Re: Negative zeros?

[Stefan van der Walt]

The article John referred to is available at

http://eins2002.info.uni-karlsruhe.de/literatur/goldberg-floating-point.pdf

The other one that David referred to was written by William Kahan
(sometimes called father of floating-point mathematics, being the
primary architect of the IEEE standard), but isn't available for free
on the web.  Here is a complete reference:

"Branch Cuts for Complex Elementary Functions, or Much Ado About
Nothing's Sign Bit" by William Kahan in "The State of the Art in
Numerical Analysis", (eds. Iserles and Powell), Clarendon Press,
Oxford, 1987.

However, he has many other interesting writings pertaining to
Floating-Point on his webpage at

http://www.cs.berkeley.edu/~wkahan/

(like "Matlab's Loss is Nobody's Gain" and "How Java's Floating-Point
hurts Everyone Everywhere").

Regards
Stéfan

On Tue, Sep 13, 2005 at 06:05:53AM +0200, David Bateman wrote:
> Vic Norton a écrit :
> 
> >Look at the following code
> >   octave> CS = [ 0 0 0 ]
> >   CS =
> >
> >     0  0  0
> >
> >   octave> CS *= -1
> >   CS =
> >
> >     -0  -0  -0
> >
> >I find this a real annoyance. To paraphrase Gertrude Stein:
> >   A zero is a zero is a zero.
> >Is there some easy way to stop Octave from outputting negative zeros?
> >
> >Regards,
> >
> >Vic
> >
> Vic,
> 
> Maybe you should do a google search for the article "Much ado about 
> zeros sign bit" and read about the importance of having two 
> representations of zero.
> 
> Regards
> David



-------------------------------------------------------------
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
-------------------------------------------------------------