Synthetic Dataset Laboratory
Design seeded relational fixtures with bounded columns, uniqueness, missing values and enforced parent keys, then export SQLite and CSV datasets.
Application guideVerifying your session…
Using Synthetic Dataset Laboratory
Generate repeatable relational datasets from authored schema constraints and a seed.
Constraint-driven generation · Relational integrity · Repeatable fixtures
- Inputs and working data
- Table row counts, column generators, finite domains, null quotas, uniqueness, foreign keys and generator seed.
- Working in the editor
- Table relationship explorer, typed generator editor and paged generated tables.
- What can I produce?
- Related CSVs, executable SQLite fixture transaction, input schema and SHA-256 table checksums.
- What are the limits?
- 25 tables, 32 explicit columns per table, 20,000 total rows, acyclic single-column foreign keys to generated integer primary keys.
- Example
- Generate related objects twice, compare hashes, load the SQL into SQLite and verify actual foreign keys, null quotas and exported values.