Counts and scope
Cite the dated synthetic snapshot for counts and aggregates.
Citation and provenance
Make the exact data identifiable.
Cite the exact dataset, city, domains and release used. Keep the source and synthetic basis clear.
Keep these details with the analysis.
- Fidelra Africa / Striped Donkey and the dataset or sample title.
- City, domains and package or record scope.
- Release identifier and publication date from the manifest or package documentation.
- Source URL and access date.
- Relevant dictionary, verification, assumptions and limitations.
Use the dataset release identifier, which differs from the website version.
Attribution for public materials.
New Striped Donkey public dataset previews use CC BY-NC 4.0: share and adapt for non-commercial purposes with attribution. Earlier reference releases retain their original licences.
When sharing the material, give appropriate credit, link to the licence, indicate any changes and retain the notices supplied with it. Do not imply endorsement by Striped Donkey or Fidelra Africa.
Use any reasonable attribution format. Include the dataset title, publisher, source URL and licence; a release identifier and synthetic-data context help readers identify the material.
Use the licence notice for the exact artifact. Original ML-01 and health reference releases supplied under CC BY 4.0 retain those earlier grants.
Suggested attribution: Striped Donkey / Fidelra Africa. [Dataset title and release]. [Source URL]. CC BY-NC 4.0. [Describe changes, if any].
A suggested citation format.
Replace brackets with verified package details. This optional format adds release context to the attribution required by the applicable licence.
Fidelra Africa. Striped Donkey: [dataset or sample title]. [City and domain scope]. [Dataset version or release identifier], [published release date]. Synthetic data. [Source URL]. [Applicable licence and licence URL]. [Changes, if any]. Accessed [date].
For audits, cite the title, city, domain, result date and tested release. Describe internal checks accurately; they are neither independent validation nor a new run.
Use the source that matches the claim.
Release checks
Retain the original date, outcome and scope.
Reference evidence
Find calibration references and reports in each city’s index.
Rights and notices
Preserve release, provenance, ownership and applicable third-party notices.
State the synthetic basis.
Identify findings as synthetic. Generated records cannot substantiate claims about real individuals. Distinguish modeled scenarios from observations and forecasts.