A Markov Chain Implementation to create a controller priority list

markov_chain.py

#!/usr/bin/python
"""
This is a program that implements the Markov Chain algorithm 
to calculate the next best possible controller to contact for 
executing the source and destination flow modification.
"""

import scipy.linalg
import numpy as np


# Random Walk
def random_walk(state, A):
    n = 10
    start_state = 0
    prev_state = start_state

    while n>1:
        curr_state = np.random.choice([0, 1, 2], p=A[prev_state])
        prev_state = curr_state
        n -= 1

    values, left = scipy.linalg.eig(A, right=False, left=True)

    # Normalized values of the eigen vectors
    pi = left[:,0]
    pi_normalized = [(x/np.sum(pi)).real for x in pi]
    return pi_normalized

def controller_priority():
    state = {
        0: "ODL",
        1: "Ryu",
        2: "FL"
    }

    A = np.array([[0.2, 0.15, 0.65], [0.7, 0.2, 0.1], [0.9, 0.05, 0.05]])
    stats = random_walk(state, A)
    print(stats)
    res_controllers = [state[stats.index(x)] for x in sorted(stats)][::-1]
    return res_controllers

if __name__ == "__main__":
    controller_priority()