help-octave
[Top][All Lists]
Advanced

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

Re: [O] Experiences with G.Strang: Intro to Linear Algebra?


From: A. Danial
Subject: Re: [O] Experiences with G.Strang: Intro to Linear Algebra?
Date: Fri, 31 Oct 2003 23:00:56 -0600
User-agent: KMail/1.5.3

On Thursday 30 October 2003 10:43 pm, John W. Eaton wrote:
> On 30-Oct-2003, Streph Treadway <address@hidden> wrote:
> | Excellent.  I have found that book very useful.  About a year ago, I
> | modified a little of the code to run with octave.  I'll look for it.
> | As I recall, some trivial modifications were necessary.
>
> How about reporting what changes are needed?

Octave's eig() function has caused me grief when trying to run
unmodified matlab .m files.

matlab's eig() can solve both the standard (one matrix) and general 
(two matrix) eigenvalue problems.  Octave has separate functions for 
these problems:  its eig() only handles the standard eigenproblem
while qz() solves the the general eigenproblem.  The different behavior
between matlab and octave means I have to grep through matlab .m files 
for "eig", and either leave the occurrences alone or change them to "qz".
In my line of work, structural dynamics, most instances of eig() are to
solve the general eigenproblem using stiffness and mass matrices as
inputs.  However, the standard eigenproblem appears frequently as well.

Would it be feasible to overload octave's eig() function to call qz()
under the hood if two matrices are passed in?  That would eliminate
the biggest matlab/octave linear algebra incompatibility I'm aware of 
(and would save me a lot of future hassle!)       -- Al



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

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------



reply via email to

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