Householder reflections

Back to Matrix algorithms

A Householder reflection is an orthogonal, involutary matrix (i.e. a reflection), which is specified by the normal of the hyperplane w.r.t which the space is orthogonally reflected.

Definition

A Householder reflection w.r.t. is defined by

Properties

A Householder reflection is an orthogonal reflection of a vector w.r.t. the hyperplane whose normal is :

where

is symmetric:

is orthogonal and involutary:

The orthogonality of implies that it preserves the 2-norm.

Matrix decompositions

The Householder reflections are used for factoring orthogonal matrices out of a general matrix . The QR-decomposition, for example, can be computed by repeated use of Householder reflections.

A commonly reoccuring case is the following. Given a matrix , we would like to factor , where is a Householder reflection which sends the span of to the span of . For brevity, let us assume . Since preserves 2-norms, we know that the must also satisfy

Therefore, is of the form

where

This gives two choices for the reflection normal . While both of them do the specified job, one of the choices is better than the other from the numerical viewpoint, because it avoids cancellation. This choice is given by

Multiplying with the Householder reflection from the right can be used to transform the rows of , instead of columns.

QR decomposition

As an example of the use of the Householder reflection, we shall now consider the QR decomposition. Given a matrix , let its first column be . Form a Householder reflection such that the span of is mapped to the span of (the i:th standard basis vector). Then

and has as its first column . Now recurse to consider the submatrix of which is obtained by removing the first row and the first column. Apply a Householder reflection to again map the first column of the submatrix to the first standard basis vector. Finally, extend to a full block-diagonal matrix where the upper-left block is an identity matrix. By repeating this procedure, one obtains the factorization . Denoting

yields the QR-decomposition.

Bidiagonal decomposition

Very similarly to the QR-decomposition, one can factor any matrix into a bidiagonal matrix surrounded by sequences of Householder reflections from both sides. This decomposition is used as an initial stage of computing the singular value decomposition.