Ocholar/East_Africa_Maize_Yield_Insights

📊 East Africa Maize Yield Insights Visualizes maize yield trends across Kenya, Rwanda, and Tanzania using SQL, Python, and Power BI. Parses 18,000+ field folders with metadata and satellite imagery Ranks variety and seasonal performance Builds dashboards for regional yield decision-making # 🌾 East Africa Maize Yield Insights This project visualizes maize yield performance across East African trials using field-calibrated agronomic methods, Python, SQL, and Power BI. It combines raw plot measurements with seasonality, regional comparisons, and variety-level analysis — delivering actionable dashboards for decision-makers in agricultural development. --- ## 📦 Repository Structure ```text modules/ └── yield-performance-module/ ├── data/ # Cleaned dataset for analysis │ └── yield_metadata_cleaned.csv ├── exports/ # Aggregated SQL output files │ └── yearly_avg_yield_by_country.csv ├── dashboards/ │ ├── final_dashboard.pbix # Power BI dashboard │ └── visual_snapshots/ │ └── yield_dashboard_snapshot.pdf ├── SQL/ # Raw and aggregated query logic │ └── top_yielding_country_per_year.sql └── docs/ └── methodology.md # Cleaning, estimation, and logic docs --- ## 📦 Dataset Overview - **Source**: Lacuna Fund Agriculture Datasets - **Collected by**: One Acre Fund - **Regions Covered**: Kenya, Rwanda, Tanzania - **Files Processed**: 18,482 folders, each containing: - `metadata.json`: Agronomic details (yield weights, plot size, fertilizer use, GPS) - `stac.json`: Satellite image metadata (timestamps, bounding box, sensor info) --- ## 🧾 Fields Extracted - `Year` - `Season` - `Country` - `Longitude` - `Latitude` - `BoxAWidth` - `BoxALength` - `BoxBWidth` - `BoxBLength` - `BoxAWetWeight` - `BoxADryWeight` - `BoxBWetWeight` - `BoxBDryWeight` - `PlotSize_Acres` - `Variety` - `Planting Date` - `CAN_Kgs` - `DAP_Kgs` - `NPK_Kgs` - `Urea_Kgs` - `ImgID` --- ## 📊 Project Goals - Build a unified SQLite database with field-level agronomic data - Identify yield performance trends across regions and seasons - Pair dry weight measurements with imagery metadata to explore spatial correlations - Showcase technical data wrangling using non-code tools (Excel + Power Query) --- # …