QZ factorization for generalized eigenvalues
Syntax
[AA,BB,Q,Z,V] = qz(A,B)
Description
The qz
function gives access to what are normally only intermediate results in the computation of generalized eigenvalues.
[AA,BB,Q,Z,V] = qz(A,B)
produces upper triangular matrices AA
and BB
, and matrices Q
and Z
containing the products of the left and right transformations, such that
Q*A*Z = AA
Q*B*Z = BB
The qz
function also returns the generalized eigenvector matrix V
.
The generalized eigenvalues are the diagonal elements of AA
and BB
so that
A*V*diag(BB) = B*V*diag(AA)
Arguments
A,B
|
Square matrices.
|
AA,BB
|
Upper triangular matrices.
|
Q,Z
|
Transformation matrices.
|
V
|
Matrix whose columns are eigenvectors.
|
Algorithm
Complex generalizations of the EISPACK routines QZHES
, QZIT
, QZVAL
, and QZVEC
implement the QZ algorithm.
See Also
eig
Eigenvalues and eigenvectors
References
[1] Moler, C. B. and G.W. Stewart, "An Algorithm for Generalized Matrix Eigenvalue Problems", SIAM J. Numer. Anal., Vol. 10, No. 2, April 1973.
[ Previous | Help Desk | Next ]