired with their corresponding eigenvalues, then:

\[ \boxed{A=PDP^{-1}} \]

Important theorem

When Is a Matrix Diagonalizable?

An n×n matrix is diagonalizable if and only if it has n linearly independent eigenvectors. Having distinct eigenvalues is sufficient, but it is not necessary.

Important: Repeated eigenvalues do not automatically mean a matrix cannot be diagonalized. The important question is whether there are enough linearly independent eigenvectors.

Example: a diagonalizable matrix with repeated eigenvalues

\[ A= \begin{bmatrix} 3&0\\ 0&3 \end{bmatrix} \]

Both eigenvalues are 3, but there are two linearly independent eigenvectors. The matrix is already diagonal and is therefore diagonalizable.

Example: a matrix that is not diagonalizable

\[ A= \begin{bmatrix} 1&1\\ 0&1 \end{bmatrix} \]

This matrix has the repeated eigenvalue 1, but its eigenspace has dimension one. It cannot provide two linearly independent eigenvectors, so it is not diagonalizable.

Why symmetric matrices are special

Every real symmetric matrix can be orthogonally diagonalized. Its eigenvectors can be chosen to be orthonormal, so the inverse of P is simply its transpose:

\[ A=QDQ^T,\qquad Q^{-1}=Q^T \]

Practical applications

Applications of Matrix Diagonalization

1. Computing matrix powers

Matrix powers can become much easier because powers of a diagonal matrix are calculated by raising each diagonal entry to the corresponding power.

\[ A^k=PD^kP^{-1} \]

2. Differential equations

Eigenvalue decompositions help solve systems of linear differential equations and understand their long-term behavior.

\[ x'(t)=Ax(t) \]

3. Physics and engineering

Eigenvalues and eigenvectors describe natural frequencies, vibration modes, stability, and important properties of physical systems.

4. Data science and PCA

Symmetric covariance matrices can be orthogonally diagonalized. Their eigenvectors identify principal directions and their eigenvalues describe the associated variances.

Frequently asked questions

Matrix Diagonalization FAQs

1. What is matrix diagonalization?

Matrix diagonalization is the decomposition of a square matrix into A = PDP inverse. P contains linearly independent eigenvectors, while D contains their corresponding eigenvalues.

2. Can every square matrix be diagonalized?

No. A matrix must have enough linearly independent eigenvectors to form an invertible matrix P. Some matrices do not satisfy this condition.

3. Does a matrix with repeated eigenvalues remain diagonalizable?

Yes, it can. A repeated eigenvalue is not a problem if its eigenspace contains enough independent eigenvectors.

4. What is the difference between P and D?

P contains eigenvectors as columns. D contains the corresponding eigenvalues on its main diagonal, with zero entries elsewhere.

5. Why is the inverse of P equal to its transpose in this simulator?

This calculator works with real symmetric matrices and constructs orthonormal eigenvectors. The resulting eigenvector matrix is orthogonal, which means P inverse equals P transpose.

6. What is the advantage of diagonalization for matrix powers?

Rather than multiplying a matrix by itself repeatedly, we can calculate powers of its diagonal form and then transform back using P and P inverse.

7. What algorithm does this simulator use?

It uses the Jacobi eigenvalue algorithm for real symmetric matrices. The algorithm applies plane rotations to reduce off-diagonal entries until the matrix is approximately diagonal.

8. Can I use this calculator for non-symmetric matrices?

This version is specifically designed for real symmetric matrices. General non-symmetric matrices may have complex eigenvalues or may not be diagonalizable, and require a different numerical approach.

Continue learning

Related Matrix Calculators

Explore more interactive tools for eigenvalues, matrix operations, decompositions and numerical methods.