Today I present to you Livablestreets (https://livablestreets.streamlit.app/), an open source web-tool that I co-developed. Livablestreets helps you find the next home you’ll love!
Not into reading? Check out my presentation on YouTube: Le Wagon Berlin Demo Day | Batch 871 | Data Science - YouTube
Want to see the code? Here: GitHub - chvieira2/livablestreets: A tool for indexing livability conditions at street level
The code is written in Python, and hosted and published with Streamlit. It uses geographical data from OpenStreetMap Foundation using the Overpass API and scrapes housing ads from a big german real estate page. The original idea is from Laia Grobe De Luca (github: Laiagdla) and myself, with initial technical contributions from Ieva Bidermane (github: ievabiM) and Nicolas Quiroga (github: nicoquiroga941). Since September 2022 the webservice is maintained by deep-urbanism, which I am a co-founder.
Here’s how Livablestreets works:
The app creates a grid over the map of the city of interest and collects info on ~200 types of features from OpenStreetMaps (bus stops, parks, restaurants, etc). The absolut count of each feature per tile on the grid (feature density) is blurried over neighboring tiles with a gaussian function, transformed by a polynomial function to remove incorrect linear relationship to livability, and scaled according to the feature relevance.
Features are grouped in 4 main categories: Activities and Services, Mobility, Comfort and Social Life. By allowing the user to set weights for each category, Livablestreets uses the weighted mean to calculate the livability scores for each tile on the grid. That way, living conditions are estimated according to the user own needs! #givebackpowertotheuser.
The result is a colorful map of the city of interest that matches the user’s living quality standard. And if you are looking for a new home, the app scrapes ads from wg-gesucht.de to display the most up-to-date housing options (only available for cities in Germany).
Livablestreets is ready to help you finding the neighborhood you love! Try it out and share it with your network so others can also benefit.
Constructive feedback is always appreciated. Please share!