mu_reward = np.mean(list_Reward, axis=0)
std_reward = np.std(list_Reward, axis=0)
print('Mean Reward', mu_reward)
print('Std Reward', std_reward)
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument("--gpu", help="the GPU to use")
args = parser.parse_args()
if (args.gpu):
os.environ["CUDA_VISIBLE_DEVICES"] = args.gpu
env = Environment(scenario_file)
agent = Agent(env.NumActions())
reward_list_training = []
number_of_training_episodes = parameter.how_many_times / parameter.save_each
if (Train_Model):
for i in range(1, parameter.how_many_times_training + 1):
mean_scores = []
if (Train_Model):
print("Training Iteration {}, using {}".format(i, Feature))
agent.Train()
print('Mean Scores', mean_scores)
reward_list_training.append(mean_scores)
print("Mean List Reward", reward_list_training)
mu_reward_training = np.mean(reward_list_training, axis=0)
std_reward_training = np.std(reward_list_training, axis=0)
# plt.xlabel('Number of Episodes')
# plt.ylabel('Mean Reward')
# file_name = model_path+"Test_"+Feature + '_' + str(how_many_times) + '_' + str(number_of_episodes) + '.png'
# plt.savefig(file_name)
# plt.show()
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument("--gpu", help="the GPU to use")
args = parser.parse_args()
if (args.gpu):
os.environ["CUDA_VISIBLE_DEVICES"] = args.gpu
env = Environment(scenario_file)
agent = Train(env.NumActions())
reward_list_training=[]
number_of_training_episodes=parameter.how_many_times/parameter.save_each
for i in range(1,parameter.how_many_times_training+1):
mean_scores=[]
if (Train_Model):
print("Training Iteration {}, using {}".format(i,Feature))
agent.Train()
print('Mean Scores',mean_scores)
reward_list_training.append(mean_scores)
#Test_Model(agent)
print("Mean List Reward",reward_list_training)
mu_reward_training = np.mean(reward_list_training, axis=0)
std_reward_training = np.std(reward_list_training, axis=0)
number_of_steps = len(reward_list_training[0])
