Edouard Leurent

Results 176 comments of Edouard Leurent

Hi, I'm not sure, but it may because you are using a jupyter notebook to run the code, maybe this prevents from opening the rendering window. Could you try running...

As explained in https://github.com/eleurent/highway-env/issues/278, target speeds is a set of speeds that can be tracked by the vehicle. For instance, you could set target_speeds = [5 m/s, 10 m/s, 15...

Sure, you could just edit this method: https://github.com/eleurent/highway-env/blob/ea79a9557e09c4f0f867a76da4926ed1d8aac22c/highway_env/vehicle/kinematics.py#L201 and add ``` 'id': id(self) ``` And then simply add this 'id' key to the observation features, as explained in [the docs](https://highway-env.readthedocs.io/en/latest/observations/index.html#kinematics)....

You can edit `Vehicle.__init__` in kinematics.py as follows: ``` max_id : int = 0 def __init__(self, road: Road, position: Vector, heading: float = 0, speed: float = 0, predition_type: str...

In principle, the gains (TAUs) of the controllers are already in expressed in seconds, so changing the simulation timestep should not change the behaviour / response of the vehicle. That...

This is because they are create on the (i, j) lane with a longitudinal coordinate much larger than the lane length. Try reducing the `ego_spacing`, and creating the controlled vehicles...

See https://github.com/eleurent/highway-env/issues/130 and https://github.com/eleurent/highway-env/issues/63

Hi TibiGG. I am not sure what you mean exactly, since the nodes do not have positions: they are only abstract vertices in the graph which defines the topological road...

@piccupl not sure about your question either, you can specify the longitudinal displacement to any value you want, and you can also set a lateral displacement. Alternatively, you can set...

To add a new observation type, you should edit the [envs/common/observation.py](https://github.com/eleurent/highway-env/blob/master/highway_env/envs/common/observation.py) file. You can add a new Type, and in its `observe()` method you can access the current vehicle lane...