haris989 / nsga-ii Goto Github PK

Hello, I test the code and compare the results with R. Why is the number of result values different from R's result? I am a beginner to the NSGA. I appreciate if anyone has idea about it. Many thanks.

Hello Haris,
First of all thank you very much for sharing your work! I have some questions regarding the code.

1. In line 136, the output of non_dominated_sorted_solution2_1 will always be 0,1,2... for every i, so it seems that the code is meaningless. I think what we should do is to align crowding distance and front order, which seems not to appear in the code.

The code below is the original code which seems not to achieve the target aligning crowding distance and front order.

for i in range(0,len(non_dominated_sorted_solution2)):
non_dominated_sorted_solution2_1 = [index_of(non_dominated_sorted_solution2[i][j], non_dominated_sorted_solution2[i] ) for j in range(0,len(non_dominated_sorted_solution2[i]))]

2. In line 22, the definition of index_of(a,list) is not rigorous. If there are two same values which equal to a in the list, the function only outputs the first position. Thus, using the list.index(a) perhaps is a better choice.

def index_of(a,list):
    for i in range(0,len(list)):
        if list[i] == a:
            return i
    return -1

3. The operations of mutation and crossover are strange, because it seems the parent is not correlated with children at all. From the code, we can see no matter which the selected individuals a, b are, the crossover will give a new random solution.

def crossover(a,b):
    r=random.random()
    if r>0.5:
        return mutation((a+b)/2)
    else:
        return mutation((a-b)/2)

def mutation(solution):
    mutation_prob = random.random()
    if mutation_prob <1:
        solution = min_x+(max_x-min_x)*random.random()
    return solution

4. The crowding distance was also miscalculated and should have been calculated for each target and then added up, which has mentioned by other many times.

There is an error in Crowding Distance function.
It should be values1[i-1] - values1[i+i]
and values2[i-1] - values2[i+1]

Hello Haris,

First of all thank you very much for sharing this work. I have a question regarding mutation prob. You wrote if mutation_prob < 1 but as random.random() outputs values between 0 and 1, should we make it less than 1 ? and I have a question regarding multi variate objective function, how would you define the solutions and the mutation ? i did it this way:

def mutation(solution):
    mutation_prob = random.random()
    if mutation_prob <0.5:
        solution = [min_x+(max_x-min_x)*random.random(), min_x+(max_x-min_x)*random.random()]
    return solution

solution=[[min_x+(max_x-min_x)*random.random(),min_x+(max_x-min_x)*random.random()]  for i in range(0,pop_size)]

thanks

Hey,

I just remarked, that to plot all results for function1 and function2 are multplied by -1 and I can't understand why.

Cheers
Fabi

#Function to sort by values
def sort_by_values(list1, values):
    sorted_list = []
    while(len(sorted_list)!=len(list1)):
        if index_of(min(values),values) in list1:
            sorted_list.append(index_of(min(values),values))
        values[index_of(min(values),values)] = math.inf
    return sorted_list

Why do you calculate the index of the minimum value of a list to calculate three times?
And does this function not generate a return value the same as list1?