A real symmetric matrix can always be written : is orthogonal with unit eigenvectors as columns, and is diagonal with the eigenvalues. Equivalently, .
| Step | On | |
|---|---|---|
| 1 | Check symmetry. Only for a symmetric matrix are real eigenvalues and an orthonormal set of eigenvectors guaranteed. | ✓ |
| 2 | Find the eigenvalues. Solve the characteristic equation, as in the eigenvalues method. | , , |
| 3 | Find an eigenvector for each eigenvalue. Solve as in the eigenvectors method. For a repeated eigenvalue, orthogonalize its eigenspace with Gram-Schmidt. | : ; : |
| 4 | Normalize and assemble. Divide each eigenvector by its length. Column of must match diagonal entry of . | , |
| 5 | Check with the outer-product form. Add the terms (spectral decomposition); the sum must be . | ✓ |