FeatureGroup Class | Teradata Package for Python - FeatureGroup Class - Teradata Package for Python

The FeatureGroup class represents a collection of features that are related and belong to the same entity.

• FeatureGroup Properties
• FeatureGroup Methods

Syntax

FeatureGroup(name, features, entity, data_source, description=None)

Required Parameters

name

Specifies unique name of the feature group.

features

Specifies the features required to create a group.

entity

Specifies the entity associated with corresponding features.

data_source

Specifies the data source associated with features.

Optional Parameter

DataSource

Specifies human readable description for DataSource.

Example setup

>>> load_example_data('dataframe', ['sales'])
>>> df = DataFrame("sales")

Example: Create a FeatureGroup for above mentioned DataFrame

Create the features.

>>> jan_feature = Feature("sales:Jan", df.Jan)
>>> feb_feature = Feature("sales:Feb", df.Feb)
>>> mar_feature = Feature("sales:Mar", df.Mar)
>>> apr_feature = Feature("sales:Apr", df.Apr)

Create the entity.

>>> entity = Entity("sales:accounts", df.accounts)

Create the data source.

>>> data_source = DataSource("sales_source", df.show_query())

Create the feature group.

>>> fg = FeatureGroup('Sales',
...                   features=[jan_feature, feb_feature, mar_feature, apr_feature],
...                   entity=entity,
...                   data_source=data_source)

>>> fg

FeatureGroup(name=Sales, features=[Feature(name=sales:Jan), Feature(name=sales:Feb), Feature(name=sales:Mar), Feature(name=sales:Apr)], entity=Entity(name=sales:accounts), data_source=DataSource(Sales_Data))