Linear Algebra · Undergraduate

Complex Eigenvalues and Rotation

Quick answer

A rotation of the plane turns every direction, so it has no real eigenvectors. Its characteristic polynomial has complex roots instead. For a 2 × 2 real matrix, the eigenvalues then come in a conjugate pair a ± bi, and the matrix acts like a rotation by the angle of a + bi combined with a scaling by its modulus r = √(a² + b²). That describes powers at once: Aᵏ rotates by k times the angle and scales by rᵏ.

What you'll learn

  • Find complex eigenvalues from the characteristic polynomial
  • Find a complex eigenvector
  • Interpret a ± bi as a rotation angle and a scale factor
  • Describe powers of a matrix with complex eigenvalues

A matrix that turns every direction

The rotation by 90°90°, R=[0−110]R = \begin{bmatrix} 0 & -1\\ 1 & 0 \end{bmatrix}, turns every nonzero vector to a new direction. No direction stays put, so there is no real eigenvector. The algebra agrees: the characteristic polynomial is

det⁡(R−λI)=λ2+1\det(R - \lambda I) = \lambda^2 + 1

which has no real roots. Its roots are λ=±i\lambda = \pm i. Allowing complex numbers restores the eigenvalues that the geometry hid.

For any 2×22 \times 2 matrix the characteristic polynomial is λ2−(tr⁡A)λ+det⁡A\lambda^2 - (\operatorname{tr}A)\lambda + \det A, so the eigenvalues are complex exactly when (tr⁡A)2<4det⁡A(\operatorname{tr}A)^2 < 4\det A. They then come as a conjugate pair a±bia \pm bi, since a polynomial with real coefficients has conjugate roots.

Why a ± bi means rotate and scale

Consider C=[a−bba]C = \begin{bmatrix} a & -b\\ b & a \end{bmatrix}. Its eigenvalues are a±bia \pm bi. Write a+bia + bi in polar form, r(cos⁡φ+isin⁡φ)r(\cos\varphi + i\sin\varphi), as on the complex plane. Then

C=r[cos⁡φ−sin⁡φsin⁡φcos⁡φ]C = r\begin{bmatrix} \cos\varphi & -\sin\varphi\\ \sin\varphi & \cos\varphi \end{bmatrix}

a rotation by φ\varphi followed by a scaling by rr. Multiplying by CC does to vectors exactly what multiplying by a+bia + bi does to complex numbers. Every real 2×22 \times 2 matrix with eigenvalues a±bia \pm bi is similar to this CC, so in the right coordinates it is a rotation and a scaling too. The coordinates may be skewed, so its orbits trace ellipses rather than circles.

Worked examples

Common mistakes

Practice problems

  1. Find the eigenvalues of [0−220]\begin{bmatrix} 0 & -2\\ 2 & 0 \end{bmatrix} and describe the matrix.

    Answer

    ±2i\pm 2i; a rotation by 90°90° and a stretch by 22

    Full solution

    λ2+4=0\lambda^2 + 4 = 0. The form [a−bba]\begin{bmatrix} a & -b\\ b & a \end{bmatrix} has a=0a = 0, b=2b = 2, so r=2r = 2 and the angle is 90°90°.

  2. Find the eigenvalues of [21−12]\begin{bmatrix} 2 & 1\\ -1 & 2 \end{bmatrix} and their modulus.

    Answer

    2±i2 \pm i; modulus 5\sqrt{5}

    Full solution

    Trace 44, determinant 55: λ2−4λ+5=0\lambda^2 - 4\lambda + 5 = 0 gives λ=2±i\lambda = 2 \pm i, and ∣2+i∣=5|2 + i| = \sqrt{5}.

  3. For A=[1−331]A = \begin{bmatrix} 1 & -\sqrt{3}\\ \sqrt{3} & 1 \end{bmatrix}, find the eigenvalues, rr and the angle, and compute A3A^3.

    Answer

    1±3 i1 \pm \sqrt{3}\,i; r=2r = 2; 60°60°; A3=−8IA^3 = -8I

    Full solution

    Three steps rotate by 180°180° and scale by 23=82^3 = 8. A half turn sends every vector to its negative, so A3=−8IA^3 = -8I.

  4. Show that [a−bba]\begin{bmatrix} a & -b\\ b & a \end{bmatrix} has eigenvalues a±bia \pm bi.

    Answer

    The characteristic polynomial is (a−λ)2+b2(a - \lambda)^2 + b^2, which is zero when a−λ=±bia - \lambda = \pm bi.

    Full solution

    det⁡[a−λ−bba−λ]=(a−λ)2+b2\det\begin{bmatrix} a - \lambda & -b\\ b & a - \lambda \end{bmatrix} = (a - \lambda)^2 + b^2.

  5. Find an eigenvector of [0−110]\begin{bmatrix} 0 & -1\\ 1 & 0 \end{bmatrix} for λ=i\lambda = i.

    Answer

    (1,−i)(1, -i), or any nonzero multiple

    Full solution

    The first row of R−iIR - iI is (−i,−1)(-i, -1), so −iv1−v2=0-iv_1 - v_2 = 0 and v2=−iv1v_2 = -iv_1.

  6. A 2×22 \times 2 real matrix has trace 66 and determinant 1313. Find its eigenvalues.

    Answer

    3±2i3 \pm 2i

    Full solution

    λ2−6λ+13=0\lambda^2 - 6\lambda + 13 = 0 gives λ=3±9−13=3±2i\lambda = 3 \pm \sqrt{9 - 13} = 3 \pm 2i.

  7. If a matrix has eigenvalues a±bia \pm bi with a2+b2<1a^2 + b^2 < 1, what happens to AkxA^k\mathbf{x} as kk grows?

    Answer

    It spirals in toward 0\mathbf{0}.

    Full solution

    Each step rotates and multiplies the size by r<1r < 1, so the size shrinks like rkr^k while the direction keeps turning.

  8. A student says [1−111]\begin{bmatrix} 1 & -1\\ 1 & 1 \end{bmatrix} has no eigenvalues because λ2−2λ+2\lambda^2 - 2\lambda + 2 has a negative discriminant. Correct the statement.

    Answer

    It has no real eigenvalues. Its eigenvalues are the complex pair 1±i1 \pm i.

    Full solution

    The quadratic formula gives λ=2±−42=1±i\lambda = \tfrac{2 \pm \sqrt{-4}}{2} = 1 \pm i. They describe a 45°45° rotation with a 2\sqrt{2} stretch.

Frequently asked questions

When does a real matrix have complex eigenvalues?

For a 2 × 2 matrix, when its characteristic polynomial λ² − (trace)λ + det has a negative discriminant: (trace)² < 4·det.

Why do complex eigenvalues come in pairs?

The characteristic polynomial has real coefficients, so if a + bi is a root, its conjugate a − bi is a root too.

What does a complex eigenvalue mean geometrically?

For a 2 × 2 matrix with eigenvalues a ± bi, the matrix rotates by the angle of a + bi and scales by r = √(a² + b²), in a suitable coordinate system.

What are the eigenvalues of a rotation matrix?

The rotation by angle θ has eigenvalues cos θ ± i sin θ, which have modulus 1: a pure rotation, no stretching.

How do you find a complex eigenvector?

Solve (A − λI)v = 0 with complex arithmetic. For a 2 × 2 matrix one row is enough, since the two rows are multiples of each other.

What to learn next