numpy.load()函数从具有npy扩展名(.npy)的磁盘文件返回输入数组。

用法：

numpy.load(file, mmap_mode=None, allow_pickle=True, fix_imports=True, encoding=’ASCII’)

参数：

file :file-like对象，字符串或pathlib.Path。要读取的文件。 File-like对象必须支持seek()和read()方法。

mmap_mode :如果不为None，则使用给定模式memory-map文件(有关详细信息，请参见numpy.memmap

模式说明)。

allow_pickle :允许加载存储在npy文件中的腌制对象数组。

fix_imports :仅在在Python 3上加载Python 2生成的腌制文件时有用，该文件包括包含对象数组的npy /npz文件。

encoding :仅当在Python 3中加载Python 2生成的腌制文件时有用，该文件包含包含对象数组的npy /npz文件。

Returns :数据存储在文件中。对于.npz文件，必须关闭NpzFile类的返回实例，以避免泄漏文件描述符。

代码1：

# Python program explaining  
# load() function  
  
import numpy as geek 
  
a = geek.array(([i + j for i in range(3)  
                       for j in range(3)])) 
# a is printed. 
print("a is:") 
print(a) 
  
geek.save('geekfile', a) 
print("the array is saved in the file geekfile.npy") 
  
# the array is saved in the file geekfile.npy  
b = geek.load('geekfile.npy') 
  
# the array is loaded into b 
print("b is:") 
print(b) 
  
# b is printed from geekfile.npy 
print("b is printed from geekfile.npy")

输出：

a is:
[0, 1, 2, 1, 2, 3, 2, 3, 4]
the array is saved in the file geekfile.npy
b is:
[0, 1, 2, 1, 2, 3, 2, 3, 4]
b is printed from geekfile.npy

代码2：

# Python program explaining  
# load() function  
  
import numpy as geek 
  
# a and b are numpy arrays. 
a = geek.array(([i + j for i in range(3)  
                       for j in range(3)])) 
b = geek.array([i + 1 for i in range(3)]) 
  
# a and b are printed. 
print("a is:") 
print(a) 
print("b is:") 
print(b) 
  
# a and b are stored in geekfile.npz 
geek.savez('geekfile.npz', a = a, b = b) 
  
print("a and b are stored in geekfile.npz") 
  
# compressed file is loaded 
c = geek.load('geekfile.npz') 
  
print("after loading...") 
print("a is:", c['a']) 
print("b is:", c['b'])

输出：

a is:
[0 1 2 1 2 3 2 3 4]
b is:
[1 2 3]
a and b are stored in geekfile.npz
after loading...
a is:[0 1 2 1 2 3 2 3 4]
b is:[1 2 3]