Decimals to fraction strings

[François Pinard]

Mikael Olofsson <mikael at isy.liu.se> écrit:

> On 17-May-00 François Pinard wrote:
>  >  Not necessarily.  0.6667 is well approximated by 1:3, for example, while
>  >  if you force the denominator to be an exponent of 10, you will obtain a
>  >  fraction which is not only uglier, but less precise.

> Precise in what respect?

I mean that the fraction you get is further than 0.6667 than 1:3 is.

> Given the (possibly rounded) number 0.6667, we do not know if its origin
> is 2/3 or something else.  It might very well be 6667/10000.  The latter is
> definitely closer to the given number than the former, thus more precise.

Of course, if you do not tolerate any error, 6667:10000 is the only
solution. :-)

-- 
François Pinard   http://www.iro.umontreal.ca/~pinard