Skip to content
Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

This pipeline contains code to replicate the plots in:

"Driver Surge Pricing." Nikhil Garg and Hamid Nazerzadeh. Management Science, 2021. URL:

If you use the code in this directory, please cite the above work.

Data source:

Usage instructions:

  1. Download data from above link (Rides_DataA.csv, Rides_DataB.csv) and put in data/rideaustin/ folder
  2. Create a plots/ folder in home directory
  3. Run code as outlined below for various purposes. In general, contains a pipeline to run the analyses and data-preprocessing. The pipeline function takes in as input a settings dictionary, with several examples in

To obtain a clean version of the dataset with merged rows and errors replaced:

  1. Run pipeline(settings_server_preprocessing)

To replicate all plots in paper from scratch:

  1. Run pipeline(settings_server_2months). (warning: default settings use 55 cores, to do the matching in parallel. With this many cores, the code takes a few hours.)
  2. Run pipeline(settings_plotting_2months). This code will use the output of the above command and generate the plots.


  • takes the polygon.csv from online and closes the polygons so that they define fixed areas. Also selects a subset of the polygons that have less overlap. This code is not used in the current analysis.
  • takes the clean rides file and converts the trip origin/destination longitude/latitude locations to the polygons. This step is time-consuming. This code is not used in the current analysis.
  • merges Rides_DataA.csv, Rides_DataB.csv, and converts time-zones, and replaces incorrect values with NaN.


No description, website, or topics provided.






No releases published


No packages published