Random numbers

Random numbers#

Being able to generate random numbers is useful both for games and for computational techniques such as Monte Carlo methods.

Using the built-in random library#

The standard built-in random library which is part of the Python offers several useful functions:

  • random.randint(a, b) – random integer between a and b (inclusive)

  • random.random() – random float between 0.0 and 1.0

  • random.choice(seq) – random element from a sequence (typically a list)

  • random.shuffle(seq) – shuffle a list in-place (i.e. the list changes)

There are many other useful functions, see the full documentation for more information.

import random

ntries = 20
for i in range(ntries):
   print(random.randint(1,10)) 

4
6
10
9
3
8
3
9
6
5
9
7
6
3
1
10
5
9
6
4

for i in range(ntries):
    print(random.random())

0.6801894755686804
0.9094084191557079
0.8778989237796774
0.41144723463072563
0.40406117799697616
0.47482337203724945
0.7464046511319727
0.43319093725373137
0.8499297422602158
0.635046269898856
0.8367422048468426
0.677279670089847
0.35192814222375
0.2962411016873928
0.023720888050612343
0.467772937173183
0.12389269366357136
0.6245444719459522
0.9547784614389204
0.7382503515838159

l1 = ['rock','paper','scissors']
for i in range(ntries):
    print(random.choice(l1))

paper
rock
scissors
rock
paper
paper
rock
rock
paper
scissors
rock
rock
paper
rock
paper
rock
rock
rock
paper
paper

l2 = [0,1,2,3,4,5,6,7,8,9]
nshuffles = 5
for i in range(nshuffles):
    random.shuffle(l2)  # Note that l2 is changed here
    print(l2)

[0, 1, 5, 8, 6, 9, 4, 7, 2, 3]
[5, 2, 6, 4, 9, 7, 1, 3, 0, 8]
[4, 0, 6, 3, 5, 2, 1, 9, 7, 8]
[3, 6, 7, 4, 5, 8, 2, 1, 0, 9]
[4, 6, 8, 5, 1, 0, 3, 7, 9, 2]

Dungeons and Dragons dice#

d4 = lambda: random.randint(1,4)
d6 = lambda: random.randint(1,6)
d8 = lambda: random.randint(1,8)
d10 = lambda: random.randint(1,10)
d12 = lambda: random.randint(1,12)
d20 = lambda: random.randint(1,20)

d4()

3

d6()

5

d20()

3

Numpy and random numbers#

As a core package for numerical computation in Python, Numpy offers a variety of functionality for using random numbers. Often this is array-oriented (as opposed to single valued as in the built-in random module). Some examples of useful functions include:

  • np.random.random(size) - generate size random numbers uniformly in the range [0,1)

  • np.random.randint(low,high,size) - generate size random integers uniformly in the range [low,high)

  • np.random.normal(loc=0.0,scale=1.0,size=10) - generates size numbers from a normal/gaussian distribution centered at loc with standard devitation scale (note that use of the keywords is required)

See the full documentation for more possibilities.

import numpy as np

print(np.random.random(10))

[0.33220297 0.27094937 0.10991044 0.86823173 0.03204402 0.00451145
 0.2395526  0.71103928 0.84242267 0.21892699]

print(np.random.randint(1,7,10))

[2 4 5 4 4 2 2 6 2 5]

print(np.random.normal(size=10))

[-0.00737491 -0.48210976 -1.42249881 -0.3881652   0.37479936 -0.60501388
  0.22367986  0.28576499 -0.46607219  1.36660142]