mjlab.managers

Environment managers for actions, observations, rewards, terminations, commands, and curriculum.

Environment managers.

Classes:

CommandManager
CommandTerm	Base class for command terms.
NullCommandManager	Placeholder for absent command manager that safely no-ops all operations.
CurriculumManager
NullCurriculumManager	Placeholder for absent curriculum manager that safely no-ops all operations.
CommandTermCfg	Configuration for a command generator term.

class mjlab.managers.CommandManager

Bases: ManagerBase

Methods:

__init__(cfg, env)
debug_vis(visualizer)
get_active_iterable_terms(env_idx)
reset(env_ids)	Resets the manager and returns logging info for the current step.
compute(dt)
get_command(name)
get_term(name)
get_term_cfg(name)

Attributes:

active_terms

__init__(cfg: dict[str, CommandTermCfg], env: ManagerBasedRlEnv)

debug_vis(visualizer: DebugVisualizer) → None

property active_terms: list[str]

get_active_iterable_terms(env_idx: int) → Sequence[tuple[str, Sequence[float]]]

reset(env_ids: Tensor | None) → dict[str, Tensor]

Resets the manager and returns logging info for the current step.

compute(dt: float)

get_command(name: str) → Tensor

get_term(name: str) → CommandTerm

get_term_cfg(name: str) → CommandTermCfg

class mjlab.managers.CommandTerm

Bases: ManagerTermBase

Base class for command terms.

Methods:

__init__(cfg, env)
debug_vis(visualizer)
reset(env_ids)	Resets the manager term.
compute(dt)

Attributes:

command

__init__(cfg: CommandTermCfg, env: ManagerBasedRlEnv)

debug_vis(visualizer: DebugVisualizer) → None

abstract property command

reset(env_ids: Tensor | slice | None) → dict[str, float]

Resets the manager term.

compute(dt: float) → None

class mjlab.managers.NullCommandManager

Bases: object

Placeholder for absent command manager that safely no-ops all operations.

Methods:

__init__()
debug_vis(visualizer)
get_active_iterable_terms(env_idx)
reset([env_ids])
compute(dt)
get_command(name)
get_term(name)
get_term_cfg(name)

__init__()

debug_vis(visualizer: DebugVisualizer) → None

get_active_iterable_terms(env_idx: int) → Sequence[tuple[str, Sequence[float]]]

reset(env_ids: Tensor | None = None) → dict[str, Tensor]

compute(dt: float) → None

get_command(name: str) → None

get_term(name: str) → None

get_term_cfg(name: str) → None

class mjlab.managers.CurriculumManager

Bases: ManagerBase

Methods:

__init__(cfg, env)
get_term_cfg(term_name)
get_active_iterable_terms(env_idx)
reset([env_ids])	Resets the manager and returns logging info for the current step.
compute([env_ids])

Attributes:

active_terms

__init__(cfg: dict[str, CurriculumTermCfg], env: ManagerBasedRlEnv)

property active_terms: list[str]

get_term_cfg(term_name: str) → CurriculumTermCfg

get_active_iterable_terms(env_idx: int) → Sequence[tuple[str, Sequence[float]]]

reset(env_ids: Tensor | slice | None = None) → dict[str, float]

Resets the manager and returns logging info for the current step.

compute(env_ids: Tensor | slice | None = None)

class mjlab.managers.NullCurriculumManager

Bases: object

Placeholder for absent curriculum manager that safely no-ops all operations.

Methods:

__init__()
get_active_iterable_terms(env_idx)
reset([env_ids])
compute([env_ids])

__init__()

get_active_iterable_terms(env_idx: int) → Sequence[tuple[str, Sequence[float]]]

reset(env_ids: Tensor | None = None) → dict[str, float]

compute(env_ids: Tensor | None = None) → None

class mjlab.managers.CommandTermCfg

Bases: object

Configuration for a command generator term.

Attributes:

class_type
resampling_time_range
debug_vis

Methods:

__init__(*, class_type, resampling_time_range)

class_type: type[CommandTerm]

resampling_time_range: tuple[float, float]

debug_vis: bool = False

__init__(*, class_type: type[CommandTerm], resampling_time_range: tuple[float, float], debug_vis: bool = False) → None