Sort the predictions based on the probability of each mode. Args: predictions (torch.Tensor): The predicted trajectories [b, k, t, 2].

Determines wether values in batch-dim are within bounds. :param metric: metric values :param min_bound: minimum bound, defaults to None

Searches for intersecting geometries in the occupancy map :param geometry: geometries to query :return: list of tokens for in

Performs a simple euler integration. :param init: Initial state :param delta: The rate of change of the state. :param sampling_time:

Function to directly calls shapely's query function on str-tree :param geometry: geometries to query :param predicate: see sh

Create a dataset from a list of samples. :param samples: List of dataset candidate samples. :param feature_preprocessor: Feature preproce

Args: candidate_trajectories: (N, 80+1, S), sampled at 10 Hz

Extract acceleration of ego pose in simulation history over batch-dim :param states: array representation of ego state values :param acce

Returns the corresponding profile (i.e. trajectory) given an initial condition and derivatives at multiple timesteps by integration. :par

Analytical solution for min_{c} |cx_1 + (1-c)x_2|_2^2 d is the distance (objective) optimzed v1v1 = <x1,x1> v1v2 = <x

Propagates the model forward and backwards and computes/logs losses and metrics. This is called either during training, validation o

Propagates the model forward and backwards and computes/logs losses and metrics. This is called either during training, validation o

Build the three core lightning modules: LightningDataModule, LightningModule and Trainer :param cfg: omegaconf dictionary :param worker:

Get the arc length of each point in a polyline

Apply BFS to find route to target roadblock. :param target_roadblock_id: id of target roadblock :param max_depth: maximum sea

Given two equal-length sequences y and x, compute an approximation to the n-th derivative of some function interpolating the (x, y) data poin

Extract jerk of ego pose in simulation history over batch-dim :param states: array representation of ego state values :param acceleration

Extract yaw-rate of simulation history over batch-dim :param states: array representation of ego state values :param time_steps_s: time s

Find the minimum norm solution as combination of two points This is correct only in 2D ie. min_c |\sum c_i x_i|_2^2 st. \sum

performing batch-wise collision check using Separating Axis Theorem params: ego_states: [bs, (x, y, theta)], center of the ego

Args: position: (N, T, 2) angle: (N, T) shape: (N, 2) [width, length] Returns: 4 corners of oriented box (FL,

Classify collision between ego and the track. :param ego_state: Ego's state at the current timestamp. :param tracked_object: Tracked obje

Loads the schedulers state. Args: state_dict (dict): scheduler state. Should be an object returned from a call to

Performs dijkstra's shortest path to find a route to the target roadblock. :param target_roadblock: The target roadblock the path sho

img: (H, W, C) shift: (H_shift, W_shift, 0) resolution: float angle: float

Loads proposal state array and resets velocity, and curvature profile. :param proposal_states: array representation of proposals.

Builds occupancy map of drivable area. :param ego_state: EgoState

Builds occupancy map of obstacles. :param ego_state: EgoState

Check if the search should end regardless if the goal condition is met. :param depth: The current depth to check. :param targ

Check if the current edge is at the target roadblock at the given depth. :param current_edge: The edge to check. :param targe

Check if the current edge is at the target roadblock at the given depth. :param current_edge: edge to check. :param depth: cu

Compute triplet loss Args: hidden: (3*bs, D)

Compute triplet loss Args: hidden: (3*bs, D)

This method projects the initial tracking error into vehicle/Frenet frame. It also extracts initial velocity. :param current_iterati

Computes a set of planning metrics given the model's predictions and targets. :param predictions: model's predictions :param

Computes a set of planning metrics given the model's predictions and targets. :param predictions: model's predictions :param

This method computes reference velocity and curvature profile based on the reference trajectory. We use a lookahead time equal to sel

:param end_edge: The end edge to start back propagating back to the start edge. :param depth: The depth of the target edge. :

Constructs a path when goal was found. :param end_edge: The end edge to start back propagating back to the start edge. :param

Edge cost of given lane. :param lane: lane class :return: length of lane