diff_prob.py

import numpy as np
from scipy.stats import gaussian_kde


def get_cdf_of_difference(data_1, data_2, diff_value=0):
    data_1 = np.asarray(data_1)
    data_2 = np.asarray(data_2)
    kde_1 = gaussian_kde(data_1)
    data_1.sort()
    data_2.sort()
    cdf_diff = list()
    for i, x in enumerate(data_1):
        print(" {}th point is {}".format(i, x))
        cdf_2 = float(data_2[data_2 <= x].size) / data_2.size
        if cdf_2 <= 0:
            print("Skipping because at {} CDF of data_2 is 0.0".format(x))
            continue
        print("CDF of data_2 is {}".format(cdf_2))
        kde_1_x = kde_1.pdf(diff_value+x)[0]
        print("KDE of data_1 is {}".format(kde_1_x))
        print("ADDING {}".format(cdf_2*kde_1_x))
        cdf_diff.append(cdf_2*kde_1_x)

    return sum(cdf_diff)