CODEWITHSUNDEEP
python
Zzyy
Zzyy

Reputation: 25

How to avoid select two identical parents in genetic algorithm?

I'm new to genetic algorithm and trying to find a way avoid selecting two same parents, but I am failed to do that.

Assume that the initial popuation is

[[0,0,0,0,1],[0,0,0,1,0],[0,0,1,0,0],[[0,1,0,0,0].

Assume that the fitness for each individual is [2,3,4,5].

The code used to pick parent is shown below:

def select_individual(population, scores, k=4):
    first_pick = np.random.randint(0,len(population))  
    for second_pick in np.random.randint(0,len(population),k-1):
        if scores[second_pick] < scores[first_pick]:
           winner = second_pick
        else:
           winner = first_pick
        return population[winner,:]

When I tried to used the code shown below to retrieve two parents, it returned two identical parents sometimes.

for i in range(0, len(population),2):
    # Selection
    parents = [select_individual(population, scores) for _ in range(len(population))]
    parent_1 = parents [i]
    parent_2 = parents [i+1]
    
    print('Parent_1: ',np.round(parent_1,2))
    print('Parent_2: ',np.round(parent_2,2))

One of the output, for example, is shown below:

        Parent_1: [0,0,0,1,0]
        Parent_2: [0,0,0,1,0]
        Parent_1: [0,0,1,0,0]
        Parent_2: [0,0,1,0,0]
Or
        Parent_1: [0,0,0,1,0]
        Parent_2: [0,0,1,0,0]
        Parent_1: [0,0,0,1,0]
        Parent_2: [0,0,1,0,0]

Could you please tell me that what's wrong with the code and why two selected parents are identical finally? Thank you very much!

Upvotes: 1

Views: 56

Answers (1)

Sokratis Kouvaras
Sokratis Kouvaras

Reputation: 36

In select_individual() after you have selected your first pick you iterate over the same list to select your second pick. You could create a list where you have excluded the first_pick from 'population' and iterate over that instead.

Upvotes: 1

Related Questions