README.md

metadata

title: SimpleTS
emoji: 📈
colorFrom: green
colorTo: blue
sdk: docker
app_port: 8501
tags:
  - streamlit
  - time-series
  - forecasting
  - prophet
  - arima
  - holt-winters
pinned: false
short_description: Analyze and forecast time series — no code required.

SimpleTS

Interactive Streamlit dashboard for time series analysis and forecasting. Upload your CSV or use built-in demos.

Author

Eduardo Nacimiento García
📧 enacimie@ull.edu.es
📜 Apache 2.0 License

Features

• Upload CSV or use daily/monthly/weekly demo datasets
• Automatic date parsing and time series setup
• Stationarity test (ADF)
• Seasonal decomposition (trend, seasonality, residuals)
• ACF & PACF plots
• Forecasting models:
  • Holt-Winters Exponential Smoothing
  • ARIMA (configurable p,d,q)
  • Prophet (by Meta)
• Metrics: MAE, MSE, RMSE
• Interactive future forecasting
• Plotly visualizations

Demo Datasets

Three built-in demos:

• Daily (1 year, 365 points)
• Weekly (2 years, 104 points)
• Monthly (4 years, 48 points)

Deployment

Ready for Hugging Face Spaces (free tier).

⚠️ Uses sdk: docker — include Dockerfile.

Requirements

• Python 3.8+
• Streamlit, pandas, numpy, plotly, statsmodels, prophet, scikit-learn

---

💡 Tip: After uploading, select date/value columns → analyze stationarity & seasonality → choose model → forecast future values!