Momahmoses/ng-healthcare-access-analyzer

||Catchment-area analysis for hospitals and PHCs across Nigeria identifying underserved communities using PySpark, Azure and Streamlit. # Nigeria Healthcare Access Gap Analyzer Geospatial platform for identifying underserved communities across Nigeria's 36 states, measuring facility density, travel-time catchments, and healthcare coverage gaps using **GIS**, **PySpark**, **Azure**, and **Streamlit**. ## Problem Statement Over 50% of Nigerians live more than 5km from a health facility. Rural states in the North-East and North-West have fewer than 0.2 doctors per 1,000 people, far below WHO's 1:1,000 benchmark. This tool helps health planners prioritize facility construction and resource allocation. ## Tech Stack | Layer | Technology | |---|---| | Geospatial | GeoPandas, Shapely, Folium | | Big Data | PySpark on Azure Databricks | | Cloud | Azure Blob Storage, Azure Health Data Services | | Dashboard | Streamlit + Plotly | ## Project Structure ``` ng-healthcare-access-analyzer/ ├── app.py # Streamlit dashboard ├── pipeline/spark_pipeline.py # PySpark facility density & gap scoring ├── gis/spatial_analysis.py # Catchment areas & nearest-facility distance ├── data/generate_data.py # Synthetic facility and population data ├── azure/azure_config.py # Azure Blob & Databricks helpers └── requirements.txt ``` ## Quick Start ```bash pip install -r requirements.txt python data/generate_data.py streamlit run app.py ``` ## Dashboard Features - Interactive facility map with type-coded markers and state gap overlays - KPI cards: active facilities, underserved states, avg travel time - Gap score bar chart (worst to best coverage) - Facility type donut chart - Coverage vs population scatter by geopolitical zone - Priority tier ranking table with WHO benchmark comparisons ## Data Sources (Production) - **FMOH**, Federal Ministry of Health facility registry - **NBS**, Population and housing census - **WHO SARA**, Service availability and readiness assessment - **OpenStreetMap Nigeria**, Road network for travel-time calculation