Docs Data & Application Examples
The platform provides ready-to-use sample datasets and continuously updated application examples. Use sample data to walk through the forecasting workflow, or review application examples to see TimechoAI forecasts in weather, air quality, and other real-world time-series scenarios.
Data Examples
Select Data Examples in the left sidebar to open the sample data list. Each card shows the dataset name, summary, file size, and format, with actions to Download or Forecast.
Available Data
| Dataset | Sampling interval | Summary |
|---|---|---|
| Air Quality | Hourly | Urban air pollutant concentrations with environmental factors such as temperature and humidity |
| ETTh1 and ETTh2 | Hourly | Transformer operating data from an electric power system |
| ETTm1 and ETTm2 | Every 15 minutes | Higher-resolution transformer operating data from an electric power system |
| Exchange Rate | Daily | Exchange rates of multiple countries against the US dollar |
| Jena Climate | Every 10 minutes | Weather indicators including temperature, pressure, humidity, and wind direction |
| National Illness | Weekly | Influenza-like illness ratio data |
| Weather | Every 10 minutes | Multiple weather indicators including temperature and humidity |
The datasets and available items may change as the platform evolves. Refer to the Data Examples page for the current list.
Download Data
Click Download at the bottom of a dataset card to save the sample file locally. Samples use the TsFile format and can help you inspect the file structure, validate local processing, or practice uploading data to a session.
Forecast with a Sample
Click Forecast at the bottom of a dataset card. The platform opens the Create Session page and automatically loads the selected dataset. Once parsing is complete, review the columns and continue configuring the forecast.
- Check the detected time column, target variables, and covariates.
- Adjust column roles manually when needed.
- Open Forecast Settings and review the forecast steps and start position.
- Click the blue arrow on the right to submit the forecast.
- Review the forecast curve and result data in the current session.
Even when a sample loads successfully, review its column roles and forecast settings before submitting. Available target variables and covariates depend on the parsed dataset.
Application Examples
Select Application Examples in the left sidebar to view forecasting applications that run continuously on public data. Each card shows the current value, historical and forecast curves, data source, sampling interval, and last update time.
The current examples cover the following metrics:
| City | Metric | Source and update interval |
|---|---|---|
| Beijing | Temperature, humidity, and wind speed | Open-Meteo, every 15 minutes |
| Beijing | PM2.5 | Open-Meteo Air Quality, hourly |
| Guangzhou | Temperature and humidity | Open-Meteo, every 15 minutes |
| Shanghai | Temperature and humidity | Open-Meteo, every 15 minutes |
| Shanghai | PM2.5 | Open-Meteo Air Quality, hourly |
Solid lines represent observed historical data, while red dashed lines represent forecast data. Hover over a curve to inspect the observed or forecast value at a specific time.
Application examples use TimechoAI for forecasting and Apache IoTDB for time-series storage. Each card refreshes according to its data source, and the latest update time appears at the bottom of the card.
Application examples are for product evaluation, learning, and technical demonstration only. Their data and forecasts are not professional advice or a basis for decision-making.
Next Steps
- Create Session → — Learn the complete interactive forecasting workflow
- Datasets → — Manage and reuse your own time-series data files
- Model Gallery → — Review available models and their capabilities