A = (2 1; 1 2), D = diag(3, 1)
A³ entry (1,1) = (3³ + 1)/2
14.
If a matrix has enough independent eigenvectors it can be written , where the columns of are eigenvectors and is the diagonal matrix of eigenvalues. Then .
This page is easier if you are comfortable with:
Diagonal matrices are easy to raise to powers: just raise each diagonal entry to the power. So once you have and , the power of the original matrix follows. This is the main use of diagonalisation.
First find the eigenvalues from the characteristic equation, then the eigenvectors from the eigenvector equation. Build from the eigenvectors and from the eigenvalues in the same order, and find with the 2 × 2 inverse formula.
Not every matrix can be diagonalised. A repeated eigenvalue with too few eigenvectors leads to the Jordan normal form. Symmetric matrices can always be diagonalised with an orthogonal matrix. The related matrix exponential uses the same idea.
Diagonalising a matrix rewrites it in the basis of its eigenvectors, where it becomes a simple diagonal matrix. That makes powers easy.
In Aⁿ = PDⁿP⁻¹ the large eigenvalue 3 dominates quickly, which is why the entries of Aⁿ grow like 3ⁿ.
These examples are worked by hand, step by step, so you can compare your own working.
Diagonalise .
Find for .
Verify the diagonalisation of with , .
A few situations come up again and again. Knowing them saves time and helps you spot answers that cannot be right.
A³ entry (1,1) = (3³ + 1)/2
14.
A³ entry (1,2) = (3³ − 1)/2
13.
A⁵ entry (1,1)
122.
det A = 3 × 1
3.
A = PDP⁻¹ with eigenvectors in P and eigenvalues in D; then Aⁿ = PDⁿP⁻¹, so only the diagonal is raised to a power.
A = (2 1; 1 2): (1,1) entry.
(27 − 1)/2.
(243 + 1)/2.
Product of eigenvalues.
wrong ✗
PDⁿP⁻¹ ✓
Only D is powered.
always ✗
need n independent eigenvectors ✓
Defective matrices.
wrong ✗
A = PDP⁻¹ ✓
Order.
vectors ✗
eigenvalues ✓
Roles.
Try each one on paper, then open the answer. Use the solver above to see the full working.
5
40
64
41
4
243
Unlimited fresh questions with instant marking. Choose a level, work it out on paper, then type your answer.
Solve: …
Writing A = PDP⁻¹ with D diagonal.
P has eigenvectors as columns; D has the eigenvalues.
PDⁿP⁻¹.
No, only with enough independent eigenvectors.
eigenvector equation.
matrix exponential.
More in Vectors and matrices and the formula sheets.