The code has been heavily modified since the last iteration to better fit the format of an ipynb.
It now comes in two parts:
- HETI's algorithm and the problem of stacking
- Exploring the Queensland Ballot algorithm as an alternative to simulated annealing
If you came here looking for the original article, please go to the link(s) above where a much nicer interactive version of this article is hosted. The notebook files have a lot of graphs in them so may take a while to load.
Either way, this represents a significant improvement (I hope) as far as usability and code quality is concerned.