Skip to content

Latest commit

 

History

History
64 lines (43 loc) · 1.74 KB

README.rst

File metadata and controls

64 lines (43 loc) · 1.74 KB
Raw

NetLSD

NetLSD is a family of spectral graph descriptros. Given a graph, NetLSD computes a low-dimensional vector representation that can be used for different tasks.

Quick start

import netlsd
import networkx as nx

g = nx.erdos_renyi_graph(100, 0.01) # create a random graph with 100 nodes
descriptor = netlsd.heat(g) # compute the signature

That's it! Then, signatures of two graphs can be compared easily. NetLSD supports networkx, graph_tool, and igraph packages natively.

import netlsd
import numpy as np

distance = netlsd.compare(desc1, desc2) # compare the signatures using l2 distance
distance = np.linalg.norm(desc1 - desc2) # equivalent

For more advanced usage, check out online documentation.

Requirements

  • numpy
  • scipy

Installation

  1. cd netlsd
  2. pip install -r requirements.txt
  3. python setup.py install

Or simply pip install netlsd

Citing

If you find NetLSD useful in your research, we ask that you cite the following paper:

@inproceedings{Tsitsulin:2018:KDD,
 author={Tsitsulin, Anton and Mottin, Davide and Karras, Panagiotis and Bronstein, Alex and M{\"u}ller, Emmanuel},
 title={NetLSD: Hearing the Shape of a Graph},
 booktitle = {Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining},
 series = {KDD '18},
 year = {2018},
}

Misc

NetLSD - Hearing the shape of graphs.