unreal.LearningAgentsRecord

class unreal.LearningAgentsRecord(name: str = '', schema_name: Name = 'None', step_num: int = 0, observation_dim_num: int = 0, action_dim_num: int = 0, observation_compatibility_hash: int = 0, action_compatibility_hash: int = 0, tags: GameplayTagContainer = Ellipsis, observation_data: None = [], action_data: None = [])

Bases: StructBase

A single recording of a series of observations and actions.

C++ Source:

  • Plugin: LearningAgents

  • Module: LearningAgentsTraining

  • File: LearningAgentsRecording.h

Editor Properties: (see get_editor_property/set_editor_property)

  • action_compatibility_hash (int32): [Read-Only] The compatibility hash for the recorded action vectors

  • action_data (Array[float]): [Read-Write] Action data

  • action_dim_num (int32): [Read-Only] The number of dimensions in the action vector for this record

  • name (str): [Read-Write] Record’s name.

  • observation_compatibility_hash (int32): [Read-Only] The compatibility hash for the recorded observation vectors

  • observation_data (Array[float]): [Read-Write] Observation data

  • observation_dim_num (int32): [Read-Only] The number of dimensions in the observation vector for this record

  • schema_name (Name): [Read-Write] Schema’s name. Must correspond with a schema in the recording.

  • step_num (int32): [Read-Only] The number of observations and actions recorded.

  • tags (GameplayTagContainer): [Read-Write] The tags to be referenced in the imitation trainer.

property action_compatibility_hash: int

[Read-Only] The compatibility hash for the recorded action vectors

Type:

(int32)

property action_data: None

[Read-Only] Action data

Type:

(Array[float])

property action_dim_num: int

[Read-Only] The number of dimensions in the action vector for this record

Type:

(int32)

property name: str

[Read-Write] Record’s name.

Type:

(str)

property observation_compatibility_hash: int

[Read-Only] The compatibility hash for the recorded observation vectors

Type:

(int32)

property observation_data: None

[Read-Only] Observation data

Type:

(Array[float])

property observation_dim_num: int

[Read-Only] The number of dimensions in the observation vector for this record

Type:

(int32)

property schema_name: Name

[Read-Write] Schema’s name. Must correspond with a schema in the recording.

Type:

(Name)

property step_num: int

[Read-Only] The number of observations and actions recorded.

Type:

(int32)

property tags: GameplayTagContainer

[Read-Write] The tags to be referenced in the imitation trainer.

Type:

(GameplayTagContainer)