1.15 Eigenvalues and eigenvectors

1.15 Eigenvalues and eigenvectors

IB Maths: Applications and Interpretation HL: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • An eigenvector v≠0\mathbf v\neq\mathbf 0 of MM keeps its direction: Mv=λvM\mathbf v=\lambda\mathbf v, where λ\lambda is the eigenvalue.
  • Eigenvalues solve det⁡(M−λI)=0\det(M-\lambda I)=0; for 2×22\times2 this is λ2−(trace)λ+det⁡M=0\lambda^{2}-(\text{trace})\lambda+\det M=0.
  • Find each eigenvector from (M−λI)v=0(M-\lambda I)\mathbf v=\mathbf 0.
  • With distinct real eigenvalues, M=PDP−1M=PDP^{-1} and Mn=PDnP−1M^{n}=PD^{n}P^{-1}.
  • In a model xn+1=Mxn\mathbf x_{n+1}=M\mathbf x_n, the largest ∣λ∣|\lambda| dominates in the long term.
OvMv = 5v
M(1, 1) = (5, 5): the vector is stretched by 5 but keeps its direction, so (1, 1) is an eigenvector with eigenvalue 5.

Eigenvalues and eigenvectors

An eigenvector is a non-zero vector that a matrix only stretches, shrinks or reverses, never turns.

Most vectors change direction when multiplied by a matrix. An eigenvector v\mathbf v of a square matrix MM satisfies

Mv=λv,v≠0M\mathbf v=\lambda\mathbf v,\qquad\mathbf v\neq\mathbf 0

where the scalar λ\lambda is the eigenvalue. Any non-zero multiple of an eigenvector is also an eigenvector.

OvMv
Mv points the same way as v, scaled by the eigenvalue.

Worked example

Show that (11)\begin{pmatrix}1\\1\end{pmatrix} is an eigenvector of M=(4123)M=\begin{pmatrix}4&1\\2&3\end{pmatrix}.

If v\mathbf v is an eigenvector of MM, is 3v3\mathbf v one too?

The characteristic equation

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

Mv=λvM\mathbf v=\lambda\mathbf v means (M−λI)v=0(M-\lambda I)\mathbf v=\mathbf 0. For a non-zero v\mathbf v, M−λIM-\lambda I has no inverse, so

det⁡(M−λI)=0.\det(M-\lambda I)=0.

For a 2×22\times2 matrix the characteristic equation is λ2−(trace)λ+det⁡M=0\lambda^{2}-(\text{trace})\lambda+\det M=0, where the trace is a+da+d. Then put each λ\lambda into (M−λI)v=0(M-\lambda I)\mathbf v=\mathbf 0 to find its eigenvector.

1234567−224681012λdet(M − λI)λ = 2λ = 5(3.5, −2.25)λ² − 7λ + 10
For M = (4 1; 2 3), the characteristic equation λ² − 7λ + 10 = 0 has roots λ = 2 and λ = 5
  • 2×2 eigenvaluesλ2−(trace)λ+det⁡M=0\lambda^{2}-(\text{trace})\lambda+\det M=0

Worked example

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

A 2×22\times2 matrix has trace 5 and determinant 6. What are its eigenvalues?

Diagonalisation

Put the eigenvectors in the columns of PP and the eigenvalues, in the same order, on the diagonal of DD.

If a 2×22\times2 matrix MM has two distinct real eigenvalues λ1,λ2\lambda_1,\lambda_2 with eigenvectors v1,v2\mathbf v_1,\mathbf v_2:

P=(v1v2),D=(λ100λ2),M=PDP−1.P=\begin{pmatrix}\mathbf v_1&\mathbf v_2\end{pmatrix},\quad D=\begin{pmatrix}\lambda_1&0\\0&\lambda_2\end{pmatrix},\quad M=PDP^{-1}.

The columns of PP and the entries of DD must be in the same order. This is diagonalisation.

  1. 1

    Characteristic equation

    Find λ1,λ2\lambda_1,\lambda_2

  2. 2

    Eigenvectors

    One for each λ\lambda

  3. 3

    Build PP and DD

    Same order in both

  4. 4

    Check

    M=PDP−1M=PDP^{-1}

Diagonalising a matrix

Worked example

M=(1230)M=\begin{pmatrix}1&2\\3&0\end{pmatrix}. Write MM as PDP−1PDP^{-1}.

If D=(300−2)D=\begin{pmatrix}3&0\\0&-2\end{pmatrix}, which is the first column of PP?

Powers and populations

Because M=PDP−1M=PDP^{-1}, only the diagonal entries are raised to the power.

The middle factors cancel: M2=PD2P−1M^{2}=PD^{2}P^{-1} and in general

Mn=PDnP−1,Dn=(λ1n00λ2n).M^{n}=PD^{n}P^{-1},\qquad D^{n}=\begin{pmatrix}\lambda_1^{n}&0\\0&\lambda_2^{n}\end{pmatrix}.

For xn+1=Mxn\mathbf x_{n+1}=M\mathbf x_n, write x0=c1v1+c2v2\mathbf x_0=c_1\mathbf v_1+c_2\mathbf v_2 to get xn=c1λ1nv1+c2λ2nv2\mathbf x_n=c_1\lambda_1^{n}\mathbf v_1+c_2\lambda_2^{n}\mathbf v_2. An eigenvalue 11 gives a steady state, ∣λ∣<1|\lambda|<1 decays, and the largest ∣λ∣|\lambda| dominates in the long term.

0.511.522.533.544.55246810nλⁿλ = 2: growsλ = 1: steady stateλ = 0.5: decays
The size of λⁿ: a power with |λ| > 1 grows and dominates, λ = 1 is a steady state, |λ| < 1 decays (n shown as a continuous axis)

λ=1\lambda=1

Effect:
Steady state
Long-term result:
Population settles

∣λ∣<1|\lambda|<1

Effect:
Term decays
Long-term result:
Fades away

∣λ∣>1|\lambda|>1

Effect:
Term grows
Long-term result:
Dominates long term

Worked example

Using M=PDP−1M=PDP^{-1} from the previous example, find M2M^{2}.

What are the eigenvalues of M3M^{3} if MM has eigenvalues 2 and 5?

Try an exam question

The matrix M=(4123)M=\begin{pmatrix}4&1\\2&3\end{pmatrix}. Find the eigenvalues of MM and an eigenvector for each.

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Find the characteristic polynomial of MM.

1 mark, marked against the mark scheme

Revision notes