python - 通过平均或重新组合一个 numpy 2d 数组来调整大小

Question

我正在尝试在 python 中重新实现一个 IDL 函数:

http://star.pst.qub.ac.uk/idl/REBIN.html

通过平均将二维数组缩小一个整数因子。

例如:

>>> a=np.arange(24).reshape((4,6))
>>> a
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]])

我想通过取相关样本的平均值将其调整为 (2,3)，预期输出为:

>>> b = rebin(a, (2, 3))
>>> b
array([[  3.5,   5.5,  7.5],
       [ 15.5, 17.5,  19.5]])

即b[0,0] = np.mean(a[:2,:2]), b[0,1] = np.mean(a[:2,2:4]) 和以此类推。

我相信我应该 reshape 为一个 4 维数组，然后在正确的切片上取平均值，但无法弄清楚算法。你有什么提示吗？

Answer 1

这是一个基于 the answer you've linked 的示例(为清楚起见):

>>> import numpy as np
>>> a = np.arange(24).reshape((4,6))
>>> a
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]])
>>> a.reshape((2,a.shape[0]//2,3,-1)).mean(axis=3).mean(1)
array([[  3.5,   5.5,   7.5],
       [ 15.5,  17.5,  19.5]])

作为一个函数:

def rebin(a, shape):
    sh = shape[0],a.shape[0]//shape[0],shape[1],a.shape[1]//shape[1]
    return a.reshape(sh).mean(-1).mean(1)