Definite matrix

This is an old revision of this page, as edited by AxelBoldt (talk | contribs) at 10:58, 31 October 2003. The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

In linear algebra, the positive definite matrices are (in several ways) analogous to the positive real numbers. An n × n Hermitian Hermitian matrix M is said to be positive definite if it has one (and therefore all) of the following 6 equivalent properties:

(1) For all non-zero vectors z in Cn we have

z* M z > 0.

Here we view z as a column vector with n complex entries and z* as the complex conjugate of its transpose.

(2) For all non-zero vectors x in Rn we have

xT M x > 0

(where xT denotes the transpose of the column vector x).

(3) For all non-zero vectors u in Zn (all components being integers), we have

uT M u > 0.

(4) All eigenvalues of M are positive.

(5) The form

<x, y> = x* M y

defines an inner product on Cn. (In fact, every inner product on Cn arises in this fashion from a Hermitian positive definite matrix.)

(6) All the following matrices have positive determinant: the upper left 1-by-1 corner of M, the upper left 2-by-2 corner of M, the upper left 3-by-3 corner of M, ..., and M itself.

Further properties

Every positive definite matrix is invertible and its inverse is also positive definite. If M is positive definite and r > 0 is a real number, then rM is positive definite. If M and N are positive definite, then M + N is also positive definite, and if MN = NM, then MN is also positive definite. To every positive definite matrix M, there exists precisely one square root: a positive definite matrix N with N2 = M.


Negative definite, semidefinite and indefinite matrices

The Hermitian matrix M is said to be negative definite if

x* M x < 0

for all non-zero x in Rn (or, equivalently, all non-zero x in Cn). It is called positive semidefinite if

x* M x ≥ 0

for all x in Rn (or Cn) and negative semidefinite if

x* M x ≤ 0

for all x in Rn (or Cn).

A Hermitian matrix which is neither positive nor negative semidefinite is called indefinite.

Generalizations

Suppose K denotes the field R or C, V is a vector space over K, and B : V × VK is a bilinear map which is Hermitian in the sense that B(x,y) is always the complex conjugate of B(y,x). Then B is called positive definite if B(x,x) > 0 for every nonzero x in V.