This repo contains the scripts used for my Insight project. Note that for privacy reasons, data on addresses and API keys are not included in this repo.
The 'code' folder contains four scripts used to compile, process, model, and visualize the data.
geocode.py- geocode user addresses using the Googlemaps API.get_census_data.R- request Ontario census data using the cancensus API, combining with user data (geocodes, subscription records), and writing shapefiles (for making maps) and user-census data (for modeling).model_and_visualize.py- fit logistic regression model and make visualizations.VIFs.py- compute variance inflation factors for features.
The 'application' Folder contains the script (app.py) and data for the dash app (www.datenightdash.com). Note that the user locations provided in this folder have been jittered, so they are de-identified.