93cbc631abd8671e5fe3d6a9360f49426ffa8e58,deeppavlov/models/evolution/evolution_param_generator.py,ParamsEvolution,crossover,#ParamsEvolution#Any#Any#,300

Before Change

        
        offsprings = []
        scores = np.array(scores, dtype="float")
        if np.sum(scores) < self.eps:
            scores = [self.eps for _ in range(self.population_size)]

        probas_to_be_parent = scores / np.sum(scores)
        intervals = np.array([np.sum(probas_to_be_parent[:i]) for i in range(self.population_size)])

        for i in range(self.population_size - self.n_saved_best_pretrained):

After Change

        ranges = self.range_scores(scores)
        a = 1. / (1. - self.population_size)
        b = self.population_size / (self.population_size - 1.)
        probas_to_be_parent = (a * ranges + b) / np.sum(a * ranges + b)
        intervals = np.array([np.sum(probas_to_be_parent[:i]) for i in range(self.population_size)])

        for i in range(self.population_size - self.n_saved_best_pretrained):
            rs = np.random.random(2)

Instances