编写数字库之前如何验证系统准确性? [英] how to verify system accuracy before writing numeric library?

查看:75
本文介绍了编写数字库之前如何验证系统准确性?的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

在编写数字库之前如何验证系统准确性?我想解决是由病毒编辑的CPU指令集还是由病毒编辑的窗口库引起的错误.


MEALY

how to verify system accuracy before writing numeric library? I would like to troubleshooting whether inaccuracy coming from CPU instruction set edited by virus or window library edited by virus.


MEALY

推荐答案

请定义系统精度.如果您想知道浮点类型的准确性,则它们遵循IEEE标准.在Double,Float和Decimal类中将数字范围定义为标识epsilon的属性,据我所知, 基本上是发生舍入误差时,实际值与计算值之差的范围.

Please define system accuracy. If you want to know how accurate the floating point types are then they follow the IEEE standard. The numeric ranges are defined in the Double, Float and Decimal classes as properties identifying the epsilon which, in my understanding, is basically the range of difference between the real value and the calculated one when rounding errors do occur.

在所有具有相同位数的CPU上,这将是相同的值,因此,我不确定我是否对您关于CPU指令集和病毒的讨论有所了解.

AFAIK this is going to be the same value on all CPUs of the same bitness so I'm not sure I follow anything about your discussion on CPU instruction sets and viruses.


这篇关于编写数字库之前如何验证系统准确性?的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆