unreal.LearningAgentsRecording

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

Bases: DataAsset

Data asset representing an array of records.

C++ Source:

  • Plugin: LearningAgents

  • Module: LearningAgentsTraining

  • File: LearningAgentsRecording.h

Editor Properties: (see get_editor_property/set_editor_property)

  • new_schema_name (Name): [Read-Write] The schema name. Used in combination with the “AppendSchema” and “AppendRecording” buttons in the editor.

  • new_tag (GameplayTag): [Read-Write] The tag to apply to new records. Used in combination with the “AppendRecording” and “Append All Recordings from Folder” buttons in the editor.

  • recording_file (FilePath): [Read-Write] A recording file. Used in combination with the “AppendRecording” button in the editor.

  • recording_folder (DirectoryPath): [Read-Write] A folder containing .record files to load. Used in combination with the “Append All Recordings From Folder” button in the editor.

  • records (Array[LearningAgentsRecord]): [Read-Write] Set of records.

  • schema_file (FilePath): [Read-Write] A schema file. Used in combination with the “AppendSchema” button in the editor.

  • schemas (Map[Name, LearningAgentsSchema]): [Read-Write] Map of schemas.

append_recording_from_file(file, schema_name, tag) → None

Append to this recording from a file.

Parameters:
append_recording_to_asset(recording_asset) → None

Appends this recording to the given recording asset

Parameters:

recording_asset (LearningAgentsRecording)

append_schema_from_file(schema_name, schema) → None

Append schema to this recording from a file.

Parameters:
get_action_vector(record, step) -> (out_action_vector=Array[float], out_action_compatibility_hash=int32)

Get the Action Vector associated with a particular step of a given recording

Parameters:

  • record (int32) – Index of the record in the array of records.

  • step (int32) – Step of the recording

Returns:

out_action_vector (Array[float]): Output Action Vector

out_action_compatibility_hash (int32): Output Compatibility Hash for the given Action Vector

Return type:

tuple

get_observation_vector(record, step) -> (out_observation_vector=Array[float], out_observation_compatibility_hash=int32)

Get the Observation Vector associated with a particular step of a given recording

Parameters:

  • record (int32) – Index of the record in the array of records.

  • step (int32) – Step of the recording

Returns:

out_observation_vector (Array[float]): Output Observation Vector

out_observation_compatibility_hash (int32): Output Compatibility Hash for the given Observation Vector

Return type:

tuple

get_record_num() → int32

Get the number of records

Return type:

int32

get_record_step_num(record) → int32

Get the number of steps in a given record

Parameters:

record (int32)

Return type:

int32

load_recording_from_asset(recording_asset) → None

Loads this recording from the given recording asset

Parameters:

recording_asset (LearningAgentsRecording)

load_recording_from_file(file, schema_name, tag) → Array[LearningAgentsRecord]

Load this recording from a file.

Parameters:
Returns:

out_records (Array[LearningAgentsRecord]):

Return type:

Array[LearningAgentsRecord]

save_recording_to_asset(recording_asset) → None

Saves this recording to the given recording asset

Parameters:

recording_asset (LearningAgentsRecording)

save_recording_to_file(file) → None

Save this recording to a file.

Parameters:

file (FilePath)