numpy 处理三维矩阵

创建2页，3行，4列的三维矩阵

import numpy as np
t = np.random.random((2,3,4))
print(t)

[[[0.18589574 0.5723089  0.38505554 0.14135623]
  [0.73036206 0.58540968 0.57741716 0.02551514]
  [0.86965835 0.54080129 0.80261298 0.37144176]]

 [[0.83052614 0.83502562 0.47811981 0.723045  ]
  [0.33586194 0.39660702 0.58520309 0.15943745]
  [0.32893033 0.40983408 0.10889919 0.64272805]]]

提取第1页的二维矩阵（python的索引是从0开始）

print(t[1])

 [[0.83052614 0.83502562 0.47811981 0.723045  ]
 [0.33586194 0.39660702 0.58520309 0.15943745]
 [0.32893033 0.40983408 0.10889919 0.64272805]] 

按页方向求取平均值。行方向为axis=1，列方向axis=2

a = np.mean(t,0)

[[0.50821094 0.70366726 0.43158768 0.43220061]
 [0.533112   0.49100835 0.58131013 0.0924763 ]
 [0.59929434 0.47531768 0.45575608 0.50708491]]

提取第0页，第1行，第2列的数值

b = t[0][1][2]
print(b)

0.57741716