pytorch / pytorch/rl

Comment/question about the integration of a PettingZoo AEC environment with TorchRL

Open
#2,990 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
3.6k
Forks
487
Avg merge
1d 1h
Merged PRs (30d)
207

Description

I have a comment/question as far as the integration of PettingZoo AEC env with TorchRL is conducted.

In an AEC pettingzoo environment the observe() function is called only for the specific agent that has turn. However, I noticed that in the TorchRL library the observe() function is called for all the agents, even if most of them are not about to act.

We have created a custom AEC pettingzoo environment that simulates the routing decisions of vehicles. Each vehicle starts its trip at a specific timestep. When one agent/vehicle is about to act other agents may have not started their trip yet. Is it 'correct' to call the observe() function in this case? Can this negatively affect the learning process?

Is there some misunderstanding from my side or this can potentially be a problem? Or maybe should the environment by itself handle this?

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

No files, tests, or entry points are named. Start by tracing the PettingZoo AEC integration and TorchRL observation-collection behavior, then compare them with the custom environment's turn and not-yet-started-agent handling; done means documenting whether the calls are expected and whether an integration change is needed.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
ai, machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.