This folder contains the code to reproduce the numerical results (Figures 4, 5, 6 and data for 7) reported in:
``Active Ranking from Pairwise Comparisons and When Parametric Assumptions Don't Help'', by Reinhard Heckel, Nihar B. Shah, Kannan Ramchandran, and Martin J. Wainwright
An introduction and exposition of the algorithm is given in the jupyter notebook:
active_ranking_introduction.ipynb
The code is written in Python 3, running requires installations of Python, Make, and Pdflatex.
Run:
make fig4
to generate Figure 4 (this will run the code to generate the data for Figure 5, store the data in the ./fig folder, and generate the file ./tex/fig4.pdf containing the figure.)make fig5
to generate Figure 5make fig6
to generate Figure 6
All files are provided under the terms of the Apache License, Version 2.0, see the included file "apache_licence_20" for details.