🏷️ Tags: R 📊, JavaScript 🌐, Shiny ✨
This project analyses and forecasts Taiwan’s renewable energy development, with particular emphasis on the 2016 energy policy reforms.
It combines interactive visualisations and time-series forecasting to help policymakers and stakeholders quickly understand trends, assess progress toward targets, and compare forecasting models.
“Explore Taiwan’s renewable energy journey from 2005 to 2025, and see how close we are to the 20% renewable energy goal.”
- Observable – Interactive visualisations (JavaScript) 🌐
- R – Time-series forecasting, analysis, and plotting (ggplot2) 📊
- Shiny – Interactive dashboard for exploration ✨
- Renewable energy generation has steadily increased, with a notable acceleration after 2016.
- Solar energy drives most of the growth, with clear seasonal patterns (higher in summer, lower in winter).
- Wind energy shows gradual improvement, with winter peaks.
- Forecasting models suggest that reaching 20% by 2025 is unlikely, but 2027 may be more realistic.
For detailed analysis and results:
| Section | Description | Tool | Link |
|---|---|---|---|
| 🔹 R results overview | Summary tables and plots generated in R | 🇷 R |
📎 Link |
| 📊 Data Exploration (EDA) | Annual/seasonal patterns, key energy source trends | 🇷 R |
📎 Link |
| 📈 Model Comparison | ARIMA, ETS, Prophet and evaluation | 🇷 R |
📎 Link |
| 🔮 Forecasting Results | 2025 & 2030 projections vs policy targets | 🇷 R |
📎 Link |
| 🌍 Shiny Dashboard | Interactive exploration | ✨ Shiny |
📎 Link |
| 🔗 Observable Report | Interactive visualisation in JavaScript | 🌐 JS |
📎 Link |
- Built with Shiny (R)
- Select year range and energy sources to explore trends and seasonal patterns
🌍 Try the Dashboard: Shiny App Link
Detailed instructions: Learn how to interact with filters, view seasonal patterns, and interpret plots
📖 Dashboard Guide: Shiny Dashboard
- Source: Taiwan’s Open Data Platform
- Coverage: 2005–2024 (monthly)
- Features: Power generation by energy source
- Processing: Cleaned and prepared with Python; forecasting done in R
Part of the Master’s in Data Science program (Data Visualisation course) at City, University of London (2024), where it received a Distinction.
Extended using R to provide advanced forecasting and additional analysis.
This project is licensed under the MIT License.