This repository contains code to compute the Sequential Rank Agreement (SRA) between two or more ordered lists as specified in the 2019 paper Sequential rank agreement methods for comparison of ranked lists by Claus Thorn Ekstrøm, Thomas Alexander Gerds, and Andreas Kryger Jensen in the journal Biostatistics, doi:10.1093/biostatistics/kxy017.

It includes a partial verification problem in which the results of Table 1(c) are reproduced.

Dependencies

The code was written and tested on Python 2.7 for two ordered lists of the same length. It requires the following packages

csv
numpy
itertools
matplotlib

How can I use this code?

1. Save your ordered lists as columns (with a header row) in a csv.
2. Specify the filepath of the csv and the number of columns in main.py.
3. Run plot_sra(filepath, nCols) to compute the sequential rank agreement between your lists and plot how it varies with increasing list depth.

import sra
    
filepath = 'myLists.csv'
nCols = 2
plot_sra(filepath, nCols)