Created
October 16, 2022 07:31
-
-
Save SolClover/d747393a31b23f220ebcc9a188dea279 to your computer and use it in GitHub Desktop.
Evaluate agent by visualizing its actions
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# Reset environment to initial state | |
state, info = env.reset() | |
# Cycle through 50 steps redering and displaying environment state each time | |
for _ in range(50): | |
# Render and display current state of the environment | |
plt.imshow(env.render()) # render current state and pass to pyplot | |
plt.axis('off') | |
display.display(plt.gcf()) # get current figure and display | |
display.clear_output(wait=True) # clear output before showing the next frame | |
# Use greedy policy to evaluate | |
action = eval_greedy(Qtable, state) | |
# Pass action into step function | |
state, reward, done, _, info = env.step(action) | |
# Reset environment when done=True, i.e. when the agent falls into a Hole (H) or reaches the Goal (G) | |
if done: | |
# Render and display final state of the environment | |
plt.imshow(env.render()) # render current state and pass to pyplot | |
plt.axis('off') | |
display.display(plt.gcf()) # get current figure and display | |
display.clear_output(wait=True) # clear output before showing the next frame | |
state, info = env.reset() | |
env.close() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment