Re: More on trits & tetrits... (long)
On 2010-04-22 07:41:50 -0700, Dan Zuras Intervals wrote:
> Thus, if xx = 1/[0,1] = {[1,oo],{boundedFalse}} &
> yy = 1/[eps,1] = {[1,oo],{boundedTrue}} we can answer
> questions like isContained(xx,yy) or isContained(yy,xx)
> but the answer ALSO lives in the decoration.
Not necessarily: one could ignore the decoration for the definition
of isContained (the "bounded" decoration would mostly be some
information for the user).
> The other way, the 'bounded' decoration is defined in
> exactly the same way as any other decoration: one bit
> tells you if there are points for which the decoration
> is true & the other bit tells you if there are points
> for which the decoration is false.
But is this useful in pratice for the "bounded" decoration?
BTW your definition is rather ambiguous because it tells
"rounded to..." without saying the rounding-direction attribute.
> As I explained in the accompanying text, this is because
> overflow introduces a range error in the result that is
[...]
I don't see this as a problem... unless you have a practical example.
--
Vincent Lefèvre <vincent@xxxxxxxxxx> - Web: <http://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / Arénaire project (LIP, ENS-Lyon)