View source
Visit project1 Background and motivation
Disease data becomes difficult to compare when geography, reporting periods, and indicators are shown separately. I built this dashboard to bring those dimensions into one exploration interface. A reader can move between state and district views, inspect a date range, and compare temporal trends before exporting a map or report for use outside the application.
2 Data and application structure
The application uses Next.js API routes with Prisma and SQLite for its documented storage layer. Geographic and indicator records feed the map and chart components, while Zustand manages interface state such as the active selection. Keeping filters in a shared state model lets the map, legend, and comparison views refer to the same analytical context.
Leaflet supplies interactive geography and Recharts renders temporal views. Hover details provide the values behind the map’s colours, and the legend explains the active ranges. The interface supports multiple palettes, including sequential and diverging schemes, so the visual encoding can be selected for the comparison being made.
3 Comparison and export
The map uses quantile-based ranges, which distribute observations into ranked groups. This helps reveal differences within a selection, but it also means that a colour cannot be interpreted without its legend. Changing the date range or geographic scope may change the value boundaries, so comparisons must consider the displayed ranges as well as the colours.
PNG, SVG, and PDF exports allow the chosen view to leave the interactive application. The implementation uses image-generation and PDF tooling alongside map and comparison exports. Exporting is part of the workflow because a useful analysis often needs to be placed in a presentation or document with the same context that made it readable on screen.
4 Scope of the dashboard
The project is a visualization tool for the data loaded into it, not an independent source of epidemiological validation. Import utilities and sample-data generation support development and different datasets. Before interpreting any view, its reporting period, geographic coverage, and source data need to be understood; an attractive map alone cannot establish the completeness or comparability of the underlying records.