将数组的值从nan更改为零 [英] change the values of an array from nan to zero

问题描述

我在python上有一个数组A，该数组具有由numpy.nan创建的一些nan值.我想使用A[A==numpy.nan] = 0将所有nan值设置为零.它根本不会改变数组.为什么会这样?

I have an array A on python that has some nan values created by numpy.nan. I want to set all the nan values to zero using A[A==numpy.nan] = 0. It doesn't change the array at all. Why is that?

推荐答案

您要 np.isnan :

A[np.isnan(A)] = 0

您的代码存在的问题是(根据IEEE)，nan不等于任何值，甚至不等于它本身.

The problem with your code is that (according to IEEE), nan doesn't equal anything -- even itself.

此外，还有

• np.isinf->(+/-无穷大)
• np.isfinite->(不是无穷大或NaN)
• np.isposinf->(+无限)
• np.isneginf->(-无穷大)

• np.isinf -> (+/- infinity)
• np.isfinite -> (not infinity or NaN)
• np.isposinf -> (+ infinity)
• np.isneginf -> (- infinity)

这篇关于将数组的值从nan更改为零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！