Merge: lib/standard/math: Added round method to floats

Yes, I know, it can be done right now using a simple trick like `1.34.to_i.to_f`, but it seems counter-intuitive to me, why not have a round method in Float (besides, it should be a bit faster) ?

I leave it to you, is it worth it to add this to standard ?

If it is, well +1 and let's merge this :)

Pull-Request: #924
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>