Capital Crashpad: Visualizing the Airbnb Market in Washington, DC

Explore, visualize, and understand DC’s Airbnb landscape—interactive dashboards and maps powered by open data and modern web tech.

🔗 Live Website
🔗 Tableau Dashboard
🔗 Exploratory Data Analysis (EDA)

⚠️ ℹ️ Status: This project is under active, rapid refinement. Adding a summary analysis and updating the dashboard for the GitHub Pages version.

Table of Contents

• Project Overview
• Features
• Tools & Technologies
• Usage
• Gallery
• References
• License
• Acknowledgements
• Author

Project Overview

This full-stack project explores the landscape of short-term rentals in Washington, DC. Using data from Inside AirBnB, it combines:

• A live, interactive web dashboard hosted on GitHub Pages
• A Tableau story-style EDA for strategic analysis
• A PostgreSQL backend with Flask and Django implementations
• A Jupyter-based ETL pipeline that transforms and loads data for multiple outputs

The goal is to visualize the availability, pricing, and neighborhood distribution of short-term rentals and highlight interesting patterns in the DC vacation rental market.

Features

• Dynamic map with multiple layers and neighborhood overlays
• Interactive plots aggregating neighborhood trends
• Tableau dashboard for filtering and exploring rental trends
• Clean ETL pipeline to prep and load data across platforms
• Choropleth and bubble maps to compare neighborhoods
• Flask and Django versions for backend experimentation
• Mobile-friendly static version with JavaScript interactivity

Tools & Technologies

• Data: Inside AirBnB, CSV, PostgreSQL, SQLite
• Backend: Flask, Django, SQL, Jupyter, Pandas
• Frontend: JavaScript, Plotly, Leaflet, Bootstrap, HTML/CSS
• Visualization: Tableau, Choropleths, Bubble Maps, Plotly
• Deployment: GitHub Pages, Tableau Public

Usage

Website

• Visit GitHub Pages version
• Explore listings, prices, and availability
• Navigate through the embedded Tableau dashboard

ETL Pipeline

• Edit and run ./notebooks/data_pipeline_postgres_export.ipynb to clean and update source data
• Automatically loads PostgreSQL tables and exports data for Flask and GitHub Pages

Full-Stack Versions

• Flask app: flask/app.py (pulls from PostgreSQL)
• Django app: django/django_airbnb_dc (uses SQLite)

Data Backup

Restore the included PostgreSQL backup:

pg_restore -U <username> -d <dbname> -1 airbnb.backup

Note to Self

• Run all cells in ./notebooks/data_pipeline_postgres_export.ipynb to update the source data
• Recheck plotPriceAvailability annotations in plots.js after updates
• Review schema.sql and neighbourhoods_dict in data_processing.ipynb
• Export updated map_listings and price_availability views to CSV for GitHub Pages
• Leave Tableau as-is (June 2024)

Errata

The Flask / Django version of the price-availability chart has formatting issues. Be sure to toggle the allDCData filter correctly in plotPriceAvailability().

Gallery

Tableau Dashboard:

Web Dashboards Plots:

Interactive Map:

Exploratory Data Analysis:

Entity Relationship Diagram:

Course Certificates

Parts of this project served as capstones for two of HarvardX’s free online computer science courses:

• CS50's Web Programming with Python and JavaScript
• CS50's Introduction to Databases with SQL

References

Dataset provided by Inside AirBnB.

Neighborhood population and housing unit data from Census Reporter.

License

Creative Commons Attribution 4.0 International License

Acknowledgements

• Thanks to Imen Najar for early insights and support.
• Thanks to Geronimo Perez for feedback and assistance during development.

Author

Bryan Johns, October 2024
[email protected] | LinkedIn | GitHub | Portfolio
— Fluent in Data. Fluent in Human.