This project analyzes and compares solar irradiation data across Benin, Sierra Leone, and Togo to identify relative solar potential and key differences between these West African nations.

solar-analysis/
├── data/
│   ├── raw/                  # Original data files
│   ├── processed/            # Cleaned country datasets
│   └── outputs/              # Analysis results
├── notebooks/
│   ├── 1_data_cleaning.ipynb
│   ├── 2_country_analysis.ipynb
│   └── 3_comparison.ipynb    # Main comparison notebook
├── src/
│   ├── data_cleaner.py       # Data preprocessing
│   ├── feature_engineer.py   # Feature generation
│   └── visualization.py      # Plotting functions
└── README.md

1. Data Preparation

   • Load raw CSV files for each country
   • Standardize timestamps and column names
   • Handle missing values and outliers

2. Country-Specific Cleaning

   # Sample cleaning pipeline per country
   cleaner = (
       SolarDataCleaner(df)
       .add_time_features()
       .handle_missing_values()
       .remove_outliers()
       .validate_ranges()
   )
   cleaned_df = cleaner.get_clean_data()

3. Comparative Analysis

   • Generate side-by-side visualizations
   • Calculate summary statistics
   • Perform statistical significance testing

4. Key Metric Comparison

   • Global Horizontal Irradiance (GHI)
   • Direct Normal Irradiance (DNI)
   • Diffuse Horizontal Irradiance (DHI)

• Load each country's cleaned dataset
• Ensure consistent datetime indexing
• Verify all metrics exist in each dataset

# Generate comparison boxplots
plot_comparison_boxplots(
    metrics=['GHI', 'DNI', 'DHI'],
    countries=['benin', 'sierra_leone', 'togo'],
    data=cleaned_data
)

• ANOVA Testing:

  stats.f_oneway(benin_ghi, sierraleone_ghi, togo_ghi)

• Post-hoc Tests (if ANOVA significant):

  stats.tukey_hsd(benin_ghi, sierraleone_ghi, togo_ghi)

• Create markdown summary of:
  • Highest potential countries
  • Variability observations
  • Atmospheric condition inferences

1. Install requirements:

   pip install -r requirements.txt

2. Execute notebooks in order:

   1_data_cleaning.ipynb → 2_country_analysis.ipynb → 3_comparison.ipynb
   

3. View outputs in:

   • data/outputs/ for numerical results
   • notebooks/figures/ for visualizations

### Comparative Insights
1. Highest Potential: Sierra Leone shows the highest median GHI at approximately 490 W/m², closely followed by Benin.

2. Consistency: Togo demonstrates the most stable radiation (IQR ≈ 400 W/m²), indicating less variability in total irradiance.

3. Atmospheric: Togo's low DNI compared to GHI (high DHI/DNI ratio) suggests more diffuse sunlight, likely due to frequent cloud cover or atmospheric scattering.

• Seasonal variation decomposition
• Cloud cover impact analysis
• Solar farm yield simulations
• Economic feasibility projections

This systematic approach enables scientifically rigorous comparison of solar potential across West African nations, supporting data-driven energy policy decisions.