Skip to main content
Version: 0.8

tecton.SparkBatchConfig

Summary​

Configuration used to define a batch source using a Data Source Function.

The SparkBatchConfig class is used to configure a batch source using a user defined Data Source Function.

This class is used as an input to a BatchSource’s parameter batch_config. Declaring this configuration class alone will not register a Data Source. Instead, declare as a part of BatchSource that takes this configuration class instance as a parameter.

warning

Do not instantiate this class directly. Use tecton.spark_batch_config() instead.

Attributes​

  • data_delay: This attribute is the same as the data_delay parameter of the __init__ method. See below.

Methods​

__init__(...)​

Instantiates a new SparkBatchConfig.

Parameters​

  • data_source_function (Union[Callable[[SparkSession], DataFrame], Callable[[SparkSession, FilterContext], DataFrame]]) – User defined Data Source Function that takes in a SparkSession and an optional tecton.FilterContext, if supports_time_filtering=True. Returns a DataFrame.

  • data_delay (timedelta) – By default, incremental materialization jobs run immediately at the end of the batch schedule period. This parameter configures how long they wait after the end of the period before starting, typically to ensure that all data has landed. For example, if a feature view has a batch_schedule of 1 day and one of the data source inputs has a data_delay of 1 hour, then incremental materialization jobs will run at 01:00 UTC. (Default: datetime.timedelta(0))

  • supports_time_filtering (bool) – Must be set to to True if one of the following conditions is met:

    • <data source>.get_dataframe() is called with start_time or end_time
    • A feature view wraps this Data Source with a FilteredSource

    If this parameter is set to true, Tecton passes a FilterContext object into the Data Source Function, which is expect to handle its own filtering. (Default: False)

Returns​

A SparkBatchConfig class instance.

Was this page helpful?