/crime-prediction

Figure out how safe a given time and place in SF is based on ML models using recent police data

Primary LanguageJupyter Notebook

SF Crime Predictor 🔮

Figure out if a given time and place in SF is safe or not. The applied-ML model is based on millions of incident reports across 15 years made public by the San Francisco Police Department.

Preprocessed Dataset

https://drive.google.com/open?id=1JddXMyapWG0qLpecVmBYc3jgS0cdJ3za

Requirements

  • Python 3
  • npm
  • Google Maps Geocoding API Key

Getting Started

  1. Install dependencies in both services

    • cd front-end and npm i
    • cd back-end and pip3 install -r requirements.txt
  2. Export Google Maps API Key and start up Flask backend

    $ cd back-end
    $ export SFCRIME_API_KEY='<Your Google Maps API here>'
    $ python3 app.py
    
  3. Export API Key again and start up React frontend

    $ cd front-end
    $ export REACT_APP_SFCRIME_API_KEY='<Your Google Maps API here>'
    $ npm start
    

The app should be running on http://localhost:3000