← 🎯 Eigenvalues & Eigenvectors
Subtopic 6 of 6

5.6Eigenvectors: independence and orthogonality

The idea

Eigenvectors for different eigenvalues are independent, and for a real symmetric matrix they are even orthogonal.

Think of it like this: Special matrices produce tidy, right-angled axes; general ones produce slanted axes.

Why it exists

Questions like "the dot product of eigenvectors of different eigenvalues of a symmetric matrix" have a one-word answer: 0.

How it works

  1. Suppose Av = λv, Aw = μw with λ ≠ μ and A symmetric.
  2. Then λ(v·w) = (Av)·w = v·(Aw) = μ(v·w).
  3. So (λ − μ)(v·w) = 0, and v·w = 0.
  4. For a non-symmetric matrix the eigenvectors are independent but generally not orthogonal.

Worked example: GATE 2014

A 4 × 4 symmetric positive definite matrix has eigenvectors for different eigenvalues with dot product 0.

Now play with it

  • →Load the symmetric example and read the zero off-diagonals in the dot-product table.
  • →Load the non-symmetric one and see non-zero dot products.
symmetric
dot productv for λ = 1v for λ = 4
v for λ = 120
v for λ = 403

Green zeros: eigenvectors of different eigenvalues of a real symmetric matrix are orthogonal. (Repeated eigenvalues give a whole eigenspace; vectors inside it need not be orthogonal until you choose them so.)

det (area scale)3
trace4
eigenvalues3, 1

The gold shape is the unit square after the map. Dashed lines are the eigenvector directions (when real): points on them are only stretched by λ, never turned. If det is negative the picture is flipped; if det is 0 it collapses to a line.

Check yourself

The dot product of eigenvectors of a real symmetric matrix belonging to different eigenvalues is

Remember: Different eigenvalues ⇒ independent eigenvectors; symmetric ⇒ orthogonal.
Exam notes (the dense version)

Eigenvectors for different eigenvalues are linearly independent. For a real symmetric matrix, eigenvectors for different eigenvalues are also orthogonal (dot product 0). Any non-zero multiple of an eigenvector is an eigenvector, and the eigenvectors of one eigenvalue (plus 0) form a subspace, the eigenspace. A matrix with only one eigenvector direction is not diagonalisable and has a repeated eigenvalue.

Previous year questions: Eigenvalues & Eigenvectors (14)

PYQs are grouped by topic, so this is the full set for Eigenvalues & Eigenvectors; try the ones that match this subtopic. Open all 14 on the topic page.