Sample
Pulls N rows at random from a table, the standard move for audit testing and spot checks. Sampled rows keep their original order. Setting a seed makes the selection reproducible, which lets you document exactly how a test population was drawn.
Inputs
| Input | Accepts | Required |
|---|---|---|
Table | Table | Yes |
Outputs
| Output | Produces |
|---|---|
Result | Table |
Settings
Typed in on the node or set by the assistant, not wired.
| Setting | Default | Notes |
|---|---|---|
N | — | How many rows to pull, as a whole number (e.g. 25). |
Seed | — | Optional whole number that makes the sample repeatable: the same seed on the same table always picks the same rows. Leave blank for a different sample each run. |
Example
Your Expense_Reports table has 9,600 line items and the audit plan calls for
testing 25. Set N to 25 and Seed to 2026, and you get the same 25 rows every
time the pipeline runs. The selection is repeatable and defensible, and it
appears as its own tab, ready to export for the workpapers.
Tips
- Asking for more rows than the table has returns the whole table.
- To sample only large items, put a Filter (e.g.
Amount > 5000) before the Sample node.