Source: https://github.com/openai/spinningup/blob/master/spinup/algos/tf1/ppo/core.py#L45

Returns the total elapsed time in seconds, computed from the last reset() call.

Terminates the current trajectory by adding the value of the terminal state

Returns direction and waypoint at a distance ahead defined by the user. :param steps: number of steps to get the incoming waypoi

Instantiate the agent's memory; easy to subclass

Accessor for topology. This function retrieves topology from the server as a list of road segments as pairs of waypoint objec

Check if a target object is within a certain distance in front of a reference object. :param target_transform: location of the target object

Defines the architecture of the policy-network

Augmentation function used during the reinforcement learning phase

Registers a given [callback] to a specific [event]

Skips the given amount of frames

Instantiate vehicles and pedestrians in the current world

Maps a discrete array of bins into their corresponding continuous values

Maps a continuous array of values into their corresponding bins (i.e. inverse of `interpolate`)

Maps a continuous array of values into their corresponding bins (i.e. inverse of `interpolate`)

This method returns list of (carla.Waypoint, RoadOption) from origin (carla.Location) to destination (carla.Location)

Enables synchronous mode. Usage: with carla_env as env: # code...

Disables synchronous mode

:param agent: a CARLAgent instance. :param control_policy: dict that specifies the policy-branch of the network. :param contr

This function builds a networkx graph representation of topology. The topology is read from self._topology. graph node prope

This function builds a networkx graph representation of topology. The topology is read from self._topology. graph node proper

Clips actions to prevent numerical instability when computing (log-)probabilities. - Use for Beta distribution only.

Compute the size (in meters) of the entire planned route

This method finds road segments that have an unconnected end and adds them to the internal graph representation

This method finds road segments that have an unconnected end, and adds them to the internal graph representation

Retrieves data for detector, the call is non-blocking thus doesn't wait for available data.

Transforms arbitrary complex state-spaces (and, optionally, action-spaces) as input layers

features: N distances from current vehicle location to N next route waypoints' locations

9 features: - 3: is_intersection, is_junction, is_at_traffic_light - 1: speed_limit - 5: traffic_light_state

Retrieves data for sensors (i.e. camera and other) it blocks waiting until timeout is expired.

Calculates the yaw of the waypoint that represents the trigger volume of the traffic light

4 features: - 1: similarity (e.g. current heading direction w.r.t. next route waypoint) - 1: speed - 1: throttle

Controller initialization. dt -- time difference between physics control in seconds. This is can be fixed from server side

This method places zero cost links in the topology graph representing availability of lane changes.

This method places zero cost links in the topology graph representing availability of lane changes.

This method is in charge of overtaking behaviors. :param location: current location of the agent :param waypoint: cu

Estimate the throttle/brake of the vehicle based on the PID equations :param target_speed: target speed in Km/h :pa

Estimate the steering angle of the vehicle based on the PID equations :param waypoint: target waypoint :param vehicl

Spawns itself within a carla.World.

This method returns the last successive intersection edge from a starting index on the route. This helps moving past tiny in

This method returns the last successive intersection edge from a starting index on the route. This helps moving past tiny int

This method is in charge of tailgating behaviors. :param location: current location of the agent :param waypoint: cu

This method sets up a global router and returns the optimal route from start_waypoint to end_waypoint. :param start_wayp

Specifies the mapping between an actions vector and the vehicle's control.

Augmentation function used during the imitation learning phase

Callback: called before world.tick()

Module in charge of car-following behaviors when there's someone in front of us. :param vehicle: car to follow

This module is in charge of warning in case of a collision and managing possible overtaking or tailgating chances. :para

Color distortion (jitter + drop) augmentation as defined in SimCLR paper

Color drop augmentation as defined in SimCLR paper

Color jitter augmentation as defined in SimCLR paper

Euclidean distance between 3D points :param location_1: 3D points :param location_2: 3D points