Add DroneEnv environment#10
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
Codex Review: Here are some suggestions.
Reply with @codex fix comments to fix any unresolved comments.
About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you open a pull request for review, mark a draft as ready, or comment "@codex review". If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex fix this CI failure" or "@codex address that feedback".
| # Observation: [x, y, goal_x, goal_y, steps_remaining] | ||
| self.observation_space = spaces.Box( | ||
| low=0.0, | ||
| high=float(self.grid_size), | ||
| shape=(5,), | ||
| dtype=np.float32, | ||
| ) |
There was a problem hiding this comment.
[P1] Define observation bounds for steps_remaining
The observation space uses spaces.Box(low=0.0, high=float(self.grid_size), shape=(5,), dtype=np.float32), which applies the grid size as the upper bound to all five features. However _get_obs() returns steps_remaining = self.max_steps - self.steps as the last element, and max_steps defaults to 100 while grid_size defaults to 10. In any normal run steps_remaining will be >10, so the validator reports an observation drift on every step and tests/test_env_integrity_validator_clean_run will fail. The space needs per‑component bounds (e.g., separate high for the step counter) so that valid states are considered in-range.
Useful? React with 👍 / 👎.
1 participant
Summary
DroneEnvgym environment with configurable grid, action/observation spaces, and integrity validationDroneEnvviaenvpackageTesting
PYTHONPATH=src pytest(fails: ModuleNotFoundError: No module named 'numpy')https://chatgpt.com/codex/tasks/task_e_68beb2fea844832b8301bda59c53e04d