lateral op• Data kinds: any → table
• Call: import fullseye as fs; fs.ledger.pure_pursuit_curvature(pose, path, lookahead, *, start_index=None, window: 'int' = 40) -> 'Dict[str, np.ndarray]' (to call the implementation directly, import drivelateral; drivelateral.pure_pursuit_curvature(pose, path, lookahead, *, start_index=None, window: 'int' = 40) -> 'Dict[str, np.ndarray]'; from the registry, opsdrive.get("pure_pursuit_curvature"))
pure pursuit: curvature κ = 2 sin α / L_d for aiming from the rear-axle center (pose = (x, y, ψ), (N, 3) allowed) at the point a distance L_d ahead on the path.
The path is a polyline (M, 2). Find the closest point, then solve a quadratic on the segment for the first point beyond it whose distance is L_d (if the end of the path is not reached, aim at the end point and set L_d to that distance). start_index and window narrow the range searched for the closest point (speeds up repeated calls on long paths; None means the whole path).
Returns: kappa, alpha, goal (N, 2), index (closest segment), lateral (lateral offset from the path, left positive), lookahead (actual distance). For a single pose the leading dimension is dropped.
Raises ValueError: wrong shape, L_d ≤ 0, start_index out of range.
• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).
• Operator provenance and references — the sources of the research/methods this op family came from.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
• poc_driving_lateral — py -3.11 examples/poc_driving_lateral.py
table as input)course_layout · course_occupancy · course_contains · world_build · world_camera · world_move · lidar_scan · rss_longitudinal_same
lateral)friction_circle_usage · curve_speed_limit · design_min_radius · understeer_gradient · steady_cornering · bicycle_model_step · ackermann_steer_angles · offtracking_circle
*Provenance: drivelateral.py — DRIVE operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.