WeVis Datathon: Political Party Analysis

This repository contains an interactive Political Party Analysis dashboard originally developed for a datathon context. It features data visualizations detailing party characteristics, 3D PCA clustering, and longitudinal performance over 4 years.

🚀 Tech Stack

Frontend

• Framework: Next.js (App Router)
• Language: TypeScript
• Styling: Tailwind CSS
• Data Visualization (2D): D3.js (e.g., Radar/Spider Charts, Stacked Bar Charts)
• Data Visualization (3D): React Three Fiber (@react-three/fiber), @react-three/drei, Three.js (Rendered for 3D Party Clustering and PCA spatial mappings)

Backend

• Framework: Express.js (Node.js)
• Language: TypeScript
• Data Runtime Loading: Custom CSV parsing logic utilizing built-in modules (fs, csv-parse) acting as an in-memory datastore upon startup.

📊 Data Mapping

The application relies on several static .csv files stored in backend/data/ that are parsed into memory via dataStore.ts and piped to the frontend using REST APIs.

Core Datasets:

1. viz_result.csv: Contains the Principal Component Analysis (PCA) outputs (PC1, PC2, PC3) mapping each party's geographical X,Y,Z layout for the 3D Clustering component.
2. bills_cleaned_and_labeled2.csv: Contains historical records of political bills, their proposal timestamps, categories, mapping party data, and final statuses (ENACTED, REJECTED, IN_PROGRESS, etc.). Used dynamically to track "Success Rates" and "Passed Laws".
3. trend_vote.csv: Stores raw historical voting metadata.
4. vote_absence.csv: Stores attendance records reflecting party absenteeism and physical legislative representation.
5. unity.csv: Tracks the calculated internal Unity (avg_unity_score) metric of a specific political party's voting alignment.
6. entropy.csv: Contains variance metrics supporting internal data clustering measurements.