You are viewing documentation for an older version.
Click here to view the latest documentation.

inputs.yaml

Learn how to define and map input sources for your Profiles project.

less than a minute

The inputs.yaml file contains the configuration of all input sources used in Profiles. This includes the ID graph, entity vars, feature view, etc.

yaml

inputs:
    - name: input_name
        app_defaults:
            table: schema.table # key must be only one of table/view/s3/csv
            occurred_at_col: time_series_column
            row_identifier: [list of identifiers for composite key]    # optional
            ids:
                <id-config>

Key	Description
`name` Required	The input name.
`app_defaults` Required	Input configuration.
`table`\ `view`\ `s3`\ `csv` Required	Source name, see input sources.
`occurred_at_col` Required	Time series column.
`row_identifier`	List of all the identifiers that combine to create a composite primary key.
`ids` Required	All IDs in the input source, mapped to `entities` and `id_types`

Input sources

A source can be a warehouse table, view, Amazon S3 bucket, or CSV file. Only one key can be used per source.

Key	Description
`table`	Warehouse table name You can also refer to tables from other databases/schemas in the same data warehouse.
`view`	Warehouse view name.
`csv`	CSV file path. Note that this path should be relative to the project folder. Note: RudderStack recommends using CSV file as an input only if you have limited amount of data.
`s3`	Amazon S3 URI path.