This is an old revision of the document!
Table of Contents
SANDBOX
Category: Create / Input
Description
This action creates a temporary table that is typically used for one-time calculations. Sandboxes can be populated with result data of any action, or pasted from the clipboard.
Data in sandboxes survives reloads - contents don't change on partial or full reloads - although it doesn't survive re-opening the project. When a project is opened all of its sandboxes start empty.
Populating a Sandbox with data
| Option | Description |
|---|---|
| Paste data | This option will paste data that has been copied to the clipboard. Options for headers, delimiters, and decimal provide fine-tuning of the pasted data. |
| Import dataset | This option allows for browsing to and selecting an external dataset (.dset file) to populate the Sandbox with. |
Creating and Populating a Sandbox from another action
By right-clicking an action that generates output data and selecting Send to sandbox/module > New sandbox, you create a new sandbox populated with the data from the source action. Selecting New sandbox in new group creates a new group and inserts a new sandbox action populated with the data from the source action.
Use cases
Besides ad hoc calculations, sandboxes can be used in a few more cases:
- Together with Either table transformation Sandbox can be used to safely insert an arbitrary dataset into any point of a transformation chain, which can be useful for debugging and development purposes. In this scenario insert Either table and point its second input to a table created using Sandbox transformation.
- Sandboxes can be used to create an empty table as a starting point for further transformations. Some transformations (e.g. Run external program might not need an input dataset, but they don't create a table. In this case a generic empty table can be created using Sandbox, and then Run external program can be inserted after it.