- We implemented the Graph Reduction algorithm on adjacency matrices, with some minor optimizations:
- By counting the number of remaining processes, and comparing it to the number of processes eliminated so far, we can detect when the algorithm has failed a single time, after which it'll fail forever.
-
### Benefits:
- The graph reduction algorithm is very elegant, but also easy to optimize
- "Deleting" a node from an adjacency matrix can be performed by zeroing its respective row and column and re-running the algorithm, which is very simple!
### Drawbacks:
- The adjacency matrix representation is used without interpretation, making the output somewhat hard to read
- Our implementation of knot detection didn't pan out.