Recalculating Reward

[vyasakash231]

Hi.., I tried your code and it's not working, so I made a small change and it's working now. can you please confirm that you also faced the same issue?

# Recalculates the reward
    if observation[1] > state[0][1] >= 0 and observation[1] >= 0:
        reward = 20
    if observation[1] < state[0][1] <= 0 and observation[1] <= 0:
        reward = 20
    if done and episode_step < num_episode_steps - 1:
        reward += 10000
    else:
        reward -= 25

I just changed observation[1] to observation[0] after and and it worked.

# Recalculates the reward
    if observation[1] > state[0][1] >= 0 and observation[0] >= 0:
        reward = 20
    if observation[1] < state[0][1] <= 0 and observation[0] <= 0:
        reward = 20
    if done and episode_step < num_episode_steps - 1:
        reward += 10000
    else:
        reward -= 25

Also, one more question, how did you came up with this reward func?