Skip to content

Transform Plans

A transform plan explains how source fields become governed output fields. It is the review step between raw data and reusable analytics.

Each target field has a display name, type, mapping rule, optional unit, and optional constraints. Rokks can suggest a plan from sample data, but the user remains responsible for confirming the business meaning. Once loaded, downstream pages read the resulting catalog metadata rather than the raw plan.

  1. Open a file area or source that has sampled data.
  2. Review every proposed target field.
  3. Rename fields to business-friendly labels.
  4. Confirm date, number, boolean, enum, and text types.
  5. Set units where measures need interpretation.
  6. Mark key fields when they identify records.
  7. Save or run the load.

Do not accept vague names such as Column 1 or value when the field has a business meaning. Good names improve dashboards and assistant suggestions.

Do not use a text type for values that need sorting, aggregation, time filtering, or thresholds.