我有新问题 NumPy 维度

Question

我有一个关于 Keras 和许多其他应用程序中经常使用的 Numpy 维度的问题

问题 1

shape=(10,1) 和 (10,) 有什么区别？

问题 2

在使用 NumPy 的某些应用程序中什么是“无”， 例如）输入形状=（无，32） 是不是说具体尺寸还没有设置好？

问题 3

我该如何解释： ...轴=-1 如何将轴设置为-1？

Answer 1

1) 答：形状为 (10000,) 的数组是一维数组，而形状为 (10000,1) 的数组是二维数组

import numpy as np

# Array with shape (10000,) is a 1-Dimensional Array
one=np.ones((10,))
print("1-Dimensional Array:\n",one)

#Array with shape (10000,1) is a 2-Dimensional Array.
two=np.ones((10,1)) 
print("2-Dimensional Array:\n",two)

输出：

1-Dimensional Array:
 [1. 1. 1. 1. 1. 1. 1. 1. 1. 1.]

2-Dimensional Array:
 [[1.]
 [1.]
 [1.]
 [1.]
 [1.]
 [1.]
 [1.]
 [1.]
 [1.]
 [1.]]

2) 答案： None 只是为数组对象添加一个维度。您可以使用“None”或 numpy 的“newaxis”来创建新维度。

一般提示：您也可以使用 None 代替 np.newaxis

import numpy as np

#### Import the Fashion MNIST dataset
fashion_mnist = keras.datasets.fashion_mnist
(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()

#Original Dimension
print(train_images.shape)

train_images1 = train_images[None,:,:,:]
#Add Dimension using None
print(train_images1.shape)

train_images2 = train_images[np.newaxis is None,:,:,:]
#Add dimension using np.newaxis
print(train_images2.shape)

#np.newaxis and none are same
np.newaxis is None

输出：

(60000, 28, 28)
(1, 60000, 28, 28)
(1, 60000, 28, 28)
True

3) 答案： axis 参数指定新轴在结果维度中的索引。例如，如果axis=0，它将是第一个维度，如果axis=-1，它将是最后一个维度。

# input array 
a = np.array([[ 1, 2, 3], [ -1, -2, -3]] ) 
print ("1st Input array : \n", a)  

b = np.array([[ 4, 5, 6], [ -4, -5, -6]] ) 
print ("2nd Input array : \n", b)  

# Stacking the two arrays along axis 0 
out1 = np.stack((a, b), axis = 0) 
print ("Output array along axis 0:\n ", out1) 

# Stacking the two arrays along axis 1 
out2 = np.stack((a, b), axis = 1) 
print ("Output array along axis 1:\n ", out2) 

# Stacking the two arrays along axis -1 
out3 = np.stack((a, b), axis = -1) 
print ("Output array along axis -1:\n ", out3)

输出：

1st Input array : 
 [[ 1  2  3]
 [-1 -2 -3]]

2nd Input array : 
 [[ 4  5  6]
 [-4 -5 -6]]

Output array along axis 0:
  [[[ 1  2  3]
  [-1 -2 -3]]

 [[ 4  5  6]
  [-4 -5 -6]]]

Output array along axis 1:
  [[[ 1  2  3]
  [ 4  5  6]]

 [[-1 -2 -3]
  [-4 -5 -6]]]

Output array along axis -1:
  [[[ 1  4]
  [ 2  5]
  [ 3  6]]

 [[-1 -4]
  [-2 -5]
  [-3 -6]]]