伪随机数概念

numpy.random 模块对 Python 内置的 random 进行了补充，增加了一些用于高效生成多种概率分布的样本值的函数

之所以被称作伪随机数而不是随机数是因为它们都是通过算法基于随机数生成器种子在确定性的条件下生成的

常用函数

• rand

  产生指定形状的均匀分布的样本值

• randint

  从给定的上下限范围内选取随机整数

• randn

  产生正态分布（平均值为 0，标准差为 1）的样本值，类似于 MATLAB 接口

• binomial

  产生二项分布的样本值

• normal

  产生正态（高斯）分布的样本值

• beta

  产生 Beta 分布的样本值

• chisquare

  产生卡方分布的样本值

• gamma

  产生 Gamma 分布的样本值

• uniform

  产生在指定范围中均匀分布的样本值

• seed

  确定随机数生成器的种子

• permutation

  返回一个序列的随机排列或返回一个随机排列的范围

• shuffle

  对一个序列就地随机排列

import numpy as np

np.random.rand(3,4)

array([[0.25576472, 0.79349969, 0.27850507, 0.90933056],
       [0.1968004 , 0.43184137, 0.85308284, 0.63514769],
       [0.63847559, 0.69836197, 0.93565657, 0.14745512]])

np.random.randint(5, size=(3, 4))

array([[2, 1, 3, 4],
       [0, 2, 0, 0],
       [0, 3, 1, 1]])

np.random.randn(2, 4)

array([[-1.79805827,  0.89237606, -1.21227477, -0.20863569],
       [-0.81141731,  1.74030762, -1.38814223, -0.03197699]])

mu, sigma = 0, 0.1
np.random.normal(mu, sigma, 10)

array([-0.03016733,  0.11870897, -0.05943379,  0.18551091, -0.01320171,
        0.03111292, -0.20898771,  0.17931511, -0.0242517 ,  0.09924793])

# 生成指定范围的样本值
np.random.uniform(1, 2, 100)

array([1.51425789, 1.42449746, 1.45887765, 1.91509535, 1.92949115,
       1.99208297, 1.26730716, 1.06134035, 1.10288814, 1.39686364,
       1.75348267, 1.08746638, 1.68855474, 1.02284299, 1.49622678,
       1.04692763, 1.38959372, 1.09019307, 1.37183829, 1.17243861,
       1.84493604, 1.73825886, 1.5884553 , 1.09940849, 1.84580879,
       1.67676695, 1.55877251, 1.58046459, 1.33408609, 1.29782968,
       1.75106753, 1.61941412, 1.26676746, 1.31206665, 1.20164265,
       1.99517187, 1.69007977, 1.32918295, 1.6460843 , 1.92691491,
       1.77403678, 1.80892126, 1.21145148, 1.34260398, 1.34524418,
       1.9365615 , 1.47606591, 1.39958125, 1.33588063, 1.88750897,
       1.42572515, 1.44809173, 1.71830767, 1.41811896, 1.19826113,
       1.32594578, 1.67826219, 1.85311161, 1.04663232, 1.42194556,
       1.92904229, 1.83680751, 1.64237948, 1.4085441 , 1.81492615,
       1.58980861, 1.23311607, 1.74682403, 1.52345087, 1.57221728,
       1.65626852, 1.03119812, 1.48341747, 1.64399359, 1.19550523,
       1.79558785, 1.98058089, 1.33055709, 1.76171829, 1.70375583,
       1.8616869 , 1.26968364, 1.96928118, 1.40622763, 1.52917968,
       1.64951899, 1.68269751, 1.64226908, 1.41391802, 1.83384097,
       1.56406308, 1.52084994, 1.182223  , 1.49735871, 1.29772454,
       1.2376237 , 1.64392624, 1.68856876, 1.53722271, 1.84031282])