gnuplot:如何与NaN进行比较? [英] gnuplot: How to compare to NaN?

查看:49
本文介绍了gnuplot:如何与NaN进行比较?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

这里有一些关于此的主题(使用javascript,c ++等),但尚未找到适合gnuplot的主题.

There are several topics here about this (in javascript, c++,...), but haven't yet found one for gnuplot.

基本上,我想在标签中将"nan"替换为"---".

Basically, I want to replace "nan" with "---" in labels.

我知道我可以设置 NaN ="---" ,但是 NaN 的行为就像一个字符串,需要修改其他代码.

I know that I can set NaN = "---" but then NaN behaves like a string and other code needs to be adapted.

### How to compare to NaN?
NaN = GPVAL_NaN  # default value for NaN

a = NaN
print (a==NaN ? "---" : sprintf("%g",a))
print (a==NaN ? sprintf("%g",a) : "---")
print (a!=NaN ? "---" : sprintf("%g",a))
print (a!=NaN ? sprintf("%g",a) : "---")

a = 123
print (a==NaN ? "---" : sprintf("%g",a))
print (a==NaN ? sprintf("%g",a) : "---")
print (a!=NaN ? "---" : sprintf("%g",a))
print (a!=NaN ? sprintf("%g",a) : "---")
### end of code

结果:

NaN
---
NaN
---
123
---
123
---

因此,没有一种组合可以完成这项工作.

So, none of the combination does the job.

最后,我发现以下内容似乎可行.

Finally, I found that the following seems to work.

print (a/a == 1 ? sprintf("%g",a) : "---")

我的问题:

a/a 是否总是精确地为 1 ,或者在特殊情况下是否可能存在舍入错误?是这样做的方式"还是我忽略了任何事情?

Will a/a always be exactly 1 or could there potentially be rounding errors in special cases? Is this "the" way how to do it or did I overlook anything?

推荐答案

只是为了不让这个问题出现而没有答案...

just for the sake of not letting this question appearing unanswered...

一种比较变量 a 是否为 NaN (至少在gnuplot中)(并至少对其进行处理)的最简单方法:

Probably, the easiest way to compare if a variable a is NaN or not (at least in gnuplot) and do something with it:

print (a == a ? sprintf("%g",a) : "---")

在SO上对某些编程语言进行讨论之后,人们开始辩论 NaN == NaN NaN!= NaN 或两者都不是.

Following some discussions on some programming languages here on SO, people are debating about NaN==NaN or NaN!=NaN or neither of them.

如果与数据中的字符串"NaN" 进行比较,请参见上面的@Christoph注释.

And in case of comparison to the string "NaN" in data: see @Christoph's comment above.

这篇关于gnuplot:如何与NaN进行比较?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