Re: rational to floating point

Thant Tessman wrote:
> And I think that this in turn implies that if and only if
> gcd(d,b) is 1 and 'd' is not 1, then the original rational number can
> only be represented by an infinite stream of digits.


No, that doesn't work; consider writing 5/6 in base 10 (0.8333...).
gcd(6,10)=2.


The logic you want is that it's a repeating decimal if d has any prime
factors which do not divide into b. I'm not sure offhand of an
efficient way to do this test, though.


However, you might consider that a better way to cope with repeating
decimals is to print them as repeating decimals. A useful tidbit to
remember is that l, the length of the repeating part, divides d-1 (for
example, 1/7 is 0.[142857], where the 6 digits in [] repeat). Given
that, you may be able to come up with a good algorithm for the printing.


--
/=============================================================\
|John Stracke | http://www.thibault.org |HTML OK |
|Francois Thibault |========================================|
|East Kingdom |So what's the gene for belief in genetic|
|francis@thibault.org|determinism? |
\=============================================================/

Return to the comp.compilers page.
Search the comp.compilers archives again.