Data structures

Data structures define the metadata schema for datasets — the set of variables, their types, and the valid values for each. They standardize how tabular data is documented and can link to codelists to constrain categorical variables to approved values.

Screenshot

Add screenshot: Data structures list view showing name, agency, version, and status

Accessing data structures

Go to Settings → Data structures in the administrator menu.

Creating a data structure

1. Click Add data structure.
2. Fill in:
   • Agency — the responsible organization
   • Name — a unique identifier
   • Version — semantic version number (e.g., 1.0.0)
3. Click Save.

Screenshot

Add screenshot: Create data structure form

Managing variables

Each data structure contains a set of variable definitions. To manage variables:

1. Open a data structure from the list.
2. Click Variables.
3. Add or edit variables. Each variable has:
   • Name — the variable identifier
   • Type — the data type (text, numeric, date, etc.)
   • Label — a human-readable description
   • Value domain — optionally link a published codelist for categorical variables

Screenshot

Add screenshot: Variable editor within a data structure

Linking codelists

When a codelist is set as the value domain for a variable, only codes from that codelist are considered valid for the variable. Ensure the codelist is in Published status before linking it.

Status workflow

Data structures follow the same status workflow as codelists: Draft → Review → Published → Deprecated → Archived.