def __init__(self, x, y):
# call init of parent class
pygame.sprite.Sprite.__init__(self)
# init graphics with object's sprite - do not touch!
init_graphics(self, x, y, "wall")
# real coordinates of object
self.x = x
self.y = y
def __init__(self, x, y):
# call init of parent class
pygame.sprite.Sprite.__init__(self)
# init graphics with object's sprite - do not touch!
init_graphics(self, x, y, "dinner_table_active")
# real coordinates of object
self.x = x
self.y = y
# states of table
self.state = 1
def __init__(self, x, y):
# call init of parent class
pygame.sprite.Sprite.__init__(self)
# init graphics with object's sprite - do not touch!
init_graphics(self, x, y, "furnace_active")
# real coordinates of object
self.x = x
self.y = y
# lists with data
self.orderedDishes = {}
self.readyDishes = {}
# how long does this furnace cook?
# for ai learning purpose - waiter has to minimize time in restaurant
self.state = 1
def activated(self):
# serve object:
# init graphics with object's sprite - do not touch!
init_graphics(self, self.x, self.y, "furnace")
self.state = 0
def __init__(self, n, matrix_fields, num_tables, num_furnaces, num_walls, solving_method):
print("Agent: initializing object...")
# call init of parent class
pygame.sprite.Sprite.__init__(self)
# check if there is enough space for everyting in simulation
if num_tables + num_furnaces + num_walls + 1 > n*n:
print("Agent: Not enough space in restaurant for objects!")
sys.exit("N-space overflow")
self.n = n
# init restaurant - matrix of objects
self.restaurant = Matrix(n, n)
# set random coordinates of agent
self.x, self.y = matrix_fields[0][0], matrix_fields[0][1]
# init graphics with object's sprite - do not touch!
init_graphics(self, self.x, self.y, "waiter")
# add objects to restaurant - creates tables and furnaces basing on random positions in the matrix
# objects have coordinates like in matrix (0..n, 0..n):
# add ghostwaiter to restaurant to mark waiters position
self.restaurant.insert('W', self.x, self.y)
# counter counts number of used coordinates, so no object will occupy the same space in simulation
counter = 1
# add tables
for i in range(num_tables):
self.restaurant.simple_insert(DinningTable(matrix_fields[i + counter][0], matrix_fields[i + counter][1]))
# increase counter with number of used coordinates
counter += num_tables
# add furnaces
for i in range(num_furnaces):
self.restaurant.simple_insert(Furnace(matrix_fields[i + counter][0], matrix_fields[i + counter][1]))
# increase counter with number of used coordinates
counter += num_furnaces
# add furnaces
for i in range(num_walls):
self.restaurant.simple_insert(Wall(matrix_fields[i + counter][0], matrix_fields[i + counter][1]))
# calculate graph
self.graph = self.restaurant.to_graph()
self.graph2 = self.restaurant.to_graph_visited_or_not()
# set list of objects
self.objects_coordinates = matrix_fields[1:counter]
# set list of goals
self.goals = self.objects_coordinates[:]
# set permutations of goals
self.goalsPer = list(map(list, list(itertools.permutations(self.goals[:]))))
# set list of solutions
self.solutions = []
# set path
self.path = []
# set all available solving methods names
self.available_methods = ['depthfs', 'breadthfs', 'bestfs']
self.unsupervised_learning = ['rabbit', 'svm', 'dtree', 'lreg']
# set unsupervised learning safety switch
self.moves_queue = [[0, 0], [0, 0], [0, 0], [0, 0]]
# set neighbourhood
self.neighbourhood = []
self.neighbourhood_size = 5
# set solving method
self.solving_method = solving_method
# svm model variable
self.svm_data = []
self.svm_target = []
if self.solving_method == 'svm':
self.init_svm()
elif self.solving_method == "dtree":
self.init_dtree()
# run solution seeking
self.solve(self.solving_method)
self.control = True
# add steps counter
self.steps_count = 0
print("Agent: initialization completed.")