feast.infra.offline_stores.contrib.trino_offline_store.tests package
Submodules
feast.infra.offline_stores.contrib.trino_offline_store.tests.data_source module
- class feast.infra.offline_stores.contrib.trino_offline_store.tests.data_source.TrinoSourceCreator(project_name: str, fixture_request: _pytest.fixtures.FixtureRequest, **kwargs)[source]
Bases:
tests.integration.feature_repos.universal.data_source_creator.DataSourceCreator
- create_data_source(df: pandas.core.frame.DataFrame, destination_name: str, suffix: Optional[str] = None, timestamp_field='ts', created_timestamp_column='created_ts', field_mapping: Optional[Dict[str, str]] = None) feast.data_source.DataSource [source]
Create a data source based on the dataframe. Implementing this method requires the underlying implementation to persist the dataframe in offline store, using the destination string as a way to differentiate multiple dataframes and data sources.
- Parameters
df – The dataframe to be used to create the data source.
destination_name – This str is used by the implementing classes to isolate the multiple dataframes from each other.
event_timestamp_column – (Deprecated) Pass through for the underlying data source.
created_timestamp_column – Pass through for the underlying data source.
field_mapping – Pass through for the underlying data source.
timestamp_field – Pass through for the underlying data source.
- Returns
A Data source object, pointing to a table or file that is uploaded/persisted for the purpose of the test.
- create_offline_store_config() feast.repo_config.FeastConfigBaseModel [source]
- create_saved_dataset_destination() feast.infra.offline_stores.contrib.trino_offline_store.trino_source.SavedDatasetTrinoStorage [source]