general_maxsat.py

import numpy as np


def ENN_maxsat(I, n, m):
    D1 = np.zeros(n)
    for i in range(n):
        if np.any(I[2:, i] != 0):
            D1[i] = -1
        else:
            D1[i] = 1

    D2 = np.zeros(n)
    for i in range(n):
        if np.any(I[2:, i] != 0):
            D2[i] = 1
        else:
            D2[i] = -1

    D3 = np.zeros(n)
    for i in range(n):
        col_mean = np.mean(I[2:, i])
        if I[1, i] + col_mean - I[0, i] > 0:
            D3[i] = 1
        else:
            D3[i] = -1

    D4 = np.zeros(n)
    for i in range(n):
        col_mean = np.mean(I[2:, i])
        if I[0, i] + col_mean - I[1, i] > 0:
            D4[i] = 1
        else:
            D4[i] = -1

    D5 = np.zeros(n)
    for i in range(n):
        if I[0, i]:
            D5[i] = 1
        elif (not I[1, 0]) or I[0, i]:
            D5[i] = -1

    D6 = np.zeros(n)
    for i in range(n):
        if I[1, i]:
            D6[i] = 1
        elif (not I[0, 0]) or I[1, i]:
            D6[i] = -1

    S1 = np.zeros(n)
    for i in range(n):
        S1[i] = D6[i] > 0 and D1[i] > 0

    S2 = np.zeros(n)
    for i in range(n):
        S2[i] = D2[i] > 0 and D3[i] > 0

    S3 = np.zeros(n)
    for i in range(n):
        S3[i] = D5[i] > 0 and D1[i] > 0

    S4 = np.zeros(n)
    for i in range(n):
        S4[i] = D2[i] > 0 and D4[i] > 0

    C1 = 10.0 * np.sum(S3) + 2.298 * np.sum(S4) - 2.298 * np.sum(S2) - 10.0 * np.sum(S1)
    C2 = 10.0 * np.sum(S1) + 2.298 * np.sum(S2) - 2.298 * np.sum(S4) - 10.0 * np.sum(S3)
    C = [C1, C2]

    return np.exp(C) / np.sum(np.exp(C))