Class ReacherEnvV5
- Namespace
- Gym.Environments.Envs.MuJoCo
- Assembly
- Gym.Environments.dll
Reacher-v5: a two-joint planar arm that reaches for a random target, driven by 2 joint torques in [-1, 1]. The
reward charges the fingertip's distance to the target and the squared torques; the episode never ends on its own
(make's 50-step limit truncates it). Byte for byte what Gymnasium's Reacher-v5 computes on MuJoCo 4.0.0
(DR-61, DR-64).
[PyName("ReacherEnv")]
public class ReacherEnvV5 : MujocoEnv, IEnv, IDisposable, IDynamicAttributes
- Inheritance
-
ReacherEnvV5
- Implements
- Inherited Members
- Extension Methods
Remarks
Observation (float64, 10 values): the cosines and sines of the two arm angles, the target's two coordinates,
the arm's two angular velocities, and the fingertip's offset from the target in x and y. Reward:
-|fingertip - target| * reward_dist_weight - sum(action**2) * reward_control_weight, measured after the
step, the control term in the action's dtype. Info: reward_dist, reward_ctrl; the reset's info
is empty. The reset draws the target uniformly in the disc of radius 0.2 by rejection (Goal).
Not thread-safe. Reference: gymnasium/envs/mujoco/reacher_v5.py.
Its info reads typed (DR-91): ReacherInfoV5 info = step.Info; (ReacherInfoV5), with the dict's own values still behind the indexer.
Constructors
ReacherEnvV5(ReacherEnvV5Options)
Builds the instance from its options record: the properties set, the constructor's defaults for the rest.
public ReacherEnvV5(ReacherEnvV5Options options)
Parameters
optionsReacherEnvV5OptionsThe options.
Exceptions
- ArgumentNullException
Thrown when
optionsis null.
ReacherEnvV5(RenderMode, string, int, IReadOnlyDictionary<string, object?>?, double, double?, int, int, int?, string?, int, IReadOnlyDictionary<int, bool>?, EnvViewerFactory?)
Builds the env with the render mode given as the RenderMode enum (DR-84, DR-93), as
ReacherEnvV5(string, int, IReadOnlyDictionary<string, object?>?, double, double?, string?, int, int, int?, string?, int, IReadOnlyDictionary<int, bool>?, EnvViewerFactory?)
builds it from the mode's string: new ReacherEnvV5(RenderMode.RgbArray) builds what
new ReacherEnvV5(renderMode: "rgb_array") builds, the member reaching the env as Gymnasium's string
(Name(RenderMode)), which its RenderMode property then holds.
public ReacherEnvV5(RenderMode renderMode, string xmlFile = "reacher.xml", int frameSkip = 2, IReadOnlyDictionary<string, object?>? defaultCameraConfig = null, double rewardDistWeight = 1, double? rewardControlWeight = null, int width = 480, int height = 480, int? cameraId = null, string? cameraName = null, int maxGeom = 1000, IReadOnlyDictionary<int, bool>? visualOptions = null, EnvViewerFactory? viewerFactory = null)
Parameters
renderModeRenderModeThe render mode, which the env must list in its
metadata["render_modes"]as for the string; it behaves exactly as its Gymnasium string does in the other constructor.xmlFilestringThe model: a bundled file name (
xml_file, DR-63).frameSkipintPhysics steps per env step (
frame_skip).defaultCameraConfigIReadOnlyDictionary<string, object>The free camera's settings (
default_camera_config); null is Gymnasium's default (trackbodyid0), and an empty dictionary leaves MuJoCo's camera as it is.rewardDistWeightdoubleThe weight of the fingertip's distance to the target (
reward_dist_weight; Gymnasium's default is the int 1).rewardControlWeightdouble?The weight of the squared torques (
reward_control_weight); null is Gymnasium's default, the Python int 1, which keeps an integer action's cost annp.int64(a given weight is a Python float).widthintThe rendered frame's width (
width).heightintThe rendered frame's height (
height).cameraIdint?The camera to render from (
camera_id).cameraNamestringThe camera to render from by name (
camera_name).maxGeomintThe most geoms a frame draws (
max_geom).visualOptionsIReadOnlyDictionary<int, bool>MuJoCo visual flags to switch (
visual_options).viewerFactoryEnvViewerFactoryGym.NET's
viewer_factory(DR-20) forhumanmode.
Remarks
The mode is required and comes first, so new ReacherEnvV5() keeps meaning the other constructor, without a
mode; leave the mode out there for no rendering. Every other parameter means what it means there.
Exceptions
- ArgumentOutOfRangeException
Thrown for a value RenderMode does not declare (a cast integer such as
(RenderMode)99).- NotSupportedException
xmlFileis not a bundled model.- OverflowError
A frame skip of 0 (Python's
OverflowErrorderiving the frame rate).- ValueError
Both a camera id and a camera name.
ReacherEnvV5(string, int, IReadOnlyDictionary<string, object?>?, double, double?, string?, int, int, int?, string?, int, IReadOnlyDictionary<int, bool>?, EnvViewerFactory?)
Creates the env (Gymnasium's ReacherEnv.init): the model, its spaces and the frame rate the time step gives.
public ReacherEnvV5(string xmlFile = "reacher.xml", int frameSkip = 2, IReadOnlyDictionary<string, object?>? defaultCameraConfig = null, double rewardDistWeight = 1, double? rewardControlWeight = null, string? renderMode = null, int width = 480, int height = 480, int? cameraId = null, string? cameraName = null, int maxGeom = 1000, IReadOnlyDictionary<int, bool>? visualOptions = null, EnvViewerFactory? viewerFactory = null)
Parameters
xmlFilestringThe model: a bundled file name (
xml_file, DR-63).frameSkipintPhysics steps per env step (
frame_skip).defaultCameraConfigIReadOnlyDictionary<string, object>The free camera's settings (
default_camera_config); null is Gymnasium's default (trackbodyid0), and an empty dictionary leaves MuJoCo's camera as it is.rewardDistWeightdoubleThe weight of the fingertip's distance to the target (
reward_dist_weight; Gymnasium's default is the int 1).rewardControlWeightdouble?The weight of the squared torques (
reward_control_weight); null is Gymnasium's default, the Python int 1, which keeps an integer action's cost annp.int64(a given weight is a Python float).renderModestringGymnasium's
render_mode:human,rgb_array,depth_array,rgbd_tupleor null.widthintThe rendered frame's width (
width).heightintThe rendered frame's height (
height).cameraIdint?The camera to render from (
camera_id).cameraNamestringThe camera to render from by name (
camera_name).maxGeomintThe most geoms a frame draws (
max_geom).visualOptionsIReadOnlyDictionary<int, bool>MuJoCo visual flags to switch (
visual_options).viewerFactoryEnvViewerFactoryGym.NET's
viewer_factory(DR-20) forhumanmode.
Exceptions
- NotSupportedException
xmlFileis not a bundled model.- OverflowError
A frame skip of 0 (Python's
OverflowErrorderiving the frame rate).- ValueError
Both a camera id and a camera name.
Properties
ClassMetadata
Gymnasium's class attribute metadata: the four render modes, the frame rate set per instance. A new dict on every read.
public static Metadata ClassMetadata { get; }
Property Value
Goal
The target the last reset drew (goal), float64 (2,), or null before the first
reset. Owned by the env, replaced by each reset: do not dispose it.
[NDBorrowed]
public NDArray? Goal { get; }
Property Value
RewardControlWeight
The weight of the squared torques (_reward_control_weight), read every step. Assigning one makes it a
Python float, as assigning 1.0 in Python would: an integer action's cost then becomes an np.float64.
[PyName("_reward_control_weight")]
public double RewardControlWeight { get; set; }
Property Value
RewardDistWeight
The weight of the fingertip's distance to the target (_reward_dist_weight), read every step.
[PyName("_reward_dist_weight")]
public double RewardDistWeight { get; set; }
Property Value
Methods
Dispose(bool)
Closes, then releases the last target array and everything MujocoEnv holds.
protected override void Dispose(bool disposing)
Parameters
ResetModel()
The episode's start (reset_model): positions uniform(-0.1, 0.1, nq) + init_qpos; the target drawn
as uniform(-0.2, 0.2, 2) until its norm is below 0.2, into the last two positions; velocities
init_qvel + uniform(-0.005, 0.005, nv) with the target's zeroed.
protected override NDArray ResetModel()
Returns
- NDArray
The first observation.
Step(NDArray)
Applies action for FrameSkip physics steps and reports the transition (step).
public override StepResult<NDArray> Step(NDArray action)
Parameters
actionNDArrayThe 2 torques, float32 (the action space's samples) or any real dtype.
Returns
- StepResult<NDArray>
The observation, the float64 reward,
terminated=False,truncated=Falseand the info.
Exceptions
- ArgumentNullException
actionis null.- ValueError
The action is not of shape
(2,), with Gymnasium's message.- NotSupportedException
An action of a dtype whose cost NumPy keeps in that dtype (float16, complex).
- KeyError
The model lacks the body
fingertiportarget(MuJoCo's message).