数组逐行替换
import numpy as np
init_matrix = np.random.uniform(-1,1,[2,3])
test_map = map(lambda row : row / 2, init_matrix)
#下面这句话是python2的语法
#python3使用: test_matrix = np.array(list(test_map))
test_matrix = np.array(test_map)
print(init_matrix)
print(test_matrix)
init_matrix:
array([[-0.48006762, 0.6910708 , 0.53442369],
[ 0.60233271, 0.40702649, -0.70795476]])
test_matrix:
array([[-0.24003381, 0.3455354 , 0.26721185],
[ 0.30116635, 0.20351324, -0.35397738]])
生成多维数组，元素随机
import numpy as np
#[0,10),左闭右开,整数
np.random.randint(0,10,size=[3,3])
array([[4, 8, 2],
[5, 3, 6],
[5, 9, 5]])
#[0.0,1.0)
np.random.random(size=[3,3])
array([[ 0.54587373, 0.43366728, 0.91261908],
[ 0.3425949 , 0.83666684, 0.5742228 ],
[ 0.29507124, 0.16803488, 0.04462426]])
dict按照key排序生成list、dict按照value排序生成list
