Transparency
Methodology
How our data is built, what we guarantee and what we deliberately do not do.
The nature of the data
Every series on this site is synthetic. It is generated by a deterministic algorithm with a fixed seed and reproduces the shape of real observations: trends, seasonality, orders of magnitude and the character of the noise. These are not measurements and must not be cited as factual values.
This was a deliberate choice. The project's purpose is to demonstrate methods of working with and visualizing data, not to republish yet another copy of open observations. Synthetic data gives us full control over a series' properties and guarantees that a method demonstration depends on no one else's license.
Reproducibility
The same seed always yields the same series. Every chart on the site is therefore reproducible to the byte: rebuilding the project changes not a single point. The generation logic lives in one module rather than being scattered across pages.
Visualization principles
- Departure, not absolute. Where it is meaningful, we show the anomaly against a baseline.
- Honest axes. We don't crop an axis to exaggerate a trend, and we label units.
- Uncertainty is visible. Confidence bands are shown, not hidden.
- Accessibility. The palette is colour-blind safe, charts carry text descriptions, and animation is disabled under
prefers-reduced-motion.
Formats and license
Each dataset is available as machine-readable JSON on the Datasets page and released under the CC BY 4.0 license. You are free to use the data with attribution — provided you note its synthetic nature.
What we do not do
We do not present synthetic series as observations, we make no forecasts, and we give no policy or investment advice. If you need real measurements for research or a decision, consult primary scientific sources and use this site as an example of presentation.