View a markdown version of this page

Using an intermediate table in analyses - AWS Clean Rooms

Using an intermediate table in analyses

After you populate an intermediate table and attach an analysis rule, you can reference it in queries just like a configured table.

Discovering intermediate tables

All collaboration members can view intermediate tables on the Tables tab of a collaboration. The table type is shown as "Intermediate table."

Referencing in SQL

Reference an intermediate table by name in your SQL query, just like you reference a configured table. For example:

SELECT * FROM my_intermediate_table

Referencing in PySpark

In a PySpark analysis template, reference an intermediate table through the context object:

context['referencedTables']['my_intermediate_table']

Requirements at analysis time

AWS Clean Rooms enforces the following requirements when you run an analysis that references an intermediate table:

  • The intermediate table schema status must be Ready (populated and analysis rule attached).

  • The analysis rule is validated against inherited constraints from base tables.

  • Deferred controls are enforced, including disallowedOutputColumns and allowedResultReceivers.

Using in ML workflows

You can reference intermediate tables in ML workflows through the dataSource.protectedQueryInputParameters configuration.

Viewing the schema

To view the columns available in an intermediate table, choose the Schema tab on the intermediate table details page. You can view the column name and type. Choose Edit schema to update column types.

Viewing dependencies

To view the base tables that an intermediate table depends on, choose the Dependencies tab. You can view the table name, owner, type, and parent type for each dependency.