This is in fact exactly the rounding specified by the IEEE floating point standard IEEE 754 (1985 and 2008). It is intended to make rounding unbiased.
確定! 回上一頁