total harmonic resistance for edge deletions

code for the paper 'Introducing Total Harmonic Resistance for Graph Robustness under Edge Deletions' (arxiv).

This repository contains the greedy algorithm for THR and our experimental setup (via simexpal)

how to run experiments

generating / loading instances

The case study instances can be generated using the generate-osm.py script. Requirements for pip can be found in the top-level requirements.txt. Additionally, the NetworKit version found in extern/networkit has to be installed via pip install extern/networkit.

All other instances can be found on snap, networkrepository, and/or konect. Before running the experiments, all files have to be converted to the networkit binary format (eg by just reading and saving all files once with networkit in python; for the MatrixMarket format, a converter tool is available in apps).

launching experiments

1. navigate to experiments dir
2. install python requirements: pip install -r requirements.txt (you should use a different venv to the one used for instance installing!)
3. build with simexpal simex b make
4. launch experiments simex e launch
5. evaluate: run the jupyter notebooks contained in this directory.