This pull request begins integration of my work with John's existing project. Some points to note:
My contributions do not yet follow the atomic task structure used by John. This is planned to follow---I need to figure out a way to handle shared Python code between tasks.
Some analysis (311 and procurement) is not yet present. These Jupyter notebooks are particularly messy and their reconstruction is a work in progress.
I believe the Makefile for the data generating process can handle a cold start---that is, make all downloads all necessary data, does all geocoding etc., and produces .parquet files ready for analysis. I will test this tomorrow on a clean computer.
A Mapbox API key is expected to be located at ~/.keys/mapbox_be.key. This can trivially be changed to use another geolocating service.
This pull request begins integration of my work with John's existing project. Some points to note:
make alldownloads all necessary data, does all geocoding etc., and produces .parquet files ready for analysis. I will test this tomorrow on a clean computer.