Geospatial Data Visualization (Streamlit + Mapbox)

Live Demo

Available here: https://streamlit-geodata-visualization-jyiexkhtva-lm.a.run.app/

Description

This repo is a demonstration of parsing geo-coordinates from address data using Mapbox.com and visualizing via awesome Streamlit library.

Repo consists of two parts:

1. Geocoding of text addresses using Mapbox.com (see scripts folder)
2. Visualization of geocoded data using Streamlit and its pydeck integration (streamlit_app.py)

Dataset: NYC-Housing-Data-2003-2019

Setup

Docker

1. Run docker build -t geodata .
2. Run docker run -it -p 80:80 geodata
3. Open http://127.0.0.1:80/

Virtual Environment

1. Create virtual environment using your favorite tool (e.g. pyenv). I used Python 3.9.15 for development, but code should work for all 3.8+ versions.
2. Run pip install -r requirements.txt
3. Run streamlit run streamlit_app.py

Deploy

1. Set up an account on Google Cloud Platform
2. Install gcloud cli
3. Run gcloud run deploy (deployment via Cloud Run)
4. Note: make sure to reply y when prompted Allow unauthenticated invocations to [streamlit-geodata-visualization] (y/N)?