Create a federated table-Ant Privacy-Preserving Computation Platform-阿里云帮助中心

A federated table is a collection of files distributed among different participants. These files are used for a single purpose, such as a training dataset. This topic provides an API example for creating a federated table.

Important

All placeholders in the API, such as "$df0", must be enclosed in single or double quotation marks.

Function path

fascia.biz.api.dataframe.create_fed_dataframe

Function definition

def create_fed_dataframe(uid='${UID}',
                         data_partitions=[${DATA_PARTITIONS}],
                         filter_columns=[${FILTER_COLUMNS}])

Request parameters

Name	Type	Required	Description
uid	String	Required	The unique ID of the federated table to create.
data_partitions	List	Required	A list of data source placeholders for the federated table. These placeholders are defined in the input configuration.
filter_columns	List	Optional	The columns to include in the output. If this parameter is set to None, all common columns from each party's data are included in the output by default. The default value is None.

Request example

from fascia.biz.api.dataframe import create_fed_dataframe

# Create a federated table
fed_df = create_fed_dataframe(uid='$output0', data_partitions=['$df0', '$df1'])

$output0 is an output parameter that is preconfigured and treated as a constant in the code.
$df0 and $df1 are input parameters that are preconfigured and treated as constants in the code.