bug in modulus?

[Christophe]

jantod at gmail.com a écrit :
> I think there might be something wrong with the implementation of
> modulus.
> 
> Negative float values close to 0.0 break the identity "0 <= abs(a % b)
> < abs(b)".
> 
>     print 0.0 % 2.0 # => 0.0
>     print -1e-010 % 2.0 # =>1.9999999999
> 
> which is correct, but:
> 
>     print -1e-050 % 2.0 # => 2.0
>     print -1e-050 % 2.0 < 2.0 # => False
> 
> This means I can't use modulus to "wrap around" before it reaches a
> certain value. I'm using Python 2.4.2 on WindowsXP.
> 
> Thanks
> Janto
> 

Consider that -1e-050 % 2.0 = 2.0 - 1e-050

Now, test that :
 >>> 2.0 - 1e-050 == 2.0
True

Floating point numbers just don't have the required precision to 
represent 2.0 - 1e-050. For your specific problem, if you really want 
the result to be < 2.0, the the best you can do is admit that floating 
point operations have errors and return 0.0 when the modulus operation 
equals 2.0.