随机数：Python3.7的random模块详解

随机数模块random

作者:Shawn
python3.7
手册
https://docs.python.org/3/library/random.html#module-random
部分图片转自wiki

常用函数

random.seed()

random.seed(a=None, version=2)

• 设置随机种子 ，用于同步不同运行环境的随机数。

random.getstate()

random.getstate()

• 获得当前状态，用于恢复状态

random.setstate()

random.setstate(state)

• 恢复状态

>>> import random
>>> s=random.getstate()
>>> random.random()
0.15441857485858956
>>> random.random()
0.6330314601528841
>>> random.setstate(s)
>>> random.random()
0.15441857485858956
>>> random.random()
0.6330314601528841
>>> random.random()
0.04725013105129261

random.getrandbits()

random.getrandbits(k)

• 生成占内存k位以内的随机整数，硬核秃头专属。

>>> import random
>>> random.getrandbits(10)
674
>>> random.getrandbits(10)
10
>>> random.getrandbits(10)
745
>>> random.getrandbits(10)
560
>>> random.getrandbits(10)
162

random.random()

random.random()

• 随机产生一个[0,1.)数字。

>>> random() 
0.37444887175646646

random.uniform()

random.uniform(a, b)

• 产生一个a、b区间的随机数。

>>> uniform(2.5, 10.0)
3.1800146073117523

random.randrange()

random.randrange(start, stop[, step])

• 整数随机。
• 功能等同于choice(range(start, stop, step))。

>>> randrange(10)                        # 0到9随机
7
>>> randrange(0, 101, 2)                 # 0到100随机偶数
26

random.randint()

random.randint(a, b)

• 返回一个[a,b]的随机整数。
• 功能等同于randrange(a, b+1)。

random.choice()

random.choice(seq)

• 返回对象中的一个随机元素。

random.choices()

random.choices(population, weights=None, *, cum_weights=None, k=1)

• 随机选择，是random.choice(seq)的升级版本。

>>> choice(['win', 'lose', 'draw'])
'draw'

random.sample()

random.sample(population, k)

• 随机取样

>>> sample([10, 20, 30, 40, 50], k=4)
[40, 10, 50, 30]

---

• choices与sample的区别：
  
  • choices在抽取随机元素时是包含重复元素的，即：一个元素可能会被抽取多次。
  • 反之，在sample中，抽取的元素是不重复的。
  • 所以，在抽取元素大于样本集总数时，choices会继续而sample会报错：

>>> import random
>>> random.choices([1,2,3,4],k=4)
[4, 3, 4, 4]
>>> random.sample([1,2,3,4],k=4)
[3, 1, 2, 4]
>>> random.choices([1,2,3,4],k=5)
[3, 3, 2, 3, 2]
>>> random.sample([1,2,3,4],k=5)
Traceback (most recent call last):
  File "", line 1, in 
    random.sample([1,2,3,4],k=5)
  File "F:\python\lib\random.py", line 321, in sample
    raise ValueError("Sample larger than population or is negative")
ValueError: Sample larger than population or is negative
>>> 

---

random.shuffle()

random.shuffle(x[, random])

• 打乱序列

>>> deck = 'ace two three four'.split()
>>> shuffle(deck)
>>> deck
['four', 'two', 'ace', 'three']

不同的分布模式

random.triangular(low, high, mode)
https://en.wikipedia.org/wiki/Triangular_distribution
random.betavariate(alpha, beta)
https://en.wikipedia.org/wiki/Beta_distribution
random.expovariate(lambd)
https://en.wikipedia.org/wiki/Exponential_distribution
random.gammavariate(alpha, beta)
https://en.wikipedia.org/wiki/Gamma_distribution
random.gauss(mu, sigma)
https://en.wikipedia.org/wiki/Normal_distribution
random.lognormvariate(mu, sigma)
https://en.wikipedia.org/wiki/Log-normal_distribution
random.normalvariate(mu, sigma)
https://en.wikipedia.org/wiki/Normal_distribution
random.vonmisesvariate(mu, kappa)
https://en.wikipedia.org/wiki/Von_Mises_distribution
random.paretovariate(alpha)
https://en.wikipedia.org/wiki/Pareto_distribution
random.weibullvariate(alpha, beta)
https://en.wikipedia.org/wiki/Weibull_distribution

• random.triangular(low, high, mode)
  
  • 以三角分布的概率分布返回随机数。
    

import random
import collections
counter=collections.Counter() # 方便偷懒，用Counter统计次数。
for i in range(10000):
    step=str(round(random.triangular(0,9,7)))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')

================= RESTART: F:\PyWorkspace\blank.py =================
Counter({'7': 1963, '6': 1946, '5': 1617, '4': 1264, '8': 1105, '3': 944, '2': 668, '1': 323, '9': 138, '0': 32})
0
1***
2*******
3*********
4*************
5****************
6*******************
7********************
8***********
9*

• random.betavariate(alpha, beta)
  
  • 以beta分布的概率分布返回0到1之间的随机数。
    

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.betavariate(2,5)*10))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')


Counter({'2': 2424, '3': 2173, '1': 1941, '4': 1512, '5': 1000, '6': 416, '0': 316, '7': 167, '8': 48, '9': 3})
0***
1*******************
2************************
3**********************
4***************
5**********
6****
7**
8
9

• random.expovariate(lambd)
  
  • 以指数分布的概率分布返回随机数。

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.expovariate(1)))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')


============== RESTART: F:\PyWorkspace\LeetCode\expovariate.py ==============
Counter({'0': 3950, '1': 3862, '2': 1403, '3': 484, '4': 192, '5': 73, '6': 18, '7': 12, '8': 4, '9': 2})
0****************************************
1***************************************
2**************
3*****
4**
5*
6
7
8
9

• random.gammavariate(alpha, beta)
  
  • 以gamma分布的概率分布返回随机数。

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.gammavariate(9,0.5)))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')


============== RESTART: F:\PyWorkspace\LeetCode\gammavariate.py ==============
Counter({'4': 2671, '5': 2244, '3': 2054, '6': 1303, '2': 698, '7': 616, '8': 236, '1': 103, '9': 89, '0': 32})
0
1*
2*******
3*********************
4***************************
5**********************
6*************
7******
8**
9*

• random.gauss(mu, sigma)
  
  • 以高斯分布的概率分布返回随机数。

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.gauss(5,1)))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')

================= RESTART: F:\PyWorkspace\LeetCode\gauss.py =================
Counter({'5': 3803, '4': 2429, '6': 2418, '3': 660, '7': 582, '2': 59, '8': 41, '9': 6, '1': 2})
0
1
2*
3*******
4************************
5**************************************
6************************
7******
8
9

• random.lognormvariate(mu, sigma)
  
  • 以对数正态分布的概率分布返回随机数。

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.triangular(5,0.02)))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')


============= RESTART: F:\PyWorkspace\LeetCode\lognormvariate.py =============
Counter({'3': 3232, '2': 3168, '4': 1626, '1': 1621, '0': 178, '5': 175})
0**
1****************
2********************************
3********************************
4****************
5**
6
7
8
9

• random.normalvariate(mu, sigma)

  • 同random.gauss(mu, sigma)

• random.vonmisesvariate(mu, kappa)

  • 以von Mises分布的概率分布返回随机数。
  • 又作圆上正态分布

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.vonmisesvariate(3.14,2)))
    counter.update(step)

print(counter)

for i in range(7):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')


============ RESTART: F:\PyWorkspace\LeetCode\vonmisesvariate.py ============
Counter({'3': 4795, '4': 2609, '2': 1732, '5': 437, '1': 296, '6': 87, '0': 44})
0
1***
2*****************
3************************************************
4**************************
5****
6*

• random.paretovariate(alpha)
  
  • 以Pareto分布的概率分布返回随机数。

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.paretovariate(2)))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')



============= RESTART: F:\PyWorkspace\LeetCode\paretovariate.py =============
Counter({'1': 5644, '2': 2853, '3': 830, '4': 347, '5': 179, '6': 82, '7': 76, '8': 45, '9': 36, '0': 27})
0
1********************************************************
2*****************************
3********
4***
5**
6*
7*
8
9

• random.weibullvariate(alpha, beta)
  
  • 以Weibull分布的概率分布返回随机数。

import random
import collections
counter=collections.Counter()
for i in range(10000):
    step=str(round(random.weibullvariate(1,5)*5))
    counter.update(step)

print(counter)

for i in range(10):
    print(str(i),end='')
    value=round(counter[str(i)]/100.)
    for j in range(value):
        print('*',end='')
    print('')



============= RESTART: F:\PyWorkspace\LeetCode\weibullvariate.py =============
Counter({'5': 3601, '4': 2847, '6': 1753, '3': 1247, '2': 294, '7': 239, '1': 17, '8': 2})
0
1
2***
3************
4****************************
5************************************
6******************
7**
8
9