This repository contains code and data for reproducing our study of the effect of Nike Vaporfly shoes on marathon performance.

reproduce.txt - instructions for reproduction

scrape.py - code for scraping data from marathonguide.com

generate_sampled_dataset.R - sample performances from scraped data

merge.R - merge sampled data with shoe data

explore_data.R - exploration of data

analysis.R - fit linear mixed model to data

explore_fit.R - exploration of model fit

race_codes.csv - information about sampled races

men_matches.csv, women_matches.csv - name matches

men_shoe.csv, women_shoe.csv - manually curated shoe datasets