将遮罩阵列2d应用于3d [英] Apply Mask Array 2d to 3d

问题描述

我想将2维(NxM数组)的蒙版应用于3维数组(KxNxM数组).我该怎么办?

I want to apply a mask of 2 dimensions (an NxM array) to a 3 dimensional array (a KxNxM array). How can I do this?

2d =纬度x纬度

3d =时间x纬度x lon

3d = time x lat x lon

import numpy as np

a = np.array(
    [[[ 0,  1,  2],
      [ 3,  4,  5],
      [ 6,  7,  8]],

     [[ 9, 10, 11],
      [12, 13, 14],
      [15, 16, 17]],

     [[18, 19, 20],
      [21, 22, 23],
      [24, 25, 26]]])

b = np.array(
    [[ 0, 1, 0],
     [ 1, 0, 1],
     [ 0, 1, 1]])

c = np.ma.array(a, mask=b)  # this behavior is wanted 

推荐答案

有很多不同的方法可供选择.您想要做的是将(较低维度的)蒙版与具有额外维度的数组对齐:重要的是，您要使两个数组中的元素数量相同，如第一个示例所示:

There are quite a few different ways to choose from. What you want to do is align the mask (of lower dimension) to the array that has the extra dimension: the important part is that you get the number of elements in both arrays the same, as the first example shows:

np.ma.array(a, mask=np.concatenate((b,b,b)))  # shapes are (3, 3, 3) and (9, 3)
np.ma.array(a, mask=np.tile(b, (a.shape[0],1)))  # same as above, just more general as it doesn't require you to specify just how many times you need to stack b.
np.ma.array(a, mask=a*b[np.newaxis,:,:])  # used broadcasting

这篇关于将遮罩阵列2d应用于3d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！