def train_4v4():
envs_config_paths = [
Path(JSONS_FOLDER, 'configs', '4v4', 'all_equal.json'),
Path(JSONS_FOLDER, 'configs', '4v4', 'more_horizontally.json'),
Path(JSONS_FOLDER, 'configs', '4v4', 'more_vertically.json')
]
training_state = get_frap_training_4v4()
training_state.junctions = [
"gneJ1", "gneJ2", "gneJ3", "gneJ4", "gneJ7", "gneJ8", "gneJ9",
"gneJ10", "gneJ14", "gneJ15", "gneJ16", "gneJ17", "gneJ20", "gneJ21",
"gneJ22", "gneJ23"
]
evaluator = Evaluator.from_file(
Path(JSONS_FOLDER, 'evaluators', '4v4_eq_vert_hori.json'))
main_train(
training_state,
[
SumoEnv.from_config_file(env_config_path, 3000)
for env_config_path in envs_config_paths
],
evaluator,
Path('saved', 'aaai-multi', 'frap', '4v4'),
)
def train_aaai():
env_config_path = Path(JSONS_FOLDER, 'configs', 'aaai_random.json')
evaluator = Evaluator.from_file(
Path(JSONS_FOLDER, 'evaluators', 'example_test.json'))
main_train(
get_frap_training(),
SumoEnv.from_config_file(env_config_path),
evaluator,
Path('saved', 'aaai-random', 'frap'),
)
def train_2v2():
env_config_path = Path(JSONS_FOLDER, 'configs', '2v2', 'all_equal.json')
training_state = get_frap_training_2v2()
training_state.junctions = ['gneJ25', 'gneJ26', 'gneJ27', 'gneJ28']
evaluator = Evaluator.from_file(
Path(JSONS_FOLDER, 'evaluators', '2v2_small_subset.json'))
main_train(
training_state,
SumoEnv.from_config_file(env_config_path, 3000),
evaluator,
Path('saved', 'aaai-multi', 'frap'),
)
def main():
# evaluator = Evaluator.from_file("jsons/evaluators/example_test.json")
# evaluator = Evaluator.from_file("jsons/evaluators/osm_test.json")
# evaluator = Evaluator.from_file("jsons/evaluators/2v2_small_subset.json")
evaluator = Evaluator.from_file('jsons/evaluators/4v4_eq_vert_hori.json')
# evaluator = Evaluator.from_file('jsons/evaluators/example_test.json')
osm_phase_map = {
0: [
"gneE8_0", "gneE8_1", "166869096#1_0", "166869096#1_1",
"166869096#1_2", "166869096#1_3", "gneE16_0", "gneE16_1"
],
1: ["gneE7_0", "gneE7_1", "gneE18_0", "gneE18_1"],
2:
["gneE8_0", "gneE8_1", "gneE8_2", "gneE8_3", "gneE16_0", "gneE16_1"],
3: [
"166869096#1_0", "166869096#1_1", "gneE1_0", "gneE1_1", "gneE18_0",
"gneE18_1"
]
}
controller1 = VehicleNumberController(osm_phase_map)
phase_map = get_phase_map()
vehicle_number_controller_2x2 = [
VehicleNumberController(phase_map[tls_id])
for tls_id in phase_map.keys()
]
phase_map_1v1 = get_phase_map_1v1()
vehicle_number_controller_1x1 = [
VehicleNumberController(phase_map_1v1[tls_id])
for tls_id in phase_map_1v1.keys()
]
pressure_controller_1x1 = [
VehicleNumberPressureController(tls_id, phase)
for tls_id, phase in phase_map_1v1.items()
]
timed_controller = TimedCyclicSwitchController(list(range(8)), [5] * 8)
controller_rand = RandomSwitchController(list(range(8)))
phase_map = get_phase_map()
pressure_controller_2x2 = [
VehicleNumberPressureController(tls_id, phase)
for tls_id, phase in phase_map.items()
]
model_1v1_training_state = TrainingState.from_path(
str(
Path(PROJECT_ROOT, 'saved', 'aaai-random', 'frap',
'frap_2020-11-12.23-17-52-665056', 'states',
'ep_9_frap_2020-11-12.23-17-52-665056.tar')))
model_1v1 = model_1v1_training_state.model
model_1v1_controller = ModelController(model_1v1.eval())
model_2v2_training_state = TrainingState.from_path(
str(
Path(PROJECT_ROOT, 'saved', 'aaai-multi', 'frap',
'frap_2020-12-13.12-43-58-417983', 'states',
'ep_181_frap_2020-12-13.12-43-58-417983.tar')))
model_2v2 = model_2v2_training_state.model
model_2v2_controller = ModelController(model_2v2.eval())
# training_state = TrainingState.from_path(
# Path('ep_181_frap_2020-12-13.12-43-58-417983.tar'))
# model1 = training_state.model
# model1 = model1.eval()
# controller_mod = ModelController(model1)
controllers_2cyc_2model = [
controller_rand, timed_controller, vehicle_number_controller_2x2,
pressure_controller_2x2
]
controllers_1x1 = [
# controller_rand,
# timed_controller,
# vehicle_number_controller_1x1,
# pressure_controller_1x1
model_1v1_controller
]
phase_map_4v4 = get_phase_map_four_v_four()
controllers_4v4_vehicle = [
VehicleNumberController(phase_map_4v4[tls_id])
for tls_id in phase_map_4v4.keys()
]
controllers_4v4_pressure = [
VehicleNumberPressureController(tls_id, phase_map_4v4[tls_id])
for tls_id in phase_map_4v4.keys()
]
controller_4v4_timed = TimedCyclicSwitchController(list(range(8)),
[10] * 8)
training_state = TrainingState.from_path(
str(
Path('saved', 'aaai-multi', 'frap', '4v4',
'frap_2020-12-13.16-24-11-166272', 'states',
'ep_246_frap_2020-12-13.16-24-11-166272.tar')))
model_4v4 = training_state.model
model_4v4 = model_4v4.eval()
model_4v4_controller = ModelController(model_4v4)
# metrics = evaluator.evaluate_traffic_controllers(
# [controller_rand, controllers_4v4_vehicle, controllers_4v4_pressure], render=False)
metrics = evaluator.evaluate_traffic_controllers([
model_4v4_controller, controller_4v4_timed, controllers_4v4_vehicle,
controller_rand
],
render=True)
for env_n in metrics[0].keys():
print(env_n)
for m in metrics:
print(m[env_n])
