Edouard Leurent

Sorry for not replying, let me know if this is still needed.

Oooooh, that's a great catch @Shuchengzhang888! thanks :) Yeah, this observation normalisation - denormalisation in the attention visualisation is really ugly. Maybe it would be better to directly include vehicle...

(`env.seed()` was the gym

Oh okay, it looks like I temporarily disabled seeding when transitioning to the gym 0.26 API, and I completely forgot about it afterward. Anyway, this is now fixed at HEAD,...

Hi, no I have not. If you try, let me know how it goes :)

> highway_env compatible of handling multi-agent Graph Reinforcement Learning? In this case, the observation space is the product of the observation matrix O, adjacency matrix A, and mask matrix M...

I'm not sure what could be causing this. I would try to run the code you used to make the Polylanes in debugger mode and check that the coordinates of...

Hi, sorry I can't seem to find any trace of the code that appears failing, namely `self.new_action = safety_supervisor(action)` and `self.check_collision(vehicle, other, other.trajectories[t]` both supposedly in abstract.py. I think this...

Hi, I have not been able to reproduce the issue, [see this colab](https://colab.research.google.com/drive/1tqACS8XRBlk9qUzsCc44brZllNs6EWT5?usp=sharing) where everything seems to be working fine.

"Lane indices" apply to "road segments" where all lanes are side by side and go in the same direction (this is the criterion to allow vehicles to do a lane...