Python创建类的多个实例

Question

I'm trying to create a NumPy array of objects as follows...

import numpy as np

class Pixel:
    def __init__(self):
        self.r = 0
        self.g = 0
        self.b = 0

class Image:
    def __init__(self):
        self.pixels = np.full((4, 4), Pixel())

if __name__ == '__main__':
    image = Image()
    print(image.pixels)

The output being...

[[<__main__.Pixel object at 0x02A92F70>
  <__main__.Pixel object at 0x02A92F70>]
 [<__main__.Pixel object at 0x02A92F70>
  <__main__.Pixel object at 0x02A92F70>]]

which appears to have created an array of the same object. My question is how to initialise the array with a different object for each pixel?

Kind regards

Answer 1

You can use np.array function and initialize it from standard python list as following:

class Image:
def __init__(self):
    self.pixels = np.array([Pixel() for i in range(2 * 2)]).reshape([2, 2])

The result is as following:

[[<__main__.Pixel object at 0x00000218B89F6438>
  <__main__.Pixel object at 0x00000218B89F64A8>]
 [<__main__.Pixel object at 0x00000218B89F4D68>
  <__main__.Pixel object at 0x00000218B89F43C8>]]

As you see they are not the same object.

Note that it will not hart your time complexity since in any case you need to initialize O(n) objects.

Answer 2

Thanks for all of the replies, most helpful. The method I have found that works best for my particular case is to use the numpy.empty() method to create an array of the correct size and then iterate through the elements filling the array with new objects. It might not be the most efficient but it is the most readable to me.