[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: octave

From: A Scottedward Hodel
Subject: Re: octave
Date: Tue, 02 Oct 2001 14:38:12 -0500
User-agent: Microsoft-Entourage/9.0.2509

On 10/2/01 2:27 PM, "Dirk Bethmann" <address@hidden> wrote:

> Dear Mr. Hodel,
> I am interessted in a complex valued qz decomposition which is not
> available in my current octave release (2.1.31). My questions are:
> 1. is the mentioned complex qz decomposition available in the current
> octave version?
> 2. or: is there a simple way from the real valued qz decomposition (such
> that there are 2x2 blocks indicating complex eigenvalues) to the complex one?
> I would be very thankful if you are able to help me with my problem.
> Yours,
> Dirk Bethmann

I don't think the complex qz routines from LAPACK have been integrated into
Octave.  You should be able to take the existing interface to the
real-valued LAPACK qz routines and add routines to liboctave that would
allow the use of complex-valued qz.  I've thought about doing this in the
past but haven't had time or cause to do so.

Regarding the use of real-valued qz to perform complex qz work, yes I think
it's possible by separating your unknowns into their real and imaginary
parts and "fooling" the system into doing complex math.  However, this
doubles the size of your problem, which may not be desirable since qz is an
O(n^3) algorithm.



Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:
How to fund new projects:
Subscription information:

reply via email to

[Prev in Thread] Current Thread [Next in Thread]