school_choice_python

This repository contains Python code to compute assignments for school choice problems Abdulkadiroglu and Sönmez (2003) according to different assignment procedures. It uses and modifies code from Jeremy Kun, stable-marriage, (2014), GitHub repository, https://github.com/j2kun/stable-marriages, described in one of Jeremy's blog posts at http://jeremykun.com/2014/04/02/stable-marriages-and-designing-markets/.

Features

The main scripts in this repository are

DASolver.py : an almost exact copy of stablemarriage.py from https://github.com/j2kun/stable-marriages allowing to compute the outcome of a school choice problem in which schools have many seats available following the Deferred-Acceptance procedure. The only difference with stablemarriage.py lies in the naming of the classes, which has been altered to fit with the school choice context (classes Suitor and Suited become Student and School).
IASolver.py : a variation on stablemarriage.py allowing to compute the outcome of a school choice problem using the Immediate Acceptance procedure, also know as Boston Mechanism. Another difference with stablemarriage.py is the addition of a verbose option to the main function in IASolver.py which prints a description of the steps followed by the Immediate Acceptance procedure as the procedure unfolds.
Extensions.md: describes how to use DASolver.py and IASolver.py in settings where
- The total number of seats in all the schools is smaller than the total number of students, and some students end up unassigned.
- The school choice procedure are constrained in the sense that students can only rank a subset of schools.
- Some schools are not acceptable to some students.

Abdulkadiroglu, Atila, and Tayfun Sönmez. "School choice: A mechanism design approach." The American Economic Review 93.3 (2003): 729-747.
Haeringer, Guillaume, and Flip Klijn. "Constrained school choice." Journal of Economic Theory 144.5 (2009): 1921-1947.