numpy reshape和resize的区别（一清二楚）

主要两点区别。

区别一：

resize 无返回值（返回值为None），会改变原数组。

reshape 有返回值，返回值是被reshape后的数组，不会改变原数组。

import numpy as np

A = np.array([1, 2, 3, 4, 5, 6])

print("A:\n", A)

A_resize = A.resize((2, 3))
print("A_resize:\n", A_resize)
print("A(after resize):\n", A)

print('-'*10)

B = np.array([1, 2, 3, 4, 5, 6])

print("B:\n", B)

B_reshape = B.reshape((2, 3))
print("B_reshape:\n", B_reshape)
print("B(after reshape):\n", B)

区别二：

resize 可以放大或者缩小原数组的形状：放大时，会用0补全剩余元素；缩小时，直接丢弃多余元素。

reshape 要求reshape前后元素个数相同，否则会报错，无法运行。

import numpy as np

A = np.array([1, 2, 3, 4, 5, 6])

print("A:\n", A)

# 放大
A_resize = A.resize((3, 4))
print("A_resize:\n", A_resize)
print("A(after resize):\n", A)

# 缩小
A_resize = A.resize((2, 2))
print("A_resize:\n", A_resize)
print("A(after resize):\n", A)

print('-'*10)

B = np.array([1, 2, 3, 4, 5, 6])

print("B:\n", B)

B_reshape = B.reshape((3, 4))  # 这句会报错，reshape前后元素个数应当相同
print("B_reshape:\n", B_reshape)
print("B(after reshape):\n", B)