The characteristic polynomial
Eigenvalues are the roots of a polynomial built from the determinant. Finding them, including repeated and complex roots, is the first step of eigenvalue analysis.
The characteristic equation
\(\lambda\) is an eigenvalue of A exactly when \(A - \lambda I\) is singular, that is, when \(\det(A - \lambda I) = 0\). This is the characteristic equation. For an n x n matrix, \(\det(A - \lambda I)\) is a polynomial of degree n in \(\lambda\), the characteristic polynomial, so A has at most n eigenvalues.
The 2 x 2 case
For \(A = \begin{bmatrix} a & b \\ c & d \end{bmatrix}\), \(\det(A - \lambda I) = \lambda^2 - (a + d)\lambda + (ad - bc)\). The coefficient \(a + d\) is the trace, the sum of the diagonal, and the constant term is the determinant. So for 2 x 2 matrices, \(\lambda^2 - (\operatorname{tr} A)\lambda + \det A = 0\).
Trace, determinant, and eigenvalues
For any n x n matrix, counted with multiplicity and allowing complex values, the sum of the eigenvalues equals the trace and their product equals the determinant: \(\lambda_1 + \dots + \lambda_n = \operatorname{tr} A\) and \(\lambda_1 \cdots \lambda_n = \det A\). These identities provide quick checks on computed eigenvalues.
Algebraic and geometric multiplicity
An eigenvalue's algebraic multiplicity is its multiplicity as a root of the characteristic polynomial. Its geometric multiplicity is the dimension of its eigenspace. The geometric multiplicity is at least 1 and at most the algebraic multiplicity. For \(\begin{bmatrix} 2 & 1 \\ 0 & 2 \end{bmatrix}\), eigenvalue 2 has algebraic multiplicity 2 but geometric multiplicity 1.
Complex eigenvalues
Real matrices can have complex eigenvalues, which come in conjugate pairs \(a \pm bi\). The rotation matrix by angle \(\theta\) has eigenvalues \(\cos\theta \pm i\sin\theta\). Complex eigenvalues signal rotation: in dynamical systems they produce oscillations, spiralling inwards or outwards depending on whether their modulus is less than or greater than 1.
Similar matrices share eigenvalues
If \(B = P^{-1}AP\), then \(\det(B - \lambda I) = \det(P^{-1}(A - \lambda I)P) = \det(A - \lambda I)\), so similar matrices have the same characteristic polynomial and the same eigenvalues, trace, and determinant. Eigenvalues are properties of the underlying transformation, not of the coordinates chosen.
Computing eigenvalues in practice
For n of 5 or more, there is no general formula for polynomial roots in radicals (the Abel-Ruffini theorem), and in any case computing the characteristic polynomial is numerically unstable. Software uses the QR algorithm, developed independently by John Francis and Vera Kublanovskaya around 1961, which repeatedly factors and recombines the matrix until eigenvalues appear on the diagonal.
Worked example
Find the eigenvalues of \(A = \begin{bmatrix} 4 & 1 \\ 2 & 3 \end{bmatrix}\). The trace is 7 and the determinant is \(12 - 2 = 10\), so \(\lambda^2 - 7\lambda + 10 = 0\), giving \((\lambda - 5)(\lambda - 2) = 0\): \(\lambda = 5\) and \(\lambda = 2\). Check: \(5 + 2 = 7\) and \(5 \times 2 = 10\). For \(\lambda = 5\), \(A - 5I\) has rows (-1, 1) and (2, -2), giving eigenvector (1, 1).
Key takeaways
Eigenvalues solve det(A - lambda I) = 0, the characteristic equation; the characteristic polynomial has degree n, so there are at most n eigenvalues. For 2 x 2 matrices, lambda^2 - (trace) lambda + det = 0. Eigenvalues sum to the trace and multiply to the determinant. Algebraic multiplicity counts repeated roots; geometric multiplicity is the eigenspace dimension, never larger. Real matrices can have complex conjugate eigenvalue pairs, which signal rotation. Similar matrices share characteristic polynomials and eigenvalues. Practical computation uses the QR algorithm (Francis and Kublanovskaya, about 1961), not polynomial roots.
Flashcard
det(A - lambda I) = 0
Check your understanding
A 2 x 2 matrix has trace 6 and determinant 8. What are its eigenvalues?
