in reply to Rounding errors problem

Math::Round was recently updated to deal with floating point rounding. Check out sub nearest(). Cheers...