concepts.benchmark.namo.namo_polygon.namo_polygon_env.NamoPolygonEnv#

class NamoPolygonEnv[source]#

Bases: PymunkDefaultEnv

Methods

`add_additional_drawing_region_circle`(center, ...)
`add_additional_drawing_region_rect`(top_left, ...)
`attach_object`(body)
`can_control`(body)	Check if the body can be controlled by the user.
`detach_object`()
`distance_to_body`(body)
`get_jittered_velocity`(velocity)
`get_observation`()	Get an observation dict of the current state of the environment.
`humanplay_mainloop`()	Run a mainloop so that the user can control the objects in the environment.
`move_cfree_trajectory`(trajectory)
`move_cfree_trajectory_with_attached_object`(...)
`non_physical_execute_trajectory`(body_name, ...)	Execute a trajectory without physical simulation.
`render`()	Render the environment.
`render_and_display`([render_mode, trajectory])
`reset`(**kwargs)
`reset_scene`()
`reset_world`()
`set_additional_step_callable`(fn)
`set_agent_pos`(pos)
`set_current_attached_body`(body)
`set_current_selection`(body)
`start_recording`()
`step`(steps[, render_mode, callback])	Step the simulation for a number of steps.
`stop_recording`()

Attributes

`RENDER_MODES`
`SCREENSHOT_TITLE`
`TITLE`
`agent`
`agent_goal`
`agent_pos`
`agent_start`
`obstacles`

__init__(scene_json, **kwargs)[source]#

__new__(**kwargs)#

add_additional_drawing_region_circle(center, radius, color)#

Parameters:

center (Tuple[int, int])
radius (int)
color (Tuple[int, int, int])

add_additional_drawing_region_rect(top_left, bottom_right, color)#

Parameters:

top_left (Tuple[int, int])
bottom_right (Tuple[int, int])
color (Tuple[int, int, int])

attach_object(body)[source]#

Parameters:: body (Body)

can_control(body)#

Check if the body can be controlled by the user. This function can be overridden by subclasses.

Return type:: bool

detach_object()[source]#

distance_to_body(body)[source]#

Parameters:: body (Body)
Return type:: float

get_jittered_velocity(velocity)#

Parameters:: velocity (Tuple[float, float])
Return type:: Tuple[float, float]

get_observation()#: Get an observation dict of the current state of the environment.

humanplay_mainloop()#: Run a mainloop so that the user can control the objects in the environment.

move_cfree_trajectory(trajectory)[source]#

Return type:: NamoPolygonExecutionResult

move_cfree_trajectory_with_attached_object(trajectory)[source]#

Return type:: NamoPolygonExecutionResult

non_physical_execute_trajectory(body_name, trajectory)#

Execute a trajectory without physical simulation. This is done by manually setting the positions of the bodies.

Parameters:

body_name (str) – the name of the body to be moved.
trajectory (List[Tuple[float, float]]) – a list of positions to be set to the body.

Return type:

None

render()#

Render the environment. If recording is enabled, the rendered frames will be stored in the recorded_frames list. Use start_recording() and stop_recording() to control the recording process.

The rendering is done by calling the debug_draw method of the PymunkWorld object.

Return type:: None

render_and_display(render_mode='image', trajectory=None)#

Parameters:

render_mode (str | None)
trajectory (List[Any] | None)

Return type:

None

reset(**kwargs)#

reset_scene()[source]#

reset_world()#

Return type:: PymunkWorld

set_additional_step_callable(fn)#

Parameters:: fn (Callable[[PymunkDefaultEnv], None])

set_agent_pos(pos)[source]#

Parameters:: pos (Tuple[float, float])

set_current_attached_body(body)[source]#

Parameters:: body (Body | None)

set_current_selection(body)#

Parameters:: body (Body)

start_recording()#

step(steps, render_mode='image', callback=None)#

Step the simulation for a number of steps.

Parameters:

steps (int) – the number of steps to be executed.
render_mode (str | None) – the mode of rendering. Can be one of [‘image’, ‘human’, ‘pose’, ‘state’]. Set to None to disable any rendering.
callback (Callable[[], bool] | None) – a callback function that returns False to stop the simulation.

Return type:

List[Any]

stop_recording()#

RENDER_MODES = ['image', 'human', 'pose', 'state']#

SCREENSHOT_TITLE = 'screenshot'#

TITLE = 'Pymunk Default'#

property agent#

property agent_goal: Tuple[float, float]#

property agent_pos#

property agent_start: Tuple[float, float]#

property obstacles: Sequence[Body]#