A variable describes the data.


Add variable to View

This operation adds or updates a derived variable in a view for each selected variable.

Categorize this variable to another

This operation adds or overwrites a variable in a view and allows to recode its values.

Categorize another variable to this

This operation maps another variable’s values to the current variable categories.

Custom derivation

Derive a variable by editing its derivation script manually.


Removes the variable from the table.


This section displays the proerties of the variable:

  • Name
  • Entity Type
  • Value Type
  • Repeatable
  • Unit
  • Referenced Entity Type
  • Mime Type
  • Occurence Group
  • Index


Some variables can have categories defined. The list of categories is displayed with a summary information:

  • label (mapped label category attribute if this one is defined)
  • missing (if the category indeicates a missing answer)

Edit Categories

This operation allows the addition, edition abd deletion of a variable’s categories. Categories can also be removed or reordered by selecting one or multiple categories.


Some variables can have attributes defined. The list of attributes is displayed with full information:

  • namespace
  • language
  • value

Add Attributes

Thois operation addds a new attribute. The combination of namespace and name must be unique.


To assign the attribute to another namespace, change its name or set its value. When editing multiple attributes only the namespace can be modified.


Remove the attributes.


Statistical summary of the variable:

  • variables with categories:
    • frequency plot
  • variables without categories:
    • histogram
    • normal probability plot
    • summary data: N, Min, Max, Mean, Median, etc
    • frequencies of missiong and non-missing values


Derived variables (i.e when the table is a view) are persisted in Opal’s embedded Version Control System which tracks all changes to a script over time. One practical use case is revising the history of changes and if necessary revert the script to a previous revision.

Script History Revisions

Each time a script is edited a new history revision is created or ‘committed’ to Opal’s version control system.

Commit Differences

Commit revisions are organized in a descending order, i.e., the latest commit at the top of the history stack. A simple ‘diff’ compares the changes between two immediate commits. Opal also offers a comparison between any revisions to the current revision.

Reverting Changes

By editing and saving an older revision, a script content is reverted to its previous version. This operation is tracked as a new revision.

Review Commit Differences

The commit differences are ordered by the oldest changes first (denoted in red) followed by the latest changes (denoted in green).


Values can be displayed for a specific identifier or can be filtered to match to certain criteria.


Specify the access rights to a particular variable and its content

View with summary Permission

Allow the user to see the variable details with its data summary. Does not allow values querying. It induces the read-only access to the parent table and datasource.