tft.DatasetMetadata

Metadata about a dataset used for the "instance dict" format.

tft.DatasetMetadata(
    schema: schema_pb2.Schema
)

Used in the notebooks

Used in the tutorials
Preprocessing data with TensorFlow Transform

Caution: The "instance dict" format used with DatasetMetadata is much less efficient than TFXIO. For any serious workloads you should use TFXIO with a tfxio.TensorAdapterConfig instance as the metadata. Refer to Get started with TF-Transform for more details.

This is an in-memory representation that may be serialized and deserialized to and from a variety of disk representations.

Attributes
schema

Methods

from_feature_spec

View source

@classmethod
from_feature_spec(
    feature_spec: Mapping[str, common_types.FeatureSpecType],
    domains: Optional[Mapping[str, common_types.DomainType]] = None
) -> _DatasetMetadataType

Creates a DatasetMetadata from a TF feature spec dict.

__eq__

View source

__eq__(
    other
)

Return self==value.

__ne__

View source

__ne__(
    other
)

Return self!=value.