Graph Drawing Benchmark

Welcome to Graph Drawing Benchmark! The purpose of the repository is to provide a standard benchmark and a set of toolkits for evaluating graph layout algorithms. The benchmark includes a comprehensive set of evaluation metrics and baseline layout algorithms. It aims to streamline and simplify the evaluation process for researching and developing graph layout algorithms.

Layout Evaluation

python main.py eval \
--layout-json example_layouts.json \
--baselines neato,kamada_kawai \
--metrics-output metrics.csv \
--spc-output spc.csv \
--metrics stress,xing,ring \
--split test

• --layout-json: Path to the input JSON file.
• --baselines: Comma-separated list of baseline layout algorithms.
• --metrics-output: Path to the output CSV file for evaluation metrics.
• --spc-output: Path to the output CSV file for metrics SPCs.
• --metrics: Comma-separated list of evaluation metrics.
• --split: Dataset split to evaluate. (train, val, test)

Input format

Input file is a JSON file with the following format:

{
  "<graph1_id>": [
    [<node1_x>, <node1_y>],
    [<node2_x>, <node2_y>],
    ...
  ],
  "<graph_id_2>": [
    [<node1_x>, <node1_y>],
    [<node2_x>, <node2_y>],
    ...
  ],
  ...
}

See example_layouts.json for an example.

Show graph IDs for a specific dataset split

python main.py info splits --name <split>

Show available baseline layout algorithms

python main.py info baselines

Please see Baselines.md for more details on the baseline layout algorithms.

Show available evaluation metrics

python main.py info metrics

Please see Metrics.md for more details on the evaluation metrics.

Citation

If you used our code or find our work useful in your research, please consider citing:

@ARTICLE{smartgd,
  author={Wang, Xiaoqi and Yen, Kevin and Hu, Yifan and Shen, Han-Wei},
  journal={IEEE Transactions on Visualization and Computer Graphics}, 
  title={SmartGD: A GAN-Based Graph Drawing Framework for Diverse Aesthetic Goals}, 
  year={2024},
  volume={30},
  number={8},
  pages={5666-5678},
  keywords={Layout;Graph drawing;Deep learning;Generative adversarial networks;Stress;Generators;Training data;Deep learning for visualization;generative adversarial networks;graph visualization},
  doi={10.1109/TVCG.2023.3306356}}

@article{deepgd,
author={Wang, Xiaoqi and Yen, Kevin and Hu, Yifan and Shen, Han-Wei},
journal={IEEE Computer Graphics and Applications},
title={DeepGD: A Deep Learning Framework for Graph Drawing Using GNN},
year={2021},
volume={41},
number={5},
pages={32-44},
doi={10.1109/MCG.2021.3093908}
}