def main():
logging.info("test xworld navigation goal functions")
map_config_files = ['./xworld/map_examples/example7.json']
ego_centrics = [True]
for map_config_file in map_config_files:
for ego_centric in ego_centrics:
args = xworld_args.parser().parse_args()
args.ego_centric = ego_centric
args.map_config = map_config_file
args.show_frame = False
env = xworld_navi_goal.XWorldNaviGoal(args)
for i in range(2): #2
env.reset()
for j in range(200): #20
action = env.agent.random_action()
next_state, teacher, done = env.step(action)
env.display()
if done:
logging.info(
"test world navigation goal functions done")
break
logging.info("test world navigation goal functions done")
"""
Apply procedures of training for a QN
Args:
exp_schedule: exploration strategy for epsilon
lr_schedule: schedule for learning rate
"""
# initialize
self.initialize()
# model
self.evaluate()
if __name__ == '__main__':
# make env
args = xworld_args.parser().parse_args()
args.visible_radius_unit_side = config.visible_radius_unit_side
args.visible_radius_unit_front = config.visible_radius_unit_front
args.ego_centric = config.ego_centric
args.map_config = config.map_config_file
env = xworld_navi_goal.XWorldNaviGoal(args)
env.teacher.israndom_goal = False
env.teacher.goal_id = 0
# exploration strategy
exp_schedule = LinearExploration(env, config.eps_begin,
config.eps_end, config.eps_nsteps)
# learning rate schedule
lr_schedule = LinearSchedule(config.lr_begin, config.lr_end,
config.lr_nsteps)
