[Tutor] is this a bug? [Research papers on floating point]
Danny Yoo
dyoo@hkn.eecs.berkeley.edu
Wed, 2 Jan 2002 12:17:04 -0800 (PST)
On Wed, 2 Jan 2002, Gregor Lingl wrote:
>
> I understand; Remco braught it to the point:
>
> > The problem is that, when converted to binary, 66.6 is repeating. There is
> > no way to represent it exactly in a finite amount of binary digits.
> >
>
> It depends on the number beeing a repeating binary, not a repeating decimal
Hmmmm! I did a quick search on Google, and found an interesting paper on
this topic:
How to Read Floating Point Numbers Accurately:
http://citeseer.nj.nec.com/william90how.html
The math does seem a bit intimidating at first; I'd have to really study
it to understand it. Still, perhaps it might interest someone on Tutor
who's interested in floating-point representation issues.
The rest of the web site:
http://citeseer.nj.nec.com/directory.html
looks incredibly cool for anyone with a appetite for CS research papers;
this site appear to archive lots of good papers on computer science!