rialto.metadata package

Subpackages

• rialto.metadata.data_classes package
  • Submodules
  • rialto.metadata.data_classes.feature_metadata module
    • FeatureMetadata
      • FeatureMetadata.add_group()
      • FeatureMetadata.description
      • FeatureMetadata.from_spark()
      • FeatureMetadata.group
      • FeatureMetadata.name
      • FeatureMetadata.to_tuple()
      • FeatureMetadata.value_type
  • rialto.metadata.data_classes.group_metadata module
    • GroupMetadata
      • GroupMetadata.add_features()
      • GroupMetadata.description
      • GroupMetadata.features
      • GroupMetadata.frequency
      • GroupMetadata.from_spark()
      • GroupMetadata.fs_name
      • GroupMetadata.key
      • GroupMetadata.name
      • GroupMetadata.to_tuple()
  • Module contents

Submodules

rialto.metadata.enums module

class rialto.metadata.enums.Schedule(value)

Bases: Enum

Schedule options

daily = 'daily'

monthly = 'monthly'

unscheduled = 'unscheduled'

weekly = 'weekly'

yearly = 'yearly'

class rialto.metadata.enums.ValueType(value)

Bases: Enum

Value options

nominal = 'nominal'

numerical = 'numerical'

ordinal = 'ordinal'

rialto.metadata.metadata_manager module

class rialto.metadata.metadata_manager.MetadataManager(session, schema_path: str | None = None)

Bases: object

Metadata storage i/o

get_feature(group_name: str, feature_name: str) → FeatureMetadata

Get metadata of one feature

Parameters:

• group_name – string name of feature group

• feature_name – string name of feature

Returns:

FeatureMetadata object

get_group(group_name: str) → GroupMetadata

Get metadata of one feature group

Parameters:

group_name – string name of feature group

Returns:

GroupMetadata object

update(group_md: GroupMetadata, features_md: List[FeatureMetadata])

Save or refresh information about generated features and their groups

Parameters:

• group_md – GroupMetadata object

• features_md – list of FeatureMetadata objects

Returns:

rialto.metadata.utils module

rialto.metadata.utils.class_to_catalog_name(class_name) → str

Map python class name of feature group (CammelCase) to databricks compatible format (lowercase with underscores)

Parameters:

class_name – Python class name

Returns:

feature storage name

Module contents