def run(args):
# logger
make_logger(args.result_dir)
# make envs
env = make_env(args)
# make config
config = make_config(args)
# make planner
planner = make_planner(args, config)
# make model
model = make_model(args, config)
# make controller
controller = make_controller(args, config, model)
# make simulator
runner = make_runner(args)
# run experiment
history_x, history_u, history_g = runner.run(env, controller, planner)
# plot results
plot_results(args, history_x, history_u, history_g=history_g)
save_plot_data(args, history_x, history_u, history_g=history_g)
def main():
parser = argparse.ArgumentParser()
parser.add_argument("--env", type=str, default="TwoWheeledTrack")
# parser.add_argument("--env", type=str, default="CartPole")
parser.add_argument("--save_anim", type=bool_flag, default=1)
parser.add_argument("--controller_type", type=str, default="NMPCCGMRES")
# parser.add_argument("--controller_type", type=str, default="MPPI")
parser.add_argument("--result_dir", type=str, default="./result")
parser.add_argument("--use_learning", type=str, default=True)
parser.add_argument("--num_train_steps_per_iter",
type=np.ndarray,
default=[40000, 80000, 120000, 160000])
parser.add_argument("--relabel_with_expert", type=str, default=True)
args = parser.parse_args()
trainer = IL_trainer(args)
initial_expertdata_path = os.path.join(args.result_dir + '/' +
args.controller_type)
print(
'Running rl_trainer for {0} with controller type {1} \n Using Dagger: {2}'
.format(args.env, args.controller_type, args.relabel_with_expert))
# config = make_config(args)
traj_steps = trainer.env.config[
"max_step"] # make sure that the sampled trajectory length is the same as the initial training trajectory length
trainer.run_training_loop(n_iter=4,
initial_expertdata=initial_expertdata_path,
relabel_with_expert=args.relabel_with_expert,
start_relabel_with_expert=1,
traj_steps=traj_steps)
print('Finish training ...')
# still testing, need to refactor
make_logger(args.result_dir)
config = make_config(args)
planner = make_planner(args, config)
history_x, history_u, history_g, cost = trainer.run(planner)
plot_results(history_x, history_u, history_g=history_g, args=args)
save_plot_data(history_x,
history_u,
history_g=history_g,
cost=cost,
args=args)
if args.save_anim:
animator = Animator(env=trainer.env, args=args)
print("first in history_x", history_x[0])
animator.draw(history_x, history_g)
def run(args):
make_logger(args.result_dir)
env = make_env(args)
config = make_config(args)
planner = make_planner(args, config)
model = make_model(args, config)
controller = make_controller(args, config, model)
runner = make_runner(args)
history_x_all, history_u_all, history_g_all, history_cost_all = [], [], [], [
] # this is the collection list of n_sample trajectories
for iter_sample in range(args.n_sample):
print("Sampling {} th trajectory generated by expert policy:".format(
iter_sample))
history_x, history_u, history_g, cost = runner.run(
env, controller, planner)
history_x_all.append(history_x)
history_u_all.append(history_u)
history_g_all.append(history_g)
history_cost_all.append(cost)
plot_results(
history_x, history_u, history_g=history_g,
args=args) # no need to change now, just see the plot of the last traj
save_plot_data(history_x_all,
history_u_all,
history_g=history_g_all,
cost=history_cost_all,
args=args) # save lists
if args.save_anim:
animator = Animator(env, args=args)
animator.draw(history_x, history_g)