Varying Noise-Weight.py

import Max_Sat_interface as mxs
import group_testing_function as gtf
import numpy as np
import scipy.spatial.distance as sd
import matplotlib.pyplot as plt
n = 100


noiseless = False

class trial:
    def __init__(self, var):
        self.E = [] #on number of tests
        self.P = []
        self.t_e = [] #temporary error
        self.t_h = [] #temporary hamming distance
        self.var = var

def main():
    x, k = gtf.generate_input(n, 5/n)


    #avoid k = 0, trivial example
    while k != 4:
        x, k = gtf.generate_input(n, 5 / n)


    noise_probability = 0.05

    T = [75]#number of tests


    trials = []

    noise_weights = [0.7, 0.74, 0.76, 0.8, 0.82, 0.84, 0.88, 0.9, 0.95, 1 ,1.02, 1.04, 1.1, 1.15, 1.2]

    for i in T:
        trials.append(trial(i))

    #for every t number of tests
    for w in noise_weights:
        print(w)

        for tr in trials:
            tr.t_e = [] #blank temporary
            tr.t_h = [] #blank temporary

        for i in range(300):
            print(i)

            for tr in trials:
                a = gtf.generate_pool_matrix(n, k, tr.var)
                y = gtf.get_results(tr.var, a, x, noiseless, noise_probability)
                mxs.output(n, tr.var , x, y, a, noiseless, w)

                r, noise = mxs.call_Max_Sat(n)

                #calculating hamming distance between model result and input x

                hamming_distance = (n)*sd.hamming(x,r)
                tr.t_h.append(hamming_distance)

                #there's an error?
                if hamming_distance > 0:
                    tr.t_e.append(1)
                else:
                    tr.t_e.append(0)

        for tr in trials:
            tr.E.append(np.mean(tr.t_h))
            tr.P.append(1 - np.mean(tr.t_e))

    X = []

    for tr in trials:

        plt.plot(noise_weights, tr.P, "bo")
        plt.plot(noise_weights, tr.P, "g")
        plt.title("Error trend of Max_Sat applied to Group Testing with  n = " + str(n) + " k = " + str(k))
        plt.xlabel("weight to the noise")
        plt.ylabel("Probability of success")
        plt.savefig("PS, n = " + str(n) + " k = " + str(k) + "nT = " + str(tr.var) + ".png")#todo complete
        plt.show()

        plt.plot(noise_weights, tr.E, "bo")
        plt.plot(noise_weights, tr.E, "g")
        plt.title("Error trend of Max_Sat applied to Group Testing with  n = " + str(n) + " k = " + str(k))
        plt.xlabel("Number of tests t")
        plt.ylabel("Error (Hamming Distance)")
        plt.savefig("HD, n = " + str(n) + " k = " + str(k) + "nT = " + str(tr.var) + ".png")  # todo complete
        plt.show()



main()