Denny-B-Justin/pia-frontend-malawi

# Malawi Health Facility Accessibility Dashboard A **Databricks Data App** built with Dash and Folium that visualises the location and population accessibility of health facilities across Malawi. It allows users to simulate the addition of up to 30 optimally-placed new facilities and instantly see the projected improvement in population coverage. --- World Bank Data Analytics: Malawi Geospatial Hub: ## Table of Contents - Overview - Features - Project Structure - Architecture - Data Sources - Prerequisites - Local Setup - Environment Variables - Running the App - Deploying to Databricks Apps - How the Optimisation Logic Works - Troubleshooting --- ## Overview This dashboard is part of the **GoAT (Geospatial Optimisation and Accessibility Tool)** initiative. It connects to Unity Catalog tables in Databricks to visualise: - All **existing** health facilities in Malawi (~1,258 sites) - **Proposed new** facility locations ranked by optimisation score - **Population accessibility** percentage before and after simulated additions The optimisation is pre-computed using an Integer Linear Programming (ILP) model (Gurobi) and stored in a results table. The app reads and visualises those results interactively. --- ## Features - **Interactive Folium map** — dark-themed, with layer controls and hover tooltips - **Orange markers** — existing health facilities - **White markers (sky-blue ring)** — proposed new facilities - **KPI scorecards** — existing count, new count, accessibility %, total facilities - **Slider (0–30)** — simulate adding 1 to 30 new facilities in real time - **In-memory TTL cache** — Databricks is queried once per session; subsequent interactions are instant - **OAuth M2M authentication** — service principal credentials for secure production access --- ## Project Structure ``` your-app/ │ ├── app.py # Dash application — layout, callbacks, UI ├── queries.py # QueryService singleton — Databricks SQL + cache ├── utils.py …