Archives of the TeradataForum
Message Posted: Mon, 31 Jul 2000 @ 20:09:25 GMT
| Subj: || || Re: True rounding algorithm |
| From: || || Barry Hull |
The way I've always understood it and the way it's worked to my knowledge is that all of those would have been rounded to 100. The case
where it would be rounded down would be if the number to the left of the decimal was even. So, 98.5 (and all of it's equivalents) would be
rounded to 98.
Are you saying that you're seeing something different? If 99.500 rounded to 99, then I would consider that a bug.