规范化深度嵌套列表Python中的所有元素

Question

我有一个深度嵌套的int列表形式的图像数据：

len(train_data_imgs) = 3889       # number of images in set
len(train_data_imgs[0]) = 100     # height
len(train_data_imgs[0][0]) = 100  # width
len(train_data_imgs[0][0][0]) = 3 # these are ints - RGB pixel values

如何在0和1之间迭代以使它们正常化？只需将每个数字除以255即可。

Answer 1

使用NumPy包在一行中执行：

# Assuming an image stored in a nested list | here NumPy array
lst = np.arange(27).reshape(3,3,3)
lst
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]]])
lst = lst/255 # That's what you should look for
lst
array([[[0.        , 0.00392157, 0.00784314],
[0.01176471, 0.01568627, 0.01960784],
[0.02352941, 0.02745098, 0.03137255]],
[[0.03529412, 0.03921569, 0.04313725],
[0.04705882, 0.05098039, 0.05490196],
[0.05882353, 0.0627451 , 0.06666667]],
[[0.07058824, 0.0745098 , 0.07843137],
[0.08235294, 0.08627451, 0.09019608],
[0.09411765, 0.09803922, 0.10196078]]])