This repository provides the code used in a manuscript titled 'Supervised temporal link prediction in large-scale real-world networks'.

Temporal networks used in this study are downloaded from http://konect.cc/ or http://snap.stanford.edu.

The structure of the directories is as follows:

• code: Contains all code that is not part of tlp package.
• data: Folder where all downloads and intermediate results are stored.
• teexgraph: External dependency to calculate the diameters and shortest path lengths in a network really fast. See .gitmodules.
• tlp: Here most of my code resides. I import this folder as a package in most Jupyter Notebooks.
• cleanup.sh: Clean up Python/ iPython cache folders.
• environment.yml: Can be used to create Python environment with Conda.
• install.sh: Install teexGraph.
• spec-list.txt: Can be used to create exactly the same Python environment.