unreal.LearningAgentsActions

class unreal.LearningAgentsActions(outer: Object | None = None, name: Name | str = 'None')

Bases: BlueprintFunctionLibrary

Learning Agents Actions

C++ Source:

  • Plugin: LearningAgents

  • Module: LearningAgents

  • File: LearningAgentsActions.h

classmethod get_angle_action(object, element, relative_angle=0.000000, angle_scale=90.000000, tag='AngleAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_angle_location=[0.000000, 0.000000, 0.000000], visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → float or None

Get the value for an angle action. Returned angle is in degrees.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_angle (float) – The relative angle to transform the angle by.

  • angle_scale (float) – The scale used to control the overall magnitude of the outputted scale action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_angle_location (Vector) – A location for the visual logger to display the angle in the world.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_angle (float): The output angle value.

Return type:

float or None

classmethod get_angle_action_radians(object, element, relative_angle=0.000000, angle_scale=1.570796, tag='AngleAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_angle_location=[0.000000, 0.000000, 0.000000], visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → float or None

Get the value for an angle action. Returned angle is in radians.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_angle (float) – The relative angle to transform the angle by.

  • angle_scale (float) – The scale used to control the overall magnitude of the outputted scale action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_angle_location (Vector) – A location for the visual logger to display the angle in the world.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_angle (float): The output angle value.

Return type:

float or None

classmethod get_bitmask_action(object, element, enum, tag='BitmaskAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → int32 or None

Get the bitmask value of a bitmask action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • enum (Enum) – The Enum

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_bitmask_value (int32): The output bitmask value.

Return type:

int32 or None

classmethod get_bool_action(object, element, tag='BoolAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → bool or None

Get the value for a bool action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_value (bool): The output bool value.

Return type:

bool or None

classmethod get_continuous_action(object, element, tag='ContinuousAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Array[float] or None

Get the values of a continuous action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_values (Array[float]): The output values.

Return type:

Array[float] or None

classmethod get_continuous_action_num(object, element, tag='ContinuousAction') → int32 or None

Get the number of values in a continuous action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_num (int32): The output number of values.

Return type:

int32 or None

classmethod get_direction_action(object, element, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], tag='DirectionAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_direction_location=[0.000000, 0.000000, 0.000000], visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_arrow_length=100.000000, visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Vector or None

Get the value for a direction action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_transform (Transform) – The relative transform to transform the direction by.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_direction_location (Vector) – A location for the visual logger to display the direction in the world.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_arrow_length (float) – The length of the arrow to display to represent the direction.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_direction (Vector): The output direction value.

Return type:

Vector or None

classmethod get_either_action(object, element, tag='EitherAction') → (out_either=LearningAgentsEitherAction, out_element=LearningAgentsActionObjectElement) or None

Get the sub-action of an either action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_either (LearningAgentsEitherAction): The output either specifier.

out_element (LearningAgentsActionObjectElement): The output sub-action.

Return type:

tuple or None

classmethod get_encoding_action(object, element, tag='EncodingAction') → LearningAgentsActionObjectElement or None

Get the sub-action of an encoding action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_element (LearningAgentsActionObjectElement): The output sub-action.

Return type:

LearningAgentsActionObjectElement or None

classmethod get_enum_action(object, element, enum, tag='EnumAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → uint8 or None

Get the enum value of an enum action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • enum (Enum) – The Enum

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_enum_value (uint8): The output enum value.

Return type:

uint8 or None

classmethod get_exclusive_discrete_action(object, element, tag='DiscreteExclusiveAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → int32 or None

Get the index for an exclusive discrete action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_index (int32): The output index.

Return type:

int32 or None

classmethod get_exclusive_union_action(object, element, tag='ExclusiveUnionAction') → (out_element_name=Name, out_element=LearningAgentsActionObjectElement) or None

Get the chosen sub-action for an exclusive union action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_element_name (Name):

out_element (LearningAgentsActionObjectElement): The output sub-action.

Return type:

tuple or None

classmethod get_float_action(object, element, float_scale=1.000000, tag='FloatAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → float or None

Get the value for a float action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • float_scale (float) – The scale used to control the overall magnitude of the outputted float action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_value (float): The output float value.

Return type:

float or None

classmethod get_inclusive_discrete_action(object, element, tag='DiscreteInclusiveAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Array[int32] or None

Get the indices for an inclusive discrete action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_indices (Array[int32]): The output indices.

Return type:

Array[int32] or None

classmethod get_inclusive_discrete_action_num(object, element, tag='DiscreteInclusiveAction') → int32 or None

Get the number of indices for an inclusive discrete action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_num (int32): The output number of indices.

Return type:

int32 or None

classmethod get_inclusive_union_action(object, element, tag='InclusiveUnionAction') → Map[Name, LearningAgentsActionObjectElement] or None

Get the chosen sub-actions for an inclusive union action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_elements (Map[Name, LearningAgentsActionObjectElement]): The output sub-actions.

Return type:

Map[Name, LearningAgentsActionObjectElement] or None

classmethod get_inclusive_union_action_num(object, element, tag='InclusiveUnionAction') → int32 or None

Get the number of sub-actions for an inclusive union action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_num (int32): The output number of sub-actions.

Return type:

int32 or None

classmethod get_inclusive_union_action_to_arrays(object, element, tag='InclusiveUnionAction') → (out_element_names=Array[Name], out_elements=Array[LearningAgentsActionObjectElement]) or None

Get the chosen sub-actions for an inclusive union action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_element_names (Array[Name]): The output sub-action names.

out_elements (Array[LearningAgentsActionObjectElement]): The output sub-actions.

Return type:

tuple or None

classmethod get_location_action(object, element, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], location_scale=100.000000, tag='LocationAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Vector or None

Get the value for a location action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_transform (Transform) – The relative transform to transform the location by.

  • location_scale (float) – The scale used to control the overall magnitude of the outputted location action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_location (Vector): The output location value.

Return type:

Vector or None

classmethod get_null_action(object, element, tag='NullAction') → bool

Get a null action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

Return type:

bool

classmethod get_optional_action(object, element, tag='OptionalAction') → (out_option=LearningAgentsOptionalAction, out_element=LearningAgentsActionObjectElement) or None

Get the sub-action of an option action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_option (LearningAgentsOptionalAction): The output optional specifier.

out_element (LearningAgentsActionObjectElement): The output sub-action.

Return type:

tuple or None

classmethod get_pair_action(object, element, tag='PairAction') → (out_key=LearningAgentsActionObjectElement, out_value=LearningAgentsActionObjectElement) or None

Get the sub-actions of a pair action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_key (LearningAgentsActionObjectElement): The output key sub-element.

out_value (LearningAgentsActionObjectElement): The output value sub-element.

Return type:

tuple or None

classmethod get_rotation_action(object, element, relative_rotation=[0.000000, 0.000000, 0.000000], rotation_scale=90.000000, tag='RotationAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_rotation_location=[0.000000, 0.000000, 0.000000], visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Rotator or None

Get the value for a rotation action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_rotation (Rotator) – The relative rotation to transform the rotation by.

  • rotation_scale (float) – The scale used to control the overall magnitude of the outputted rotation action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_rotation_location (Vector) – A location for the visual logger to display the rotation in the world.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_rotation (Rotator): The output rotation value.

Return type:

Rotator or None

classmethod get_rotation_action_as_quat(object, element, relative_rotation, rotation_scale=90.000000, tag='RotationAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_rotation_location=[0.000000, 0.000000, 0.000000], visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Quat or None

Get the value for a rotation action as a quaternion.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_rotation (Quat) – The relative rotation to transform the rotation by.

  • rotation_scale (float) – The scale used to control the overall magnitude of the outputted rotation action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_rotation_location (Vector) – A location for the visual logger to display the rotation in the world.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_rotation (Quat): The output rotation value.

Return type:

Quat or None

classmethod get_scale_action(object, element, relative_scale=[1.000000, 1.000000, 1.000000], scale=1.000000, tag='ScaleAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Vector or None

Get the value for a scale action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_scale (Vector) – The relative scale to transform the scale by.

  • scale (float) – The scale used to control the overall magnitude of the outputted scale action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_scale (Vector): The output scale value.

Return type:

Vector or None

classmethod get_static_array_action(object, element, tag='StaticArrayAction') → Array[LearningAgentsActionObjectElement] or None

Get the entries of a static array action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_elements (Array[LearningAgentsActionObjectElement]): The output sub-elements.

Return type:

Array[LearningAgentsActionObjectElement] or None

classmethod get_static_array_action_num(object, element, tag='StaticArrayAction') → int32 or None

Get the number of entries in a static array action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_num (int32): The output number of entries.

Return type:

int32 or None

classmethod get_struct_action(object, element, tag='StructAction') → Map[Name, LearningAgentsActionObjectElement] or None

Get the sub-actions for a struct action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_elements (Map[Name, LearningAgentsActionObjectElement]): The output sub-actions.

Return type:

Map[Name, LearningAgentsActionObjectElement] or None

classmethod get_struct_action_num(object, element, tag='StructAction') → int32 or None

Get the number of sub-actions for a struct action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_num (int32): The output number of sub-actions.

Return type:

int32 or None

classmethod get_struct_action_to_arrays(object, element, tag='StructAction') → (out_element_names=Array[Name], out_elements=Array[LearningAgentsActionObjectElement]) or None

Get the sub-actions for a struct action.

Parameters:
Returns:

true if the provided Element is the correct type, otherwise false.

out_element_names (Array[Name]): The output sub-action names.

out_elements (Array[LearningAgentsActionObjectElement]): The output sub-actions.

Return type:

tuple or None

classmethod get_transform_action(object, element, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], location_scale=100.000000, rotation_scale=1.000000, scale_scale=1.000000, tag='TransformAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Transform or None

Get the value for a transform action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_transform (Transform) – The relative transform.

  • location_scale (float) – The scale used to control the overall magnitude of the outputted location action.

  • rotation_scale (float) – The scale used to control the overall magnitude of the outputted rotation action.

  • scale_scale (float) – The scale used to control the overall magnitude of the outputted scale action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_transform (Transform): The output transform value.

Return type:

Transform or None

classmethod get_velocity_action(object, element, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], velocity_scale=200.000000, tag='VelocityAction', visual_logger_enabled=False, visual_logger_listener=None, visual_logger_agent_id=-1, visual_logger_velocity_location=[0.000000, 0.000000, 0.000000], visual_logger_location=[0.000000, 0.000000, 0.000000], visual_logger_color=[0.000000, 0.000000, 1.000000, 1.000000]) → Vector or None

Get the value for a velocity action.

Parameters:

  • object (LearningAgentsActionObject) – The Action Object

  • element (LearningAgentsActionObjectElement) – The Action Object Element

  • relative_transform (Transform) – The relative transform to transform the velocity by.

  • velocity_scale (float) – The scale used to control the overall magnitude of the outputted velocity action.

  • tag (Name) – The tag of the corresponding action. Must match the tag given during Specify.

  • visual_logger_enabled (bool) – When true, debug data will be sent to the visual logger.

  • visual_logger_listener (LearningAgentsManagerListener) – The listener object which is making this action. This must be set to use logging.

  • visual_logger_agent_id (int32) – The agent id associated with this action.

  • visual_logger_velocity_location (Vector) – A location for the visual logger to display the velocity in the world.

  • visual_logger_location (Vector) – A location for the visual logger information in the world.

  • visual_logger_color (LinearColor) – The color for the visual logger display.

Returns:

true if the provided Element is the correct type, otherwise false.

out_velocity (Vector): The output velocity value.

Return type:

Vector or None

classmethod log_action(object, element) → None

Logs an Action Object Element. Useful for debugging.

Parameters:
classmethod make_angle_action(object, angle, relative_angle=0.000000, angle_scale=90.000000, tag='AngleAction') → LearningAgentsActionObjectElement

Make Angle Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_angle_action_radians(object, angle, relative_angle=0.000000, angle_scale=1.570796, tag='AngleAction') → LearningAgentsActionObjectElement

Make Angle Action Radians

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_bitmask_action(object, enum, bitmask_value, tag='BitmaskAction') → LearningAgentsActionObjectElement

Make Bitmask Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_bool_action(object, value, tag='BoolAction') → LearningAgentsActionObjectElement

Make Bool Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_continuous_action(object, values, tag='ContinuousAction') → LearningAgentsActionObjectElement

Make Continuous Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_direction_action(object, direction, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], tag='DirectionAction') → LearningAgentsActionObjectElement

Make Direction Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_either_a_action(object, a, tag='EitherAction') → LearningAgentsActionObjectElement

Make Either AAction

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_either_action(object, element, either, tag='EitherAction') → LearningAgentsActionObjectElement

Make Either Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_either_b_action(object, b, tag='EitherAction') → LearningAgentsActionObjectElement

Make Either BAction

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_encoding_action(object, element, tag='EncodingAction') → LearningAgentsActionObjectElement

Make Encoding Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_enum_action(object, enum, enum_value, tag='EnumAction') → LearningAgentsActionObjectElement

Make Enum Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_exclusive_discrete_action(object, index, tag='DiscreteExclusiveAction') → LearningAgentsActionObjectElement

Make Exclusive Discrete Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_exclusive_union_action(object, element_name, element, tag='ExclusiveUnionAction') → LearningAgentsActionObjectElement

Make Exclusive Union Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_float_action(object, value, float_scale=1.000000, tag='FloatAction') → LearningAgentsActionObjectElement

Make Float Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_inclusive_discrete_action(object, indices, tag='DiscreteInclusiveAction') → LearningAgentsActionObjectElement

Make Inclusive Discrete Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_inclusive_union_action(object, elements, tag='InclusiveUnionAction') → LearningAgentsActionObjectElement

Make Inclusive Union Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_inclusive_union_action_from_arrays(object, element_names, elements, tag='InclusiveUnionAction') → LearningAgentsActionObjectElement

Make Inclusive Union Action from Arrays

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_location_action(object, location, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], location_scale=100.000000, tag='LocationAction') → LearningAgentsActionObjectElement

Make Location Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_null_action(object, tag='NullAction') → LearningAgentsActionObjectElement

Make Null Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_optional_action(object, element, option, tag='OptionalAction') → LearningAgentsActionObjectElement

Make Optional Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_optional_null_action(object, tag='OptionalAction') → LearningAgentsActionObjectElement

Make Optional Null Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_optional_valid_action(object, element, tag='OptionalAction') → LearningAgentsActionObjectElement

Make Optional Valid Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_pair_action(object, key, value, tag='PairAction') → LearningAgentsActionObjectElement

Make Pair Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_rotation_action(object, rotation, relative_rotation=[0.000000, 0.000000, 0.000000], rotation_scale=90.000000, tag='RotationAction') → LearningAgentsActionObjectElement

Make Rotation Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_rotation_action_from_quat(object, rotation, relative_rotation, rotation_scale=90.000000, tag='RotationAction') → LearningAgentsActionObjectElement

Make Rotation Action from Quat

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_scale_action(object, scale, relative_scale=[1.000000, 1.000000, 1.000000], tag='ScaleAction') → LearningAgentsActionObjectElement

Make Scale Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_static_array_action(object, elements, tag='StaticArrayAction') → LearningAgentsActionObjectElement

Make Static Array Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_struct_action(object, elements, tag='StructAction') → LearningAgentsActionObjectElement

Make Struct Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_struct_action_from_arrays(object, element_names, elements, tag='StructAction') → LearningAgentsActionObjectElement

Make Struct Action from Arrays

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_transform_action(object, transform, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], location_scale=100.000000, tag='TransformAction') → LearningAgentsActionObjectElement

Make Transform Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod make_velocity_action(object, velocity, relative_transform=[[0.000000, 0.000000, 0.000000], [-0.000000, 0.000000, 0.000000], [1.000000, 1.000000, 1.000000]], velocity_scale=200.000000, tag='VelocityAction') → LearningAgentsActionObjectElement

Make Velocity Action

Parameters:
Return type:

LearningAgentsActionObjectElement

classmethod specify_angle_action(schema, tag='AngleAction') → LearningAgentsActionSchemaElement

Specifies a new angle action. This represents an action which is an angle sampled from a Gaussian distribution centered around zero.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_bitmask_action(schema, enum, prior_probabilities, tag='BitmaskAction') → LearningAgentsActionSchemaElement

Specifies a new bitmask action. This represents an action which is an inclusive choice from entries of an Enum, sampled from a Bernoulli distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • enum (Enum) – The Enum type.

  • prior_probabilities (Map[uint8, float]) – The prior probabilities of each enum element. Can be left empty to use a probability of 0.5 for each element.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_bitmask_action_from_array(schema, enum, prior_probabilities, tag='BitmaskAction') → LearningAgentsActionSchemaElement

Specifies a new bitmask action. This represents an action which is an inclusive choice from entries of an Enum, sampled from a Bernoulli distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • enum (Enum) – The Enum type.

  • prior_probabilities (Array[float]) – The prior probabilities of each enum element. Can be left empty to use a probability of 0.5 for each element.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_bool_action(schema, prior_probability=0.500000, tag='BoolAction') → LearningAgentsActionSchemaElement

Specifies a new bool action. This represents an action which is either true or false.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • prior_probability (float) – The prior probability of this action being true.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_continuous_action(schema, size, tag='ContinuousAction') → LearningAgentsActionSchemaElement

Specifies a new continuous action. This represents an action made up of several float values sampled from a Gaussian distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • size (int32) – The number of float values in the action.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_direction_action(schema, tag='DirectionAction') → LearningAgentsActionSchemaElement

Specifies a new direction action. This represents an action which is a direction sampled from a Gaussian distribution and normalized.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_either_action(schema, a, b, prior_probability_of_a=0.500000, tag='EitherAction') → LearningAgentsActionSchemaElement

Specifies a new either action. This represents an action which is either action A or action B.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_encoding_action(schema, element, encoding_size=128, hidden_layer_num=1, activation_function=LearningAgentsActivationFunction.ELU, tag='EncodingAction') → LearningAgentsActionSchemaElement

Specifies a new encoding action. This represents an action which will be a decoding of another sub-action using a small neural network.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • element (LearningAgentsActionSchemaElement) – The sub-action.

  • encoding_size (int32) – The encoding size used to decode this sub-action.

  • hidden_layer_num (int32) – The number of hidden layers used to decode this sub-action.

  • activation_function (LearningAgentsActivationFunction) – The activation function used to decode this sub-action.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_enum_action(schema, enum, prior_probabilities, tag='EnumAction') → LearningAgentsActionSchemaElement

Specifies a new enum action. This represents an action which is an exclusive choice from entries of an Enum, sampled from a Categorical distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • enum (Enum) – The Enum type.

  • prior_probabilities (Map[uint8, float]) – The prior probabilities of each enum element. Can be left empty to use a uniform distribution over elements. Should sum to one.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_enum_action_from_array(schema, enum, prior_probabilities, tag='EnumAction') → LearningAgentsActionSchemaElement

Specifies a new enum action. This represents an action which is an exclusive choice from entries of an Enum, sampled from a Categorical distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • enum (Enum) – The Enum type.

  • prior_probabilities (Array[float]) – The prior probabilities of each enum element. Can be left empty to use a uniform distribution over elements. Should sum to one.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_exclusive_discrete_action(schema, size, prior_probabilities, tag='DiscreteExclusiveAction') → LearningAgentsActionSchemaElement

Specifies a new exclusive discrete action. This represents an action which is an exclusive choice from a number of discrete options, sampled from a Categorical distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • size (int32) – The number of discrete options in the action.

  • prior_probabilities (Array[float]) – The prior probabilities of each option. Can be left empty to use a uniform distribution over options. Should sum to one.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_exclusive_union_action(schema, elements, prior_probabilities, tag='ExclusiveUnionAction') → LearningAgentsActionSchemaElement

Specifies a new exclusive union action. This represents an action which is an exclusive choice from a number of named sub-actions, sampled from a Categorical distribution.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_exclusive_union_action_from_arrays(schema, element_names, elements, prior_probabilities, tag='ExclusiveUnionAction') → LearningAgentsActionSchemaElement

Specifies a new exclusive union action. This represents an action which is an exclusive choice from a number of named sub-actions, sampled from a Categorical distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • element_names (Array[Name]) – The names of the sub-actions.

  • elements (Array[LearningAgentsActionSchemaElement]) – The corresponding sub-actions. Must be the same size as ElementNames.

  • prior_probabilities (Array[float]) – The prior probabilities of each option. Can be left empty to use a uniform distribution over options. Should sum to one.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_float_action(schema, tag='FloatAction') → LearningAgentsActionSchemaElement

Specifies a new float action. This represents an action which is a single float sampled from a Gaussian distribution. It can be used as a catch-all for situations where a type-specific action does not exist.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_inclusive_discrete_action(schema, size, prior_probabilities, tag='DiscreteInclusiveAction') → LearningAgentsActionSchemaElement

Specifies a new inclusive discrete action. This represents an action which is an inclusive choice from a number of discrete options, sampled from a Bernoulli distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • size (int32) – The number of discrete options in the action.

  • prior_probabilities (Array[float]) – The prior probabilities of each option. Can be left empty to use a probability of 0.5 for each option.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_inclusive_union_action(schema, elements, prior_probabilities, tag='InclusiveUnionAction') → LearningAgentsActionSchemaElement

Specifies a new inclusive union action. This represents an action which is an inclusive choice from a number of named sub-actions, sampled from a Bernoulli distribution.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_inclusive_union_action_from_arrays(schema, element_names, elements, prior_probabilities, tag='InclusiveUnionAction') → LearningAgentsActionSchemaElement

Specifies a new inclusive union action. This represents an action which is an inclusive choice from a number of named sub-actions, sampled from a Bernoulli distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • element_names (Array[Name]) – The names of the sub-actions.

  • elements (Array[LearningAgentsActionSchemaElement]) – The corresponding sub-actions. Must be the same size as ElementNames.

  • prior_probabilities (Array[float]) – The prior probabilities of each option. Can be left empty to use a probability of 0.5 for each option.

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_location_action(schema, tag='LocationAction') → LearningAgentsActionSchemaElement

Specifies a new location action. This represents an action which is a location sampled from a Gaussian distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_null_action(schema, tag='NullAction') → LearningAgentsActionSchemaElement

Specifies a new null action. This represents an empty action and can be useful when an action is needed which does nothing.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_optional_action(schema, element, prior_probability=0.500000, tag='OptionalAction') → LearningAgentsActionSchemaElement

Specifies a new optional action. This represents an action which may or may not be generated.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_pair_action(schema, key, value, tag='PairAction') → LearningAgentsActionSchemaElement

Specifies a new pair action. This represents an action which is made up of a key and value sub-actions.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_rotation_action(schema, tag='RotationAction') → LearningAgentsActionSchemaElement

Specifies a new rotation action. This represents an action which is a rotation sampled from a Gaussian distribution in the angle-axis space.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_scale_action(schema, tag='ScaleAction') → LearningAgentsActionSchemaElement

Specifies a new scale action. This represents an action which is a scale sampled from a Gaussian distribution in the log space.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_static_array_action(schema, element, num, tag='StaticArrayAction') → LearningAgentsActionSchemaElement

Specifies a new static array action. This represents an action which is a fixed sized array of some sub-action.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_struct_action(schema, elements, tag='StructAction') → LearningAgentsActionSchemaElement

Specifies a new struct action. This represents an action which is made up of a number of named sub-actions.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_struct_action_from_arrays(schema, element_names, elements, tag='StructAction') → LearningAgentsActionSchemaElement

Specifies a new struct action. This represents an action which is made up of a number of named sub-actions.

Parameters:
Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_transform_action(schema, tag='TransformAction') → LearningAgentsActionSchemaElement

Specifies a new transform action.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod specify_velocity_action(schema, tag='VelocityAction') → LearningAgentsActionSchemaElement

Specifies a new velocity action. This represents an action which is a velocity sampled from a Gaussian distribution.

Parameters:

  • schema (LearningAgentsActionSchema) – The Action Schema

  • tag (Name) – The tag of this new action. Used during action object validation and debugging.

Returns:

The newly created action schema element.

Return type:

LearningAgentsActionSchemaElement

classmethod validate_action_object_matches_schema(schema, schema_element, object, object_element) → bool

Validates that the given action object matches the schema. Will log errors on objects that don’t match.

Parameters:
Returns:

true if the object matches the schema

Return type:

bool