Simulated Annealing

def SA...

def generate_random_state():
    state = [0 for i in range(n//2)] + [1 for i in range(n-(n//2))]
    return random_shuffle(state)

def action(state):
    # 1/2 probability of success
    # u = randint(0, n-1)
    # v = randint(0, n-1)
    # state[u], state[v] = state[v], state[u]

    # 100% success
    ind0 = [i for i in range(n) if state[i] == 0]
    ind1 = [i for i in range(n) if state[i] == 1]
    u = randint(0, len(ind0)-1)
    v = randint(0, len(ind1)-1)
    state[ind0[u]], state[ind1[v]] = state[ind1[v]], state[ind0[u]]
    return state

def h(state):
    sumA = sum([X[i] for i in range(n) if state[i] == 1]) # sum all elements in A
    sumB = sum([X[i] for i in range(n) if state[i] == 0]) # sum all elements in B
    return abs(sumA - sumB)