Natural barriers facing female cyclists and how to overcome them: a cross national examination of bikesharing schemes
Description
Data and software for examining the role of natural barriers in over 200 million trips from bikesharing data from 10 cities. The data is integrated with weather data, gradient data, and sunrise/sunset data and examined by gender (male / female self-reported) to check for significant differences between gender using Generalized Additive Models (GAMs).
Files
Steps to reproduce
The major file for weather analysis is "analysis.R". A summary of the GAM models (with linear terms for precipitation and wind) is found in the two "parsimonious-models" output files. The bike sharing usage data, except for Minneapolis 2011/12, Brisbane, and Helsinki, is available from public web repositories as CSV files (some zipped). "cc-daylight.R" calculates gendered percentages of trips before sunrise and after sunset "get-elevation-sd.R" and "gradient-mf-testing.R" contain great circle elevation code and gradient testing functions "gen-age.R" produces the gender/age plot "get-elev.R" is example code for extracting Brisbane/NYC elevations "get-jaxa-elevation" files illustrate how to extract JAXA elevations (lat/long files available on request) "gradient-electric-classic.R" analyses electric vs classic gradients in Bay Area / NYC systems "[cityname]-el.csv" contains summaries of the elevations of bike sharing stations (from JAXA data) "[cityname].csv" contains historical weather data for each city (using bilinear interpolation from oikolab.com ERA5 data for each city point) "[cityname]-gen-age.txt" contains gender/age count data "get-gender-age.bash" is a shell script for getting the gender/age counts (sans Brisbane) "average.sh" is example shell code for averaging summary data The Excel files contain figures for the paper and the data used to produce these figures.