Energy Explorer – Interactive Time-Series Analysis Dashboard

Energy Explorer is an interactive data analysis dashboard for exploring global energy consumption, electricity generation, and emissions data from Our World in Data.

The application focuses on time-series analysis and helps users understand trends, temporal dependencies, and the statistical structure of energy-related indicators across countries.

---

🎯 Project Goals

This project was built to:

• Explore long-term developments in global energy and emissions data
• Compare time series across countries and metrics
• Analyze temporal dependencies using ACF (Autocorrelation Function) and
  PACF (Partial Autocorrelation Function)
• Assess time-series suitability for modeling approaches (e.g. ARIMA)
• Demonstrate a clean, reproducible, portfolio-ready data application

---

🧠 Key Features

• Interactive selection

  • Country (or global aggregation)
  • Time range
  • Energy / emissions metric

• Four analytical panels

  1. Time series visualization
  2. Autocorrelation (ACF)
  3. Rolling standard deviation vs. rolling mean (stationarity diagnostics)
  4. Partial autocorrelation (PACF)

• Automatic metadata handling

  • Units and descriptions loaded from the OWID codebook
  • Dynamic axis labeling based on the selected metric

---

🛠️ Technical Stack

Component	Technology
Data source	Our World in Data (CSV)
Storage & queries	DuckDB
Backend logic	Python
Visualization	Streamlit + Plotly
Statistics	NumPy, statsmodels
Deployment	Streamlit Community Cloud

---

🗂️ Project Structure

├── energy_app.py # Main Streamlit application
├── data/ # Local data (ignored in Git)
├── requirements.txt # Python dependencies
├── README.md # Project documentation
└── .gitignore

---

🚀 How to Run Locally

1. Clone the repository

git clone https://github.com//energy-explorer.git cd energy-explorer

2. Create and activate a virtual environment

python -m venv .venv source .venv/bin/activate # macOS / Linux

3. Install dependencies

pip install -r requirements.txt

4. Start the application

streamlit run energy_app.py

📊 Data Source

Our World in Data – Energy Dataset https://github.com/owid/energy-data The data is downloaded automatically on first run and stored locally. Metadata (units and descriptions) are read from the official OWID codebook.

📄 License

This project is released under the MIT License.

👤 Author

Axel Behrendt
Data analysis · Time-series · Scientific Python
Parts of the development as well as the debugging were supported by ChatGPT 5.2.