What are the different types of dimensions that can be used in a data warehouse?

In dimensional modeling, dimensions describe the context of facts and are typically categorized into several types:

1. Conformed Dimensions – shared across multiple fact tables or data marts, ensuring consistent semantics (e.g., a common Date dimension used by sales and inventory facts).
2. Degenerate Dimensions – derived from a fact table’s primary key or transaction number, often stored as a simple attribute rather than a separate dimension table (e.g., an Order Number in a Sales Fact).
3. Junk Dimensions – consolidate many low‑cardinality flags or attributes into a single dimension to reduce clutter (e.g., a Junk dimension containing Yes/No flags for various product attributes).
4. Role‑Playing Dimensions – the same dimension used in different contexts, such as a Date dimension used for Order Date, Ship Date, and Delivery Date.
5. Slowly Changing Dimensions (SCD) – handle changes over time, with types like SCD‑1 (overwrite), SCD‑2 (historical record), or SCD‑3 (add new attribute).

Choosing the right type depends on business requirements, query patterns, and data governance needs.