我想将2维掩码(NxM数组)应用于3维数组(KxNxM数组)。我怎么能这样做?
2d = lat x lon
3d =时间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
答案 0 :(得分:9)
有很多不同的选择方式。你想要做的是将掩码(较低维度)与具有额外维度的数组对齐:重要的是你得到两个数组中元素的数量相同,如第一个例子所示:
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