Data Patterns
A row that changes over time needs a rule for what the history looks like. The pattern for that is slowly changing dimensions: what to do when a row changes, from keeping nothing to keeping every version.
The rule belongs to the warehouse design, not to one pipeline. The same notes are in Data Warehouse.
The reference is Adatis. Slowly changing dimensions (SCD) by Adatis: what you can do when the information in your table changes, types 0 through 6. The figure below lays those types out side by side, with the credit to Adatis.
Last updated
Was this helpful?