Eigenvalues and eigenvectors

Eigenvalues and eigenvectors

Spec C10 · AQA A-Level Further Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • An eigenvector v\mathbf{v} (non-zero) of MM with eigenvalue λ\lambda satisfies Mv=λvM\mathbf{v}=\lambda\mathbf{v}.
  • Eigenvalues solve the characteristic equation det⁡(M−λI)=0\det(M-\lambda I)=0.
  • For 2×22\times2: λ2−(trace)λ+det⁡M=0\lambda^2-(\text{trace})\lambda+\det M=0; the eigenvalues sum to the trace and multiply to the determinant.
  • Find eigenvectors by solving (M−λI)v=0(M-\lambda I)\mathbf{v}=\mathbf{0}: the equations are dependent, so choose one value freely.
  • Eigenvectors give invariant lines through the origin; λ\lambda is the stretch factor along each.

What they are

An eigenvector keeps its line when multiplied by MM; the eigenvalue is the factor it is scaled by.

For a square matrix MM, a non-zero vector v\mathbf{v} is an eigenvector with eigenvalue λ\lambda if Mv=λvM\mathbf{v}=\lambda\mathbf{v}.

Multiplying by MM only scales v\mathbf{v}; it does not change its line.

Any non-zero multiple of an eigenvector is also an eigenvector with the same eigenvalue, so an eigenvector is found only up to a scalar multiple. The zero vector is never an eigenvector, because it satisfies the equation for every λ\lambda.

OvMv = 3vuMu
For M = [[2, 1], [1, 2]]: the eigenvector (1, 1) maps to (3, 3), the same line scaled by 3. A non-eigenvector such as (1, 0) maps to (2, 1) and leaves its line.
  • MvM\mathbf{v}λv\lambda\mathbf{v}
  • Eigenvector v\mathbf{v}non-zero, direction kept (or reversed)

v\mathbf{v} is an eigenvector of MM with eigenvalue 3. What is M(2v)M(2\mathbf{v})?

The characteristic equation

A non-zero solution of (M−λI)v=0(M-\lambda I)\mathbf{v}=\mathbf{0} exists only when M−λIM-\lambda I is singular, so det⁡(M−λI)=0\det(M-\lambda I)=0.

Rearranging Mv=λvM\mathbf{v}=\lambda\mathbf{v} gives (M−λI)v=0(M-\lambda I)\mathbf{v}=\mathbf{0}, so det⁡(M−λI)=0\det(M-\lambda I)=0: the characteristic equation.

For a 2×22\times2 matrix it is λ2−(trace)λ+det⁡M=0\lambda^2-(\text{trace})\lambda+\det M=0, where the trace is the sum of the leading diagonal. For a 3×33\times3 matrix it is a cubic.

Use the trace and determinant to check your roots.

−112345−22468xyλ = 1λ = 3det(M − λI)
For M = [[2, 1], [1, 2]] (trace 4, det 3), det(M − λI) = λ² − 4λ + 3. The roots, where the curve crosses zero, are the eigenvalues 1 and 3.
  • Eigenvaluesdet⁡(M−λI)=0\det(M-\lambda I)=0
  • 2×22\times2 shortcutλ2−(trace) λ+det⁡M=0\lambda^2-(\text{trace})\,\lambda+\det M=0

Worked example

Find the eigenvalues of M=(3122)M=\begin{pmatrix} 3 & 1 \\ 2 & 2 \end{pmatrix}.

Find the eigenvalues of (4123)\begin{pmatrix} 4 & 1 \\ 2 & 3 \end{pmatrix}.

Finding eigenvectors

Solve (M−λI)v=0(M-\lambda I)\mathbf{v}=\mathbf{0}; the equations give the same condition, so choose one component and find the rest.

For each eigenvalue λ\lambda, solve (M−λI)v=0(M-\lambda I)\mathbf{v}=\mathbf{0}.

The equations are dependent (they give the same condition) because M−λIM-\lambda I is singular. Choose one free value, such as x=1x=1, and find the rest.

If the two equations disagree, the eigenvalue is wrong.

  1. 1

    Subtract

    form M−λIM-\lambda I

  2. 2

    Solve

    (M−λI)v=0(M-\lambda I)\mathbf{v}=\mathbf{0}

  3. 3

    Dependent equations

    both rows give one condition

  4. 4

    Choose x=1x=1

    read off yy

Finding an eigenvector

Worked example

Find eigenvectors of M=(3122)M=\begin{pmatrix} 3 & 1 \\ 2 & 2 \end{pmatrix} for λ=4\lambda=4 and λ=1\lambda=1.

An eigenvector of (4123)\begin{pmatrix} 4 & 1 \\ 2 & 3 \end{pmatrix} for λ=5\lambda=5 is...

3x3 matrices

The method is the same, but the characteristic equation is a cubic and the eigenvector has three components.

Example: Q=(2−10−12−10−12)Q=\begin{pmatrix} 2 & -1 & 0 \\ -1 & 2 & -1 \\ 0 & -1 & 2 \end{pmatrix}. Expanding det⁡(Q−λI)\det(Q-\lambda I) gives (2−λ)((2−λ)2−2)=0(2-\lambda)\left((2-\lambda)^2-2\right)=0, so λ=2\lambda=2 or λ=2±2\lambda=2\pm\sqrt2.

For λ=2\lambda=2: −y=0-y=0 and −x−z=0-x-z=0, so an eigenvector is (10−1)\begin{pmatrix} 1 \\ 0 \\ -1 \end{pmatrix}.

Once one root is known, factorise the cubic. For an eigenvector set one component to 1, solve two equations, and use the third as a check.

0.511.522.533.54−4−224xy≈ 0.592≈ 3.41det(Q − λI)
The characteristic cubic (2 − λ)((2 − λ)² − 2) crosses zero at the three eigenvalues 2 − √2 ≈ 0.59, 2 and 2 + √2 ≈ 3.41.

2×22\times2 matrix

  • Quadratic characteristic equation
  • At most 2 eigenvalues
  • Eigenvector with 2 components

3×33\times3 matrix

  • Cubic characteristic equation
  • At most 3 eigenvalues
  • Eigenvector with 3 components

What type of equation is the characteristic equation of a 3×33\times3 matrix?

Geometrical meaning

An eigenvector gives an invariant line through the origin, and the eigenvalue is the scale factor along it.

When MM represents a linear transformation:

  • An eigenvector gives the direction of an invariant line through the origin.
  • The eigenvalue λ\lambda is the scale factor along that line: if λ>1\lambda>1 points move away from the origin, if 0<λ<10<\lambda<1 towards it, and if λ<0\lambda<0 they also swap sides.
  • If λ=1\lambda=1 every point on the line is invariant.

A=(1221)A=\begin{pmatrix} 1 & 2 \\ 2 & 1 \end{pmatrix} has eigenvalue 33 along y=xy=x and −1-1 along y=−xy=-x.

−4−3−2−11234−4−3−2−11234xy(1, 1)(3, 3)(1, −1)(−1, 1)y = xy = −x
Invariant lines of A: (1, 1) goes to (3, 3), (1, −1) to (−1, 1)

A matrix has eigenvalue −2-2 with eigenvector (10)\begin{pmatrix} 1 \\ 0 \end{pmatrix}. What happens to points on the xx-axis?

Try an exam question

Find the eigenvalues of M=(5214)M=\begin{pmatrix} 5 & 2 \\ 1 & 4 \end{pmatrix}, and an eigenvector corresponding to the larger eigenvalue.

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Which equation is the characteristic equation of MM?

1 mark, marked against the mark scheme

Revision notes