Zhang Ziqi

zzq.py

import numpy as np
import math

def process_number(s):
    s = s.replace(',', '')
    if s[-1] == '\n':
        s = s[:-1]
    if s[-1] == '%':
        return float(s[:-1]) * 0.01
    return float(s)

def read_data():
    rows = []
    with open('data.txt') as f:
        for line in f.readlines():
            rows.append([process_number(s) for s in line.split('\t')])
    return np.array(rows).T

if __name__ == '__main__':
    np.set_printoptions(precision=3)
    data = read_data()
    x = data[1:]
    print('Data', x)
    mu = (data - np.min(data, axis=0)[np.newaxis, :]) / (np.max(data, axis=0) - np.min(data, axis=0))[np.newaxis, :]
    mu0 = mu[0]
    mu = mu[1:]
    p = x / np.sum(x, axis=0)
    K = 1 / np.log(x.shape[0])
    E = - K * np.sum(p * np.log(p), axis=0)
    print('Entropy of each index:\n', E)
    w = E / np.sum(E)
    print('Weight of each index:\n', w)
    pH = 1 - np.sum(w * np.abs(mu - mu0[np.newaxis, :]), axis=1)
    W = pH / np.sum(pH)
    print(list(zip(*sorted(list(enumerate(W)), key=lambda x: x[1]))))