Do a Quick Test

TL;DR

A Quick Test runs an aggregated Data Element on one record and shows the result right away, no full simulation needed. Enter a record id on the Quick Test tab to check your aggregation while you build it.

What is it?

Quick Test runs on an aggregated Data Element, a Data Element with Is Aggregated turned on. You point it at one record, tweak the logic, and re-run until it’s right, all on the same tab.

It’s the fast counterpart to a full Simulation, use it while developing, then run a Simulation to validate.

	Quick Test	Simulation
Scope	One record, interactive	A full sample, as a job
Speed	Instant	Slower; runs as a queued job
Use For	Sanity-checking an aggregated Data Element while iterating	Validating performance, generating approval evidence

• It works only on aggregated Data Elements.
• It runs Python and pandas definitions. Test Spark aggregates with a full Simulation instead.

Other object types (Features, Models, Policies) do not have a Quick Test tab. To trace a Policy record end-to-end, use Run a Single Record instead.

How to Run a Quick Test

Open the Quick Test tab

Open the aggregated Data Element’s details page and switch to the Quick Test tab.

If the Data Element is not aggregated, the tab shows Enable Data Element Aggregation to use Quick Test. Turn on Is Aggregated in the Details tab first, then return.

Enter an ID

In Enter ID to start reviewing, type the ID of the record you want to test. For an application-level Data Element that is an application ID; for an account-level one, an account ID. The field checks that the value matches the ID column’s data type.

Click Submit & Review. The platform pulls just that record’s rows from the source tables and runs the aggregation on them.

Read the result

The result appears in two parts:

• Output - the aggregated value your Data Element returned for that record, shown at the top. Any print statements from your definition appear just below it, so you can trace what the logic did.
• Filtered Data - the source rows the aggregation read for that record, shown alongside the output, one table at a time. You can sort, filter, and page through each table, or open it in a new tab, to check the aggregation read what you expected.

If your definition errors out, an Error Logs section takes the place of the output and shows the traceback.

Adjust the definition and re-run until the output looks right. For example, if a maximum-balance aggregate comes back lower than you expected, open Filtered Data to check whether every account row for that application was actually read. Once the logic checks out, move on to a full Simulation.

What’s Next

• Register a Data Aggregate to set up the aggregation Quick Test runs on.
• Run a Simulation once the logic checks out, to validate it across a sample.
• For a Policy, trace a record end-to-end with Run a Single Record.