delvinso / divvy-chicago

Code for '14 Million Bike Rides in The Windy City (2013 - 2017)' on my site

Divvy Bikeshare PostgreSQL Database & Analyses

This repo contains the code necessary to setup the database and all analyses used in my post 14 Million Bike Rides in The Windy City (2013 - 2017) - An Analysis of Chicago’s Bikeshare System.

1. Download the bikeshare data.

download_concat_data.sh

Downloads the raw trip data from 2013 - 2017 into the sub-directory /data within the current folder and concatenates the raw trip data into one file.

2. Run setup_import_data.sh, which does the following:

a) Creates a database named 'divvy-bikeshare-data' with appropriate schema.

01_create_schema.sql

b) Imports raw trips, station data, community areas and census tracts.

02_import.sql

c) Cleans the raw trip and station data and maps the station data onto community areas and census tracts.

03_tables_for_analyses.sql

Analyses

SQL queries and analyses can be found in analyses/.

Static Heat Map of Trips

Additionally, if one wants to recreate the static heat map, look at get_legs_routes_key_rm.R.

NOTE: This may cost $$$$ with the new pricing changes Google implemented with their Google Maps API mid-July 2018. May want to consider looking into OSRM as a free, open-source alternative.

The legs can then be imported using 04_import_legs.sh, making sure to run it from the top level directory of the repo or the following command in a unix terminal:

psql chicago-divvy-data -c "SET datestyle = 'ISO, MDY'; COPY legs FROM 'ABSOLUTE_PATH/data/legs_start_end.csv' CSV HEADER;".

TO DO

• Review Indexes
• Review transports.R - OVER and WINDOW functions