75028b69a20261c44081f807a4a18d6d4bcb2929,cistar-dev/build-tester.py,,,#,9

Before Change

        [15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
         15, 15, 15, 15])

exp.env.terminate()

After Change

cfg_params = {"type_list": ["rl"], "start_time": 0, "end_time":3000, "cfg_path":"debug/cfg/", "num_cars":num_cars, "type_counts":{"rl": auton_cars}, "use_flows":True, "period":"1"}

initial_positions = [("top", 0), ("top", 70), ("top", 140), \
                    ("left", 0), ("left", 70), ("left", 140), \
                    ("bottom", 0), ("bottom", 70), ("bottom", 140), \
                    ("right", 0), ("right", 70), ("right", 140)]
initial_config = {"positions": initial_positions, "shuffle": False}

scenario = LoopScenario("test-exp", num_cars, type_params, cfg_params, net_params, initial_config=initial_config, generator_class=CircleGenerator)

////data path needs to be relative to cfg location

exp = SumoExperiment(SimpleVelocityEnvironment, env_params, sumo_binary, sumo_params, scenario)

logging.info("Experiment Set Up complete")

print("experiment initialized")

env = normalize(exp.env)

for seed in [1, 5, 10, 73, 56]:
    policy = GaussianMLPPolicy(
        env_spec=env.spec,
        hidden_sizes=(16,)
    )

    baseline = LinearFeatureBaseline(env_spec=env.spec)

    algo = TRPO(
        env=env,
        policy=policy,

Instances