randarium
Datasets

Warehouse Schema Generator

Create realistic data warehouse rows in common schema patterns: dimension tables with attributes, fact tables with measures and FKs, date dimensions, or Slowly Changing Dimension Type 2 with temporal validity. This data is synthetic and not real.

Also known as: data warehouse · star schema · dw schema

seeded · synthetic data

Output

No output yet — set your options and hit .
About this tool, tips & examples

What it does

The Warehouse Schema Generator produces rows in the classic data warehouse patterns: dimension tables with descriptive attributes, fact tables with measures and foreign keys, date dimensions with calendar attributes, and SCD Type 2 rows with valid_from/valid_to versioning. Up to 10,000 rows per run, seeded — star-schema test data without hand-writing it.

Common use cases

  • ETL pipeline testing — loads, upserts, and surrogate-key handling against realistic dimensional data.
  • SCD2 logic validation — temporal-validity rows exercise the trickiest merge pattern in warehousing: closing old versions, opening new ones, keeping history queryable.
  • BI development — populate a star schema so dashboards and semantic layers have something to aggregate.
  • Teaching dimensional modeling — concrete dimension/fact/date examples beat diagrams.

Settings

  • Schema kind — dimension, fact, date dimension, or SCD2.
  • Rows — 1 to 10,000, exportable as CSV, JSON, NDJSON, TSV, or Markdown.
  • Seed — identical seed + settings = identical rows, so warehouse tests stay deterministic.

Privacy note

Rows are generated locally in your browser and never uploaded. Everything is synthetic — no real business data — which is exactly what should be in a development warehouse.

FAQ

What makes SCD2 rows special? Each entity appears in multiple versions with validity windows — the current row open-ended, history closed off. Testing merge logic against generated SCD2 data catches the classic bugs: overlapping windows and orphaned current flags.

Do fact rows reference the dimensions? Fact rows carry foreign-key-shaped columns, so join-path testing works. For a fully keyed multi-table model, generate dimensions and facts with coordinated seeds and align keys in your fixture code.

A full date dimension? The date-dimension kind emits calendar attributes per row — the join-target every warehouse needs and nobody enjoys writing by hand.