torchdrivesim.map

Classes

Stopline
MapConfig	Encapsulates various map metadata, including where to find files defining the map.

Functions

_filename_defaults(→ Dict[str, str])
resolve_paths_to_absolute(→ MapConfig)
load_map_config(→ MapConfig)
store_map_config(→ None)
find_map_config(→ Optional[MapConfig])	To retrieve configs for maps not bundled with the package,
download_iai_map(→ None)	Downloads the map information using LOCATION_INFO from the Inverted AI API.
traffic_controls_from_map_config(→ Dict[str, ...)
find_wrong_way_stoplines(→ List[int])

Module Contents

class torchdrivesim.map.Stopline

actor_id: int

agent_type: str

x: float

y: float

length: float

width: float

orientation: float

__post_init__()

class torchdrivesim.map.MapConfig

Encapsulates various map metadata, including where to find files defining the map. Map definition includes a coordinate frame and traffic signals.

name: str

left_handed_coordinates: bool = False

center: Tuple[float, float] | None = None

lanelet_path: str | None = None

lanelet_map_origin: Tuple[float, float] = (0, 0)

mesh_path: str | None = None

stoplines_path: str | None = None

traffic_light_controller_path: str | None = None

iai_location_name: str | None = None

note: str | None = None

property lanelet_map: torchdrivesim.lanelet2.LaneletMap | None

property road_mesh: torchdrivesim.mesh.BirdviewMesh | None

property stoplines: List[Stopline]

property traffic_light_controller: torchdrivesim.traffic_lights.TrafficLightController | None

torchdrivesim.map._filename_defaults(name: str) → Dict[str, str]

torchdrivesim.map.resolve_paths_to_absolute(cfg: MapConfig, root: str) → MapConfig

torchdrivesim.map.load_map_config(json_path: str, resolve_paths: bool = True) → MapConfig

torchdrivesim.map.store_map_config(cfg: MapConfig, json_path: str, store_absolute_paths: bool = False) → None

torchdrivesim.map.find_map_config(map_name: str, resolve_paths: bool = True) → MapConfig | None

To retrieve configs for maps not bundled with the package, folders with corresponding names must be placed inside one of the directories listed in TDS_RESOURCE_PATH environment variable. Note that map names should be globally unique.

torchdrivesim.map.download_iai_map(location_name: str, save_path: str) → None

Downloads the map information using LOCATION_INFO from the Inverted AI API. IAI_API_KEY needs to be set for this to succeed. Basename of save_path will be used as the map name for TorchDriveSim. If the dirpath of save_path is in TDS_RESOURCE_PATH, the map will be immediately available in find_map_config.

torchdrivesim.map.traffic_controls_from_map_config(cfg: MapConfig) → Dict[str, torchdrivesim.traffic_controls.BaseTrafficControl]

torchdrivesim.map.find_wrong_way_stoplines(map_cfg: MapConfig, angle_threshold: float = np.pi / 6) → List[int]