np.expand_dims()

np.expand_dims:用于扩展数组的形状

原始数组：

import numpy as np

In [12]:
a = np.array([[[1,2,3],[4,5,6]]])
a.shape
Out[12]:
(1, 2, 3)

np.expand_dims(a, axis=0)表示在0位置添加数据,转换结果如下：

In [13]:
b = np.expand_dims(a, axis=0)
b
Out[13]:
array([[[[1, 2, 3],
         [4, 5, 6]]]])

In [14]:
b.shape
Out[14]:
(1, 1, 2, 3)

np.expand_dims(a, axis=1)表示在1位置添加数据,转换结果如下：

In [15]:
c = np.expand_dims(a, axis=1)
c
Out[15]:
array([[[[1, 2, 3],
         [4, 5, 6]]]])

In [16]:
c.shape
Out[16]:
(1, 1, 2, 3)

np.expand_dims(a, axis=2)表示在2位置添加数据,转换结果如下：

In [17]:
d = np.expand_dims(a, axis=2)
d
Out[17]:
array([[[[1, 2, 3]],
        [[4, 5, 6]]]])

In [18]:
d.shape
Out[18]:
(1, 2, 1, 3)

np.expand_dims(a, axis=3)表示在3位置添加数据,转换结果如下：

In [19]:
e = np.expand_dims(a, axis=3)
e

In [20]:
e.shape
Out[20]:
(1, 2, 3, 1)

能在(1,2,3)中插入的位置总共为4个，再添加就会出现以下的警告，要不然也会在后面某一处提示AxisError。

In [21]:
f = np.expand_dims(a, axis=4)
D:\ProgramData\Anaconda3\envs\dlnd\lib\site-packages\ipykernel_launcher.py:1: DeprecationWarning: Both axis > a.ndim and axis < -a.ndim - 1 are deprecated and will raise an AxisError in the future.
  """Entry point for launching an IPython kernel.

a = np.array([[[1,2,3],[4,5,6]]])
b = np.expand_dims(a,axis=-1)
print(b.shape)

axis = -1 也就是相当于在最后插入

(1, 2, 3, 1)