An matrix is diagonalizable over a chosen field exactly when it has a basis of linearly independent eigenvectors. Equivalently, is diagonal for some invertible . Distinct eigenvalues suffice but are not necessary. For example,
is not diagonalizable: its only eigenvalue is and its eigenspace is the one-dimensional span of . It is a nontrivial Jordan block.
A Hermitian matrix satisfies , where the dagger denotes conjugate transpose. If with , then
The denominator is positive real, so
For and , Hermiticity also gives
Thus
This is the necessary qualification to the PDF's orthogonality wording: arbitrary eigenvectors belonging to a repeated eigenvalue need not be orthogonal. Within each repeated eigenspace, Gram-Schmidt orthogonalization produces an orthonormal basis; together the eigenspaces give an orthonormal eigenbasis by the finite-dimensional spectral theorem. For example, and are nonorthogonal eigenvectors of the identity matrix.
The block has characteristic polynomial . A convenient orthonormal eigenbasis for the full real symmetric matrix is
with eigenvalues , , and , respectively. Hence its orthogonal diagonalization is
For repeated eigenvalues, every nonzero combination within the corresponding eigenspace is also an eigenvector; the displayed basis remains valid. Introduce principal axes coordinates
The quadratic form becomes
The principal-axis reduction of a quadric therefore has three regimes.
For , all coefficients are positive: the surface is an ellipsoid centred at the origin, with
For , the term disappears. The surface is an elliptic cylinder, unbounded along , whose cross-section is :
For , there is one negative coefficient, so the surface is a one-sheet hyperboloid:
Its waist at is an ellipse with semiaxes and . At other , both waist radii are multiplied by . There is no intercept on its longitudinal axis. Its asymptotic cone is . If positive eigenvalues coincide, the corresponding choice of principal axes is not unique, reflecting a rotational symmetry of the surface.
Figure 1.
Ellipsoid, elliptic cylinder, and one-sheet hyperboloid in principal coordinates
. Original calculated examples with and . The horizontal longitudinal coordinate is directed along in the original coordinates. Cylinder and hyperboloid plots show finite portions of unbounded surfaces. Details.

Articles by others on the same topic (0)

There are currently no matching articles.