agent = Agent(gamma=GAMMA, epsilon=EPSILON, batch_size=BATCH_SIZE, max_mem_size=500000,
n_actions=NUMBER_OF_ACTIONS, eps_end=0.01, input_dims=[2], lr=LR)
scores, eps_history = [], []
n_runs = 10
df = pd.DataFrame()
for i in range(n_runs):
score = 0
done = False
observation = env.reset()
run = []
while not done:
action = agent.choose_action(observation)
observation_, reward, done, info = env.step(action)
score += reward
agent.store_transition(observation, action, reward, observation_, done)
agent.learn()
observation = observation_
run.append(observation)
scores.append(score)
eps_history.append(agent.epsilon)
avg_score = np.mean(scores[-100:])
run_name = str(uuid.uuid4())
x = pd.DataFrame()
x[run_name] = run
df = pd.concat([df, x], axis=1)
run.clear()
print(f"Episode: {i}",
f"Score: {score}",
