Mersenne Twister in Python

Try to rebuild the pseudo-random algorithm Mersenne Twister, which is used in python's random library.

Also with a basic Random class and some simple methods for easily testing.

MT19937.py

Main part of the algorithm.

Convert the pseudocode in Mersenne Twister to python code.

Coefficients follow the standard of MT19937-32.

RandomClass.py

A class named Random.

Usage

Firstly, build a Random object. if no input, seed will default to 0.

>>> name = Random(seed)

 

.random():

return uniform ditribution in [0,1)

>>> name.random()
0.1786995275775844

 

.randint(begin_number, end_number):

return random int in [a,b)

>>> name.randin(1,10)
9

 

.shuffle(sequence):

shuffle the input sequence

>>> name.shuffle([1,2,3,4,5])
[2, 1, 5, 3, 4]

 

.choice(sequence, replace=True, size=1):

choice an element randomly in the sequence.

replace: choose with replacement or not.

size: the number of element to be chosen, if size != 1, will return a list contains those element.

>>> name.choice([1,2,3,4,5])
1
>>> name.choice([1,2,3,4,5],size=3)
[2, 3, 2]
>>> name.choice([1,2,3,4,5],replace=False,size=3)
[2, 5, 1]

 

.bern(p):

generate a Bernoulli Random Variable

p: the probability of True

>>> name.bern(0.5)
True
>>> name.bern(0.5)
False

 

.binomial(n, p):

generate a Binomial Random Variable

n: total times

p: probability of success

>>> name.binomial(10, 0.5)
6
>>> name.binomial(10, 0.5)
3

 

.geometric(p):

generate a Geometric Random Variable

p: probability of success

>>> name.geometric(0.5)
1
>>> name.geometric(0.5)
2

Randomness Testing

The file Testing for Randomness.ipynb contains several basic randomness testing result for this algorithm.