🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: sci.math.num-analysis
Article #578

EISPACK solution of generalized eigenvalue prob.

#578
From: Matt Wette
Date: Fri, 21 Dec 1990 22:06
25 lines
1499 bytes
This article concerns the problem of finding two orthogonal
matrices Q and Z such that 
	Q'*A*Z is in upper real Schur form
	Q'*B*Z is upper triangular
where A and B are square matrices and ' denotes transpose.

I have encountered problems with the EISPACK routine QZIT
not converging.  The problem I am trying to solve is of
order 18 and a `relative Crawford number' for the problem,
given by evaluation of the Matlab expression
	cond(A/norm(A) + j*B/norm(B)),
is about 10^2.

Does anyone know of software which is a better implementation 
of (a variant of) the QZ algorithm and/or a better working
algorithm for solving the same problem?

Matt


-- 
 _________________________________________________________________
 Matthew R. Wette           | Jet Propulsion Laboratory, 198-326
 mwette@csi.jpl.nasa.gov    | 4800 Oak Grove Dr, Pasadena,CA 91109
 -----------------------------------------------------------------
Technical Details
Message-ID: <1990Dec21.220608.11308@jato.jpl.nasa.gov>
Path: arc1-ca.pugleaf.net!utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!jato!mr-ed!mwette

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!jato!mr-ed!mwette
From: mwette@mr-ed.jpl.nasa.gov (Matt Wette)
Newsgroups: sci.math.num-analysis,comp.lang.fortran
Subject: EISPACK solution of generalized eigenvalue prob.
Message-ID: <1990Dec21.220608.11308@jato.jpl.nasa.gov>
Date: 21 Dec 90 22:06:08 GMT
Sender: news@jato.jpl.nasa.gov
Reply-To: mwette@mr-ed.jpl.nasa.gov (Matt Wette)
Organization: Jet Propulsion Laboratory, Pasadena, CA
Lines: 25
Nntp-Posting-Host: mr-ed.jpl.nasa.gov