Object3D

data_or_path: 'np.ndarray | str | pathlib.Path | TextIO | dict',
caption: 'str | None' = None, **kwargs: 'str | FileFormat3D | None'

Description

W&B class for 3D point clouds.

Args

• data_or_path:
• caption:
• kwargs:

Methods

method bind_to_run

self,
run: 'wandb.Run',
key: 'int | str',
step: 'int | str',
id_: 'int | str | None' = None,
ignore_copy_err: 'bool | None' = None

Bind this object to a particular Run.

Calling this function is necessary so that we have somewhere specific to put the file associated with this object, from which other Runs can refer to it.

Arguments

• run:
• key:
• step:
• id_:
• ignore_copy_err:

method captions

media_items: 'Sequence[Media]'

Arguments

• media_items:

method file_is_set

self

method from_file

data_or_path: 'TextIO | str',
file_type: 'FileFormat3D | None' = None

Initializes Object3D from a file or stream.

Arguments

• data_or_path: A path to a file or a TextIO stream.
• file_type: Specifies the data format passed to data_or_path. Required when data_or_path is a TextIO stream. This parameter is ignored if a file path is provided. The type is taken from the file extension.

method from_json

json_obj: 'dict',
source_artifact: 'Artifact'

Likely will need to override for any more complicated media objects.

Arguments

• json_obj:
• source_artifact:

method from_numpy

data: 'np.ndarray'

Initializes Object3D from a numpy array.

Arguments

• data: Each entry in the array will represent one point in the point cloud.

method from_point_cloud

points: 'Sequence[Point]',
boxes: 'Sequence[Box3D]',
vectors: 'Sequence[Vector3D] | None' = None,
point_cloud_type: 'PointCloudType' = 'lidar/beta'

Initializes Object3D from a python object.

Arguments

• points: The points in the point cloud.
• boxes: 3D bounding boxes for labeling the point cloud. Boxes are displayed in point cloud visualizations.
• vectors: Each vector is displayed in the point cloud visualization. Can be used to indicate directionality of bounding boxes. Defaults to None.
• point_cloud_type: At this time, only the “lidar/beta” type is supported. Defaults to “lidar/beta”.

method init_from_json

json_obj: 'dict',
source_artifact: 'Artifact'

Initialize a WBValue from a JSON blob based on the class that created it.

Looks through all subclasses and tries to match the json obj with the class which created it. It will then call that subclass’ from_json method. Importantly, this function will set the return object’s source_artifact attribute to the passed in source artifact. This is critical for artifact bookkeeping. If you choose to create a wandb.Value via it’s from_json method, make sure to properly set this artifact_source to avoid data duplication.

Arguments

• json_obj: A JSON dictionary to deserialize. It must contain a _type key. This is used to lookup the correct subclass to use.
• source_artifact: An artifact which will hold any additional resources which were stored during the to_json function.

method is_bound

self

method path_is_reference

path: 'str | pathlib.Path | None'

Arguments

• path:

method to_data_array

self

Convert the object to a list of primitives representing the underlying data.

method type_mapping

Return a map from _log_type to subclass. Used to lookup correct types for deserialization.

method with_suffix

name: 'str',
filetype: 'str' = 'json'

Get the name with the appropriate suffix.

Arguments

• name: the name of the file
• filetype: the filetype to use. Defaults to “json”.