environment.environment.SimEnv Class Reference

Custom Environment that follows gym interface. More...

Inheritance diagram for environment.environment.SimEnv:

Public Member Functions
None	__init__ (self, str env_path, str urdf, bool use_graphics, int worker_id=0)
	The SimEnv class initializer. More...
None	set_goal (self, tuple goal)
	Set the coordinates of the goal visualization. More...
None	set_workspace (self, tuple workspace)
	Set the coordinates and size of the workspace visualization. More...
None	build_wall (self, List[List[bool]] wall)
	Build a new wall The first wall will be built on a certain distance from the anchor. More...
None	remove_walls (self)
	Remove all walls.
None	replace_walls (self, List[List[bool]] wall)
	Replace all the walls, build 1 new one See build_wall for how a wall is represented. More...
np.ndarray	step (self, np.ndarray action, return_observations=True)
	Do 1 step in the unity environment. More...
np.ndarray	reset (self)
	Reset the Unity Environment, essentially starting a new episode. More...
None	pause (self, steps=200)
	For a certain amount of steps, take no action. More...
None	close (self)
	Gym interface function to close the environment.
np.ndarray	get_current_state (self)
	Get the observations from the Unity environment. More...

Public Attributes
	env_path
	urdf
	use_graphics
	worker_id
	joint_amount
	u_env
	behavior_name
	behavior_spec

Static Public Attributes
dictionary	metadata = {'render.modes': ['human']}
int	MAX_N_MODULES = 10
int	JOINT_ANGLE_STEP = 10

Detailed Description

Custom Environment that follows gym interface.

Constructor & Destructor Documentation

__init__()

None environment.environment.SimEnv.__init__	(		self,
		str	env_path,
		str	urdf,
		bool	use_graphics,
		int	worker_id = 0
	)

The SimEnv class initializer.

Parameters

env_path	Path of the environment executable.
urdf	Instance that represents the robot urdf.
use_graphics	Boolean that turns graphics on or off.

Returns

An instance of the SimEnv class.

Member Function Documentation

build_wall()

None environment.environment.SimEnv.build_wall	(		self,
		List[List[bool]]	wall
	)

Build a new wall The first wall will be built on a certain distance from the anchor.

Every subsequent wall will be built on a certain distance from the previous wall.

Walls are represented by a 2D array of booleans True means there is a tile on that 'coordinate/index' and False meaning there is not

Parameters

wall	The new wall.

get_current_state()

np.ndarray environment.environment.SimEnv.get_current_state

(

self

)

Get the observations from the Unity environment.

Returns

The observations.

pause()

None environment.environment.SimEnv.pause	(		self,
			steps = 200
	)

For a certain amount of steps, take no action.

Achieved by taking "zero" actions.

replace_walls()

None environment.environment.SimEnv.replace_walls	(		self,
		List[List[bool]]	wall
	)

Replace all the walls, build 1 new one See build_wall for how a wall is represented.

Parameters

wall	The new wall

reset()

np.ndarray environment.environment.SimEnv.reset

(

self

)

Reset the Unity Environment, essentially starting a new episode.

The effect is that OnEpisodeBegin is called in Unity.

Returns

Observations obtained after the reset.

set_goal()

None environment.environment.SimEnv.set_goal	(		self,
		tuple	goal
	)

Set the coordinates of the goal visualization.

Parameters

goal	The coordinates: (x, y, z).

set_workspace()

None environment.environment.SimEnv.set_workspace	(		self,
		tuple	workspace
	)

Set the coordinates and size of the workspace visualization.

Parameters

workspace

The coordinates and size: (x, y, z, sideLength).

step()

np.ndarray environment.environment.SimEnv.step	(		self,
		np.ndarray	action,
			return_observations = True
	)

Do 1 step in the unity environment.

Parameters

action	The action for this step.
return_observations	Whether to return observations after the step is taken.

Returns

Observations after the step is taken.

---

The documentation for this class was generated from the following file:

• src/environment/environment.py