Hands-On Neuroevolution with Python

Visual discriminator environment

The visual discriminator environment holds the generated dataset with the visual field definitions. Also, it provides methods to create the dataset and to evaluate the discriminator ANN against the dataset. The VDEnvironment Python class holds the definitions of all mentioned methods, as well as related data structures. Next, we'll look at all the significant parts of the VDEnvironment class definition:

• The class constructor is defined as follows:

    def __init__(self, small_object_positions, big_object_offset,                  field_size):        self.s_object_pos = small_object_positions        self.data_set = []        self.b_object_offset = big_object_offset        self.field_size = field_size        self.max_dist = self._distance((0, 0),  (field_size ...