-
Notifications
You must be signed in to change notification settings - Fork 30
/
Copy pathcommon_utils.py
executable file
·358 lines (312 loc) · 22.7 KB
/
common_utils.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
import time
import pathlib
import pandas as pd
from nuplan.planning.metrics.metric_engine import MetricsEngine
from nuplan.common.actor_state.vehicle_parameters import get_pacifica_parameters
from nuplan.planning.metrics.aggregator.weighted_average_metric_aggregator import WeightedAverageMetricAggregator
from nuplan.planning.metrics.evaluation_metrics.common.drivable_area_compliance import DrivableAreaComplianceStatistics
from nuplan.planning.metrics.evaluation_metrics.common.driving_direction_compliance import DrivingDirectionComplianceStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_lane_change import EgoLaneChangeStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_is_comfortable import EgoIsComfortableStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_acceleration import EgoAccelerationStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_expert_l2_error import EgoExpertL2ErrorStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_expert_l2_error_with_yaw import EgoExpertL2ErrorWithYawStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_jerk import EgoJerkStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_lat_acceleration import EgoLatAccelerationStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_lat_jerk import EgoLatJerkStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_lon_acceleration import EgoLonAccelerationStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_lon_jerk import EgoLonJerkStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_mean_speed import EgoMeanSpeedStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_progress_along_expert_route import EgoProgressAlongExpertRouteStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_yaw_acceleration import EgoYawAccelerationStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_yaw_rate import EgoYawRateStatistics
from nuplan.planning.metrics.evaluation_metrics.common.planner_expert_average_l2_error_within_bound import PlannerExpertAverageL2ErrorStatistics
from nuplan.planning.metrics.evaluation_metrics.common.ego_is_making_progress import EgoIsMakingProgressStatistics
from nuplan.planning.metrics.evaluation_metrics.common.no_ego_at_fault_collisions import EgoAtFaultCollisionStatistics
from nuplan.planning.metrics.evaluation_metrics.common.planner_expert_average_heading_error_within_bound import PlannerExpertAverageHeadingErrorStatistics
from nuplan.planning.metrics.evaluation_metrics.common.planner_expert_final_heading_error_within_bound import PlannerExpertFinalHeadingErrorStatistics
from nuplan.planning.metrics.evaluation_metrics.common.planner_expert_final_l2_error_within_bound import PlannerExpertFinalL2ErrorStatistics
from nuplan.planning.metrics.evaluation_metrics.common.planner_miss_rate_within_bound import PlannerMissRateStatistics
from nuplan.planning.metrics.evaluation_metrics.common.speed_limit_compliance import SpeedLimitComplianceStatistics
from nuplan.planning.metrics.evaluation_metrics.common.time_to_collision_within_bound import TimeToCollisionStatistics
### Parameters
T = 8 # [s] planning horizon
DT = 0.1 # [s] time interval
LENGTH = get_pacifica_parameters().front_length # [m] vehicle front length
WHEEL_BASE = get_pacifica_parameters().wheel_base # [m] vehicle wheel base
WIDTH = get_pacifica_parameters().width # [m] vehicle width
MAX_LEN = 120 # [m] max length of the path
### Simulation setting
def save_runner_reports(reports, output_dir, report_name):
"""
Save runner reports to a parquet file in the output directory.
:param reports: Runner reports returned from each simulation.
:param output_dir: Output directory to save the report.
:param report_name: Report name.
"""
report_dicts = []
for report in map(lambda x: x.__dict__, reports): # type: ignore
if (planner_report := report["planner_report"]) is not None:
planner_report_statistics = planner_report.compute_summary_statistics()
del report["planner_report"]
report.update(planner_report_statistics)
report_dicts.append(report)
df = pd.DataFrame(report_dicts)
df['duration'] = df['end_time'] - df['start_time']
save_path = pathlib.Path(output_dir) / report_name
df.to_parquet(save_path)
print(f'Saved runner reports to {save_path}')
def build_metrics_aggregators(experiment, output_dir, aggregator_metric_dir):
"""
Build a list of metric aggregators.
:param cfg: Config
:return A list of metric aggregators, and the path in which they will save the results
"""
aggregator_save_path = f"{output_dir}/{aggregator_metric_dir}"
aggregator_save_path = pathlib.Path(aggregator_save_path)
metric_aggregators = []
metric_aggregator_config = get_aggregator_config(experiment)
if not aggregator_save_path.exists():
aggregator_save_path.mkdir(exist_ok=True, parents=True)
name = metric_aggregator_config[0]
metric_weights = metric_aggregator_config[1]
file_name = metric_aggregator_config[2]
multiple_metrics = metric_aggregator_config[3]
metric_aggregators.append(WeightedAverageMetricAggregator(name, metric_weights, file_name, aggregator_save_path, multiple_metrics))
return metric_aggregators
def get_aggregator_config(experiment):
if experiment == 'open_loop_boxes':
name = 'open_loop_boxes_weighted_average'
metric_weights = {'planner_expert_average_l2_error_within_bound': 1,
'planner_expert_average_heading_error_within_bound': 2,
'planner_expert_final_l2_error_within_bound': 1,
'planner_expert_final_heading_error_within_bound': 2,
'default': 1.0}
file_name = "open_loop_boxes_weighted_average_metrics"
multiple_metrics = ['planner_miss_rate_within_bound']
challenge_name = 'open_loop_boxes'
elif experiment == 'closed_loop_nonreactive_agents':
name = 'closed_loop_nonreactive_agents_weighted_average'
metric_weights = {'ego_progress_along_expert_route': 5.0,
'time_to_collision_within_bound': 5.0,
'speed_limit_compliance': 4.0,
'ego_is_comfortable': 2.0,
'default': 1.0}
file_name = "closed_loop_agents_weighted_average_metrics"
multiple_metrics = ['no_ego_at_fault_collisions', 'drivable_area_compliance',
'ego_is_making_progress', 'driving_direction_compliance']
challenge_name = 'closed_loop_nonreactive_agents'
elif experiment == 'closed_loop_reactive_agents':
name = 'closed_loop_reactive_agents_weighted_average'
metric_weights = {'ego_progress_along_expert_route': 5.0,
'time_to_collision_within_bound': 5.0,
'speed_limit_compliance': 4.0,
'ego_is_comfortable': 2.0,
'default': 1.0}
file_name = "closed_loop_agents_weighted_average_metrics"
multiple_metrics = ['no_ego_at_fault_collisions', 'drivable_area_compliance',
'ego_is_making_progress', 'driving_direction_compliance']
challenge_name = 'closed_loop_reactive_agents'
else:
raise TypeError("Experiment type not supported!")
return name, metric_weights, file_name, multiple_metrics, challenge_name
def get_scenario_map():
scenario_map = {
'accelerating_at_crosswalk': [15.0, -3.0],
'accelerating_at_stop_sign': [15.0, -3.0],
'accelerating_at_stop_sign_no_crosswalk': [15.0, -3.0],
'accelerating_at_traffic_light': [15.0, -3.0],
'accelerating_at_traffic_light_with_lead': [15.0, -3.0],
'accelerating_at_traffic_light_without_lead': [15.0, -3.0],
'behind_bike': [15.0, -3.0],
'behind_long_vehicle': [15.0, -3.0],
'behind_pedestrian_on_driveable': [15.0, -3.0],
'behind_pedestrian_on_pickup_dropoff': [15.0, -3.0],
'changing_lane': [15.0, -3.0],
'changing_lane_to_left': [15.0, -3.0],
'changing_lane_to_right': [15.0, -3.0],
'changing_lane_with_lead': [15.0, -3.0],
'changing_lane_with_trail': [15.0, -3.0],
'crossed_by_bike': [15.0, -3.0],
'crossed_by_vehicle': [15.0, -3.0],
'following_lane_with_lead': [15.0, -3.0],
'following_lane_with_slow_lead': [15.0, -3.0],
'following_lane_without_lead': [15.0, -3.0],
'high_lateral_acceleration': [15.0, -3.0],
'high_magnitude_jerk': [15.0, -3.0],
'high_magnitude_speed': [15.0, -3.0],
'low_magnitude_speed': [15.0, -3.0],
'medium_magnitude_speed': [15.0, -3.0],
'near_barrier_on_driveable': [15.0, -3.0],
'near_construction_zone_sign': [15.0, -3.0],
'near_high_speed_vehicle': [15.0, -3.0],
'near_long_vehicle': [15.0, -3.0],
'near_multiple_bikes': [15.0, -3.0],
'near_multiple_pedestrians': [15.0, -3.0],
'near_multiple_vehicles': [15.0, -3.0],
'near_pedestrian_at_pickup_dropoff': [15.0, -3.0],
'near_pedestrian_on_crosswalk': [15.0, -3.0],
'near_pedestrian_on_crosswalk_with_ego': [15.0, -3.0],
'near_trafficcone_on_driveable': [15.0, -3.0],
'on_all_way_stop_intersection': [15.0, -3.0],
'on_carpark': [15.0, -3.0],
'on_intersection': [15.0, -3.0],
'on_pickup_dropoff': [15.0, -3.0],
'on_stopline_crosswalk': [15.0, -3.0],
'on_stopline_stop_sign': [15.0, -3.0],
'on_stopline_traffic_light': [15.0, -3.0],
'on_traffic_light_intersection': [15.0, -3.0],
'starting_high_speed_turn': [15.0, -3.0],
'starting_left_turn': [15.0, -3.0],
'starting_low_speed_turn': [15.0, -3.0],
'starting_protected_cross_turn': [15.0, -3.0],
'starting_protected_noncross_turn': [15.0, -3.0],
'starting_right_turn': [15.0, -3.0],
'starting_straight_stop_sign_intersection_traversal': [15.0, -3.0],
'starting_straight_traffic_light_intersection_traversal': [15.0, -3.0],
'starting_u_turn': [15.0, -3.0],
'starting_unprotected_cross_turn': [15.0, -3.0],
'starting_unprotected_noncross_turn': [15.0, -3.0],
'stationary': [15.0, -3.0],
'stationary_at_crosswalk': [15.0, -3.0],
'stationary_at_traffic_light_with_lead': [15.0, -3.0],
'stationary_at_traffic_light_without_lead': [15.0, -3.0],
'stationary_in_traffic': [15.0, -3.0],
'stopping_at_crosswalk': [15.0, -3.0],
'stopping_at_stop_sign_no_crosswalk': [15.0, -3.0],
'stopping_at_stop_sign_with_lead': [15.0, -3.0],
'stopping_at_stop_sign_without_lead': [15.0, -3.0],
'stopping_at_traffic_light_with_lead': [15.0, -3.0],
'stopping_at_traffic_light_without_lead': [15.0, -3.0],
'stopping_with_lead': [15.0, -3.0],
'traversing_crosswalk': [15.0, -3.0],
'traversing_intersection': [15.0, -3.0],
'traversing_narrow_lane': [15.0, -3.0],
'traversing_pickup_dropoff': [15.0, -3.0],
'traversing_traffic_light_intersection': [15.0, -3.0],
'waiting_for_pedestrian_to_cross': [15.0, -3.0]
}
return scenario_map
def get_filter_parameters(num_scenarios_per_type=None, limit_total_scenarios=None):
scenario_types = [
'starting_left_turn',
'starting_right_turn',
'starting_straight_traffic_light_intersection_traversal',
#'stopping_with_lead',
'high_lateral_acceleration',
'high_magnitude_speed',
'low_magnitude_speed',
'traversing_pickup_dropoff',
#'waiting_for_pedestrian_to_cross',
#'behind_long_vehicle',
#'stationary_in_traffic',
'near_multiple_vehicles',
'changing_lane',
'following_lane_with_lead',
]
scenario_tokens = None # List of scenario tokens to include
log_names = None # Filter scenarios by log names
map_names = None # Filter scenarios by map names
num_scenarios_per_type = num_scenarios_per_type # Number of scenarios per type
limit_total_scenarios = limit_total_scenarios # Limit total scenarios (float = fraction, int = num) - this filter can be applied on top of num_scenarios_per_type
timestamp_threshold_s = None # Filter scenarios to ensure scenarios have more than `timestamp_threshold_s` seconds between their initial lidar timestamps
ego_displacement_minimum_m = None # Whether to remove scenarios where the ego moves less than a certain amount
expand_scenarios = False # Whether to expand multi-sample scenarios to multiple single-sample scenarios
remove_invalid_goals = True # Whether to remove scenarios where the mission goal is invalid
shuffle = False # Whether to shuffle the scenarios
ego_start_speed_threshold = None # Limit to scenarios where the ego reaches a certain speed from below
ego_stop_speed_threshold = None # Limit to scenarios where the ego reaches a certain speed from above
speed_noise_tolerance = None # Value at or below which a speed change between two timepoints should be ignored as noise.
return scenario_types, scenario_tokens, log_names, map_names, num_scenarios_per_type, limit_total_scenarios, timestamp_threshold_s, ego_displacement_minimum_m, \
expand_scenarios, remove_invalid_goals, shuffle, ego_start_speed_threshold, ego_stop_speed_threshold, speed_noise_tolerance
def get_low_level_metrics():
low_level_metrics = {
'ego_acceleration': EgoAccelerationStatistics(name='ego_acceleration', category='Dynamics'),
'ego_expert_L2_error': EgoExpertL2ErrorStatistics(name='ego_expert_L2_error', category='Planning', discount_factor=1),
'ego_expert_l2_error_with_yaw': EgoExpertL2ErrorWithYawStatistics(name='ego_expert_l2_error_with_yaw', category='Planning', discount_factor=1),
'ego_jerk': EgoJerkStatistics(name='ego_jerk', category='Dynamics', max_abs_mag_jerk=8.37),
'ego_lane_change': EgoLaneChangeStatistics(name='ego_lane_change', category='Planning', max_fail_rate=0.3),
'ego_lat_acceleration': EgoLatAccelerationStatistics(name='ego_lat_acceleration', category='Dynamics', max_abs_lat_accel=4.89),
'ego_lat_jerk': EgoLatJerkStatistics(name='ego_lat_jerk', category='Dynamics'),
'ego_lon_acceleration': EgoLonAccelerationStatistics(name='ego_lon_acceleration', category='Dynamics', min_lon_accel=-4.05, max_lon_accel=2.40),
'ego_lon_jerk': EgoLonJerkStatistics(name='ego_lon_jerk', category='Dynamics', max_abs_lon_jerk=4.13),
'ego_mean_speed': EgoMeanSpeedStatistics(name='ego_mean_speed', category='Dynamics'),
'ego_progress_along_expert_route': EgoProgressAlongExpertRouteStatistics(name='ego_progress_along_expert_route', category='Planning', score_progress_threshold=2),
'ego_yaw_acceleration': EgoYawAccelerationStatistics(name='ego_yaw_acceleration', category='Dynamics', max_abs_yaw_accel=1.93),
'ego_yaw_rate': EgoYawRateStatistics(name='ego_yaw_rate', category='Dynamics', max_abs_yaw_rate=0.95),
'planner_expert_average_l2_error_within_bound': PlannerExpertAverageL2ErrorStatistics(name='planner_expert_average_l2_error_within_bound',
category='Planning', metric_score_unit='float',
comparison_horizon=[3, 5, 8], comparison_frequency=1,
max_average_l2_error_threshold=8)
}
return low_level_metrics
def get_high_level_metrics(low_level_metrics):
high_level_metrics = {
'drivable_area_compliance': DrivableAreaComplianceStatistics(name='drivable_area_compliance', category='Planning',
lane_change_metric=low_level_metrics['ego_lane_change'],
max_violation_threshold=0.3, metric_score_unit='bool'),
'driving_direction_compliance': DrivingDirectionComplianceStatistics(name='driving_direction_compliance', category='Planning',
lane_change_metric=low_level_metrics['ego_lane_change'], metric_score_unit='bool'),
'ego_is_comfortable': EgoIsComfortableStatistics(name='ego_is_comfortable', category='Violations', metric_score_unit='bool',
ego_jerk_metric=low_level_metrics['ego_jerk'],
ego_lat_acceleration_metric=low_level_metrics['ego_lat_acceleration'],
ego_lon_acceleration_metric=low_level_metrics['ego_lon_acceleration'],
ego_lon_jerk_metric=low_level_metrics['ego_lon_jerk'],
ego_yaw_acceleration_metric=low_level_metrics['ego_yaw_acceleration'],
ego_yaw_rate_metric=low_level_metrics['ego_yaw_rate']),
'ego_is_making_progress': EgoIsMakingProgressStatistics(name='ego_is_making_progress', category='Planning',
ego_progress_along_expert_route_metric=low_level_metrics['ego_progress_along_expert_route'],
metric_score_unit='bool', min_progress_threshold=0.2),
'no_ego_at_fault_collisions': EgoAtFaultCollisionStatistics(name='no_ego_at_fault_collisions', category='Dynamics', metric_score_unit='float',
ego_lane_change_metric=low_level_metrics['ego_lane_change']),
'planner_expert_average_heading_error_within_bound': PlannerExpertAverageHeadingErrorStatistics(name='planner_expert_average_heading_error_within_bound',
category='Planning', metric_score_unit='float', max_average_heading_error_threshold=0.8,
planner_expert_average_l2_error_within_bound_metric=low_level_metrics['planner_expert_average_l2_error_within_bound']),
'planner_expert_final_heading_error_within_bound': PlannerExpertFinalHeadingErrorStatistics(name='planner_expert_final_heading_error_within_bound',
category='Planning', metric_score_unit='float', max_final_heading_error_threshold=0.8,
planner_expert_average_l2_error_within_bound_metric=low_level_metrics['planner_expert_average_l2_error_within_bound']),
'planner_expert_final_l2_error_within_bound': PlannerExpertFinalL2ErrorStatistics(name='planner_expert_final_l2_error_within_bound', category='Planning',
metric_score_unit='float', max_final_l2_error_threshold=8,
planner_expert_average_l2_error_within_bound_metric=low_level_metrics['planner_expert_average_l2_error_within_bound']),
'planner_miss_rate_within_bound': PlannerMissRateStatistics(name='planner_miss_rate_within_bound', category='Planning', metric_score_unit='bool',
max_displacement_threshold=[6.0, 8.0, 16.0], max_miss_rate_threshold=0.3,
planner_expert_average_l2_error_within_bound_metric=low_level_metrics['planner_expert_average_l2_error_within_bound']),
'speed_limit_compliance': SpeedLimitComplianceStatistics(name='speed_limit_compliance', category='Violations', metric_score_unit='float',
max_violation_threshold=1.0, max_overspeed_value_threshold=2.23, lane_change_metric=low_level_metrics['ego_lane_change'])
}
high_level_metrics.update({
'time_to_collision_within_bound': TimeToCollisionStatistics(name='time_to_collision_within_bound', category='Planning', metric_score_unit='bool',
time_step_size=0.1, time_horizon=3.0, least_min_ttc=0.95,
ego_lane_change_metric=low_level_metrics['ego_lane_change'],
no_ego_at_fault_collisions_metric=high_level_metrics['no_ego_at_fault_collisions'])})
return high_level_metrics
def get_metrics_config(experiment, low_level_metrics, high_level_metrics):
if experiment == "open_loop_boxes":
metrics = [low_level_metrics['planner_expert_average_l2_error_within_bound'],
high_level_metrics['planner_expert_final_l2_error_within_bound'],
high_level_metrics['planner_miss_rate_within_bound'],
high_level_metrics['planner_expert_final_heading_error_within_bound'],
high_level_metrics['planner_expert_average_heading_error_within_bound']
]
elif experiment == 'closed_loop_nonreactive_agents' or experiment == 'closed_loop_reactive_agents':
metrics = [low_level_metrics['ego_lane_change'], low_level_metrics['ego_jerk'],
low_level_metrics['ego_lat_acceleration'], low_level_metrics['ego_lon_acceleration'],
low_level_metrics['ego_lon_jerk'], low_level_metrics['ego_yaw_acceleration'],
low_level_metrics['ego_yaw_rate'], low_level_metrics['ego_progress_along_expert_route'],
high_level_metrics['drivable_area_compliance'], high_level_metrics['no_ego_at_fault_collisions'],
high_level_metrics['time_to_collision_within_bound'], high_level_metrics['speed_limit_compliance'],
high_level_metrics['ego_is_comfortable'], high_level_metrics['ego_is_making_progress'],
high_level_metrics['driving_direction_compliance']
]
else:
raise TypeError("Experiment type not supported!")
return metrics
def build_metrics_engine(experiment, output_dir, metric_dir):
main_save_path = pathlib.Path(output_dir) / metric_dir
low_level_metrics = get_low_level_metrics()
high_level_metrics = get_high_level_metrics(low_level_metrics)
selected_metrics = get_metrics_config(experiment, low_level_metrics, high_level_metrics)
metric_engine = MetricsEngine(main_save_path=main_save_path)
for metric in selected_metrics:
metric_engine.add_metric(metric)
return metric_engine