def get_radius(self):
# faster BBs can move with greater radius
# return self._spd * 1.0 / MAX_ATTRIBUTE * MAX_RADIUS - self._image_surf_rect.width
# TODO: need to fix this!
return calc_distance(self._image_surf_rect.centerx,
self._image_surf_rect.centery, int(WIDTH / 2.0),
int(HEIGHT / 2.0))
def check_obstacles(self, vehicles):
minimumDistance = 1000
for vehicle in vehicles:
if self.vehicle_type == "Pedrestian": # and vehicle.vehicle_type == "Pedrestian":
continue
elif vehicle.RightOfPassage == 0 and self.RightOfPassage == 1 and self.vehicle_type != "Pedrestian":
# The vehicle doesn't have right of passage (not in roundabout)
continue
elif self.position == vehicle.position:
# It's our vehicle
continue
else:
distance = utils.calc_distance(self.position, vehicle.position)
angle = abs(utils.calc_angle(self.position, vehicle.position) - self.direction)
current_vision = self.vision_angle
if (
self.RightOfPassage == 0
and self.velocity < 0.75 * self.max_velocity
and vehicle.RightOfPassage == 1
):
current_vision = self.vision_angle_entrance
if distance <= self.range_of_sight and angle < current_vision:
stopDistance = (
utils.calc_stopDistance(distance, angle) - self.length * 2 - vehicle.length
) # 3 times radius
if stopDistance < minimumDistance:
minimumDistance = stopDistance
return minimumDistance
def recommendation():
json_data = request.get_json(silent=True)
df_hotels_return = df_hotels
df_tour_return = df_tours
if "city_id" in json_data:
df_hotels_return = df_hotels_return[df_hotels_return["city_id"]
== int(json_data["city_id"])]
df_tour_return = df_tour_return[df_tour_return["city_id"]
== int(json_data["city_id"])]
if "type_price_hotel" in json_data and json_data["type_price_hotel"] != "-1":
df_hotels_return = filter_by_price_hotel(
df_hotels_return, type_price=json_data["type_price_hotel"])
if "type_price_tour" in json_data and json_data["type_price_tour"] != "-1":
df_tour_return = filter_by_price_tour(
df_tour_return, type_price=json_data["type_price_tour"])
center_location = LIST_CENTER_CITY[json_data["city_id"]]
if len(df_hotels_return) > 0:
df_hotels_return["distance_calc"] = df_hotels_return["distance"].apply(
lambda x: calc_distance(x, center_location))
prob_hotels, hotel_idxes = ranking_topsis(
df_hotels_return, type_topsis="hotel")
df_hotels_return = df_hotels_return.iloc[hotel_idxes]
if len(df_tour_return) > 0:
prob_tours, tour_idxes = ranking_topsis(
df_tour_return, type_topsis="tour")
df_tour_return = df_tour_return.iloc[tour_idxes]
return jsonify({"hotels": eval(df_hotels_return.to_json(orient="records")), "tours": eval(df_tour_return.to_json(orient="records"))})
def getNearestMainNode(pos, node_list):
min_dist = 1000
for node in node_list:
dist = utils.calc_distance(pos, node)
if dist < min_dist:
min_dist = dist
end_pos = node
return (pos,end_pos)
def getNearestNodeInList(pos, any_list):
min_dist = 1000
for node in any_list:
dist = utils.calc_distance(pos, node)
if dist < min_dist:
min_dist = dist
index = any_list.index(node)
return any_list[index]
def getNearestNodeIndex(pos, node_list):
min_dist = 1000
for node in node_list:
dist = utils.calc_distance(pos, node)
if dist < min_dist:
min_dist = dist
index = node_list.index(node)
return index
def getObservation(self):
sensors = Pa10Task.getObservation(self)
# Calculate the straightline distance from the arm to the target [m]
self.distance_to_target = utils.calc_distance(self.tooltip_position,
self.target_position)
return sensors
def GetDistanceToBusStop(self,position,direction):
deltaX=self.roads[2][0]-self.roads[1][0]
deltaY=self.roads[2][1]-self.roads[1][1]
busStopX=self.roads[1][0]+deltaX*0.5
busStopY=self.roads[1][1]+deltaY*0.5
dist=utils.calc_distance((busStopX,busStopY),position)
if (abs(utils.calc_angle(position, (busStopX,busStopY)) - direction)>math.pi/2):
dist=-dist
return dist
def calc_reward(self):
"""
Calculate the agent score, which is a linear combination of the distance between the agent and the target
and the time elapsed since the beginning of the simulation
:return: the reward
"""
r1 = -utils.calc_distance(self.agent_xyz, self.tgt_xyz)
r2 = -self.time * self.time_reward_factor
return r1 + r2
def test_works_as_expected(self):
'''utils.calc_distance correctly calculates the distance between two points'''
pairs = [[[0, 0], [1, 1]], [[0, 0], [0, 1]], [[5, 2], [2, 5]],
[[-1, 3], [10, 0]]]
for pair in pairs:
distance = utils.calc_distance(*pair)
expected_distance = math.sqrt((pair[1][0] - pair[0][0])**2 +
(pair[1][1] - pair[0][1])**2)
self.assertEqual(distance, expected_distance)
def getObservation(self):
sensors = Pa10Task.getObservation(self)
# Calculate the straightline distance from the arm to the target [m]
self.distance_to_target = utils.calc_distance(
self.tooltip_position,
self.target_position
)
return sensors
def recalculate_info(currentLat, currentLong, altitude):
"""
Recalculate target heading
"""
heading = utils.calc_heading(currentLat, currentLong, this.targetLat.get(),
this.targetLong.get())
overlay.show_heading(heading)
if this.planetRadius.get() > 0:
distance = utils.calc_distance(currentLat, currentLong,
this.targetLat.get(),
this.targetLong.get(),
this.planetRadius.get(), altitude)
overlay.show_distance(distance)
def knn(all_stations,
sugg_station,
nearest_stations=[],
nearest_stations_distances=None,
nearest_stations_names=None,
index=0):
nearest_station = ''
minimum_distance = float('inf')
for station in all_stations:
diff = calc_distance(station[1], sugg_station[1])
if diff < minimum_distance:
minimum_distance = diff
nearest_station = station[0]
nearest_stations.append((nearest_station, minimum_distance))
if nearest_stations_distances:
nearest_stations_distances[index] = minimum_distance
if nearest_stations_names:
nearest_stations_names[index] = nearest_station
return (nearest_station, minimum_distance)
def test_can_see_an_object_if_angle_does_not_pass_through_center_but_passes_through_some_of_object(
self):
'''if the angle of the eye and the angle between the snake's head and the object are not equal but close enough (depending on obj size & distance) then the eye can see the object'''
''' example: ( ) - object; x - snake;
( )
|
|
x
'''
expected_seen_object = MockObject([0.5, 5], 1, visual_encoding=[1, 1])
other_objects = [expected_seen_object]
output = self.snake.look(other_objects)
snake_head = self.snake.body[0]
expected_output = expected_seen_object.visual_encoding + [
utils.calc_distance(snake_head.position,
expected_seen_object.position)
]
self.assertListEqual(output, expected_output)
def __init__(self,
agent_xyz_init,
agent_acc,
tgt_xyz_init,
tgt_traj,
max_time,
time_reward_factor=0.01):
"""
Initialize the trajectory environment
:param agent_xyz_init: the initial agent position
:param agent_acc: the acceleration available to the agent
:param tgt_xyz_init: the target initial position
:param tgt_traj: function defining the target trajectory parametrically vs. time
:param max_time: the maximum trajectory time
"""
# define the initial separation between target/agent and how close they need to be to terminate the env
self.init_sep = utils.calc_distance(agent_xyz_init, tgt_xyz_init)
self.done_sep_threshold = 0.001
# set the time equal to 0, identify max time, and define the time step, which works out to 50 Hz if the time
# is in seconds
self.time = 0
self.time_step = 0.02
self.max_time = max_time
# the time factor is a weight that determines how much the agent cares about time progressing
# Higher scores <-> lower times
self.time_reward_factor = time_reward_factor
# the agent characteristics
self.agent_xyz = agent_xyz_init
self.agent_vxyz = np.zeros(3)
self.agent_axyz = np.zeros(3)
self.agent_acc = agent_acc
# the target characteristics
self.tgt_xyz = tgt_xyz_init
self.tgt_xyz_next = self.tgt_position_calc(self.time_step)
self.tgt_traj = tgt_traj
def evaluate_possible_movement_targets(self, possible_targets):
"""The value of a story location is highest when its distance from its author is 40,
so determine the value of all possible target locations accordingly."""
result = []
# find the goal location - near either the author of the story, or its recipient (in case there's no author)
if self.authors.all().count() > 0:
author = self.authors.all()[0]
goal_location = (author.x, author.y)
elif self.recipients.all().count() > 0:
recipient = self.recipients.all()[0]
goal_location = (recipient.x, recipient.y)
else:
goal_location = None
# move toward the goal location
if goal_location != None:
goal_location = near_by_location(goal_location, min_distance=20, max_distance=40, limit=LAYER_SIZE)
for loc in possible_targets:
value = -calc_distance(goal_location, loc)
result.append( (value, loc) )
else :
result = [ (0, i) for i in possible_targets]
return result
def step(self, thrust_vector):
"""
Move the agent forward one step forward in time
:param thrust_vector: the direction the agent moves
:return: the current state (the acceleration, agent position, target position, and next target position),
the reward the agent receives for its action, and whether or not the simulation is finished
"""
# get the unit vector representation of the thrust vector
thrust_vector = utils.norm_vector(thrust_vector)
# increment time by one step
self.time += self.time_step
# update the target position
self.tgt_xyz = self.tgt_xyz_next
self.tgt_xyz_next = self.tgt_position_calc()
# update the agent position
self.agent_position_update(thrust_vector)
# calculate the reward
reward = self.calc_reward()
# determine if the simulation is finished or not. It finishes if the distance between the two is less
# than the threshold or if the time is greater than the maximum time
if utils.calc_distance(self.agent_xyz, self.tgt_xyz) < self.done_sep_threshold * self.init_sep \
or self.time > self.max_time:
done = True
reward = 0
else:
done = False
# concatenate the important information to feed into the DQN for the next prediction
cur_state = np.concatenate(self.agent_acc, self.agent_xyz,
self.tgt_xyz, self.tgt_xyz_next)
return cur_state, reward, done
def location_analysis():
reviews_path = config.Project_CONFIG['user_folder_path']
city_values = []
state_values = []
distance_values = []
std_dist_values = []
std_dist_city_values = []
city_distance_values = []
b_data = read_business_info()
files = os.listdir(reviews_path)
for user_file in files:
if not os.path.isdir(user_file):
file_json_data = open(reviews_path + user_file,
'r',
encoding="utf8")
city_stat = {}
state_stat = {}
points = []
city_points = []
for line in enumerate(file_json_data):
temp_json_data = json.loads(line[1])
b_id = temp_json_data['business_id']
b_info = b_data[b_id]
city = b_info['city']
state = b_info['state']
lat = b_info['lat']
lon = b_info['lon']
points.append((lat, lon))
if city in city_stat:
city_stat[city] += 1
else:
city_stat[city] = 1
if state in state_stat:
state_stat[state] += 1
else:
state_stat[state] = 1
max_num = city_stat[max(city_stat, key=city_stat.get)]
city_num = sum(city_stat.values())
city_values.append(max_num * 1.00 / city_num)
file_json_data = open(reviews_path + user_file,
'r',
encoding="utf8")
city_name = max(city_stat, key=city_stat.get)
for line in enumerate(file_json_data):
temp_json_data = json.loads(line[1])
b_id = temp_json_data['business_id']
b_info = b_data[b_id]
if b_info['city'] == city_name:
lat = b_info['lat']
lon = b_info['lon']
city_points.append((lat, lon))
max_num_state = state_stat[max(state_stat, key=state_stat.get)]
state_num = sum(state_stat.values())
state_values.append(max_num_state * 1.00 / state_num)
x = [p[0] for p in points]
y = [p[1] for p in points]
centroid = (sum(x) / len(points), sum(y) / len(points))
distance = 0
dist_list = []
for p in points:
distance += calc_distance(centroid[0], centroid[1], p[0], p[1])
dist_list.append(
calc_distance(centroid[0], centroid[1], p[0], p[1]))
avg_distance = distance / len(points)
distance_values.append(avg_distance)
std_dist_values.append(np.std(dist_list, ddof=1))
c_x = [c_p[0] for c_p in city_points]
c_y = [c_p[1] for c_p in city_points]
c_centroid = (sum(c_x) / len(city_points),
sum(c_y) / len(city_points))
c_distance = 0
c_dist_list = []
for c_p in city_points:
c_distance += calc_distance(c_centroid[0], c_centroid[1],
c_p[0], c_p[1])
c_dist_list.append(
calc_distance(c_centroid[0], c_centroid[1], c_p[0],
c_p[1]))
c_avg_distance = c_distance / len(city_points)
city_distance_values.append(c_avg_distance)
std_dist_city_values.append(np.std(c_dist_list, ddof=1))
print(city_values) # the proportion that the business is in the same city
print(
state_values) # the proportion that the business is in the same state
print(distance_values) # business distance
print(std_dist_values) # business distance standard deviation
print(city_distance_values) # business distance in the same city
print(std_dist_city_values
) # business distance in the same city standard deviation
def update_next_node(self,delta_t):
next_pos = self.road.GetNodePosition(self.nextNode)
if utils.calc_distance(self.position, next_pos) < self.velocity*delta_t: #We arrive at the next node
self.nextNode=self.road.GetNextNode(self.nextNode)
if self.nextNode == -1:
self.active=False
def touches(self, other):
return utils.calc_distance(self.x, self.y, other.x, other.y) <= 27
else:
print "Main bus road needed"
elif state == 2 and BusRoadStates == 3:
if len(buses['Main']) != 0:
buses['End'].append((event.pos, getNearestNodeIndex(event.pos, buses['Main'])))
busExit_list.append(getNearestMainNode(event.pos, buses['Main']))
else:
print "Main bus road needed"
elif state == 3:
alter = 0
dist = 1000
if len(roads['Main']) != 0:
nearestMainNode = getNearestNodeInList(event.pos, roads['Main'])
distMainNode = utils.calc_distance(event.pos, nearestMainNode)
if distMainNode < dist:
dist = distMainNode
alter = 1
if len(roads['Start']) != 0:
nearestStartNode = getNearestNodeInList(event.pos, [x[0] for x in roads['Start']])
distStartNode = utils.calc_distance(event.pos, nearestStartNode)
if distStartNode < dist:
dist = distStartNode
alter = 2
if len(roads['End']) != 0:
nearestEndNode = getNearestNodeInList(event.pos, [x[0] for x in roads['End']])
distEndNode = utils.calc_distance(event.pos, nearestEndNode)
if distEndNode < dist:
dist = distEndNode
