Self treated like a normal argument

So, I have the following class:

class GeneticAlgorithm():
    def breed_population(self, agents, fitness=0.10, crossover_rate=0.50, mutation_rate=0.10, mutation_degree=0.10, mutate=True):
        '''
        Crossover the weights and biases of the fittest members of the population,
        then randomly mutate weights and biases.
        '''
        # Sort by highest to lowest score
        agents.sort(key=operator.attrgetter("top_score"))

        # Get the number of breeding agents
        pop_size = len(agents)
        cutoff = (int)(fitness * pop_size)
        if cutoff < 2: return agents

        # Get number of times each parent pair needs to breed
        num_children = pop_size // cutoff

        # Initialize children
        children = [Agent() for i in range(pop_size)]

        # Breed population
        for i in range(0, cutoff, 2):
            for c in range(num_children):
                children[i*c] = self.crossover(children[i*c], agents[i], agents[i+1], crossover_rate, mutation_rate, mutation_degree, mutate)

        return children


    def crossover(self, child, parent_one, parent_two, crossover_rate, mutation_rate, mutation_degree, mutate):
        ''' Apply crossover and mutation between two parents in order to get a child. '''
        # Crossover and mutate each layer
        for i in range(len(parent_one.layers)):
            # Get weights
            p1_weights = parent_one.layers[i].get_weights()[0]
            p2_weights = parent_two.layers[i].get_weights()[0]
            
            # Cycle through layer's weights
            for x, row in enumerate(p1_weights):
                for y, _ in enumerate(row):
                    # Apply crossover
                    if (random.random() < crossover_rate):
                        p1_weights[x][y] = p2_weights[x][y]
                    
                    # Apply mutation
                    if mutate:
                        if (random.random() < mutation_rate):
                            if (random.random() > 0.50):
                                p1_weights[x][y] += p1_weights[x][y] * mutation_degree
                            else:
                                p1_weights[x][y] -= p1_weights[x][y] * mutation_degree
            
            # Set weights in child
            child.layers[i].set_weights(p1_weights)
        return child

When I try to run it, I get this error message:

agents = GeneticAlgorithm.breed_population(agents)
TypeError: breed_population() missing 1 required positional argument: 'agents'

Meaning that, for some reason, self is being treated like an actual argument instead of just giving the class access to all functions within the class.

How do I fix this?

>Solution :

The problem you are having is that you trying to call the method on the class without creating an object.

agents = GeneticAlgorithm.breed_population(agents)

Should be:

ga_obj = GeneticAlgorithm()
agents = ga_obj.breed_population(agents)