Skip to main content

Python package for graph-based clustering and semi-supervised learning

Project description

Graph-based Clustering and Semi-Supervised Learning

Clustering

This python package is devoted to efficient implementations of modern graph-based learning algorithms for both semi-supervised learning and clustering. The package implements many popular datasets (currently MNIST, FashionMNIST, and CIFAR-10) in a way that makes it simple for users to test out new algorithms and rapidly compare against existing methods. Full documentation is available, including detailed example scripts.

This package also reproduces experiments from the paper

J. Calder, B. Cook, M. Thorpe, D. Slepcev. Poisson Learning: Graph Based Semi-Supervised Learning at Very Low Label Rates., Proceedings of the 37th International Conference on Machine Learning, PMLR 119:1306-1316, 2020.

Installation

Install with

pip install graphlearning

Required packages will be installed automatically, and include numpy, scipy, sklearn, and matplotlib. Some features in the package rely on other packages, including annoy for approximate nearest neighbor searches, and torch for GPU acceleration. You will have to install these manually, if needed, with

pip install annoy torch

It can be difficult to install annoy, depending on your operating system.

To install the most recent version of GraphLearning from the github source, which is updated more frequently, run

git clone https://github.com/jwcalder/GraphLearning
cd GraphLearning
python setup.py install --user

If you prefer to use ssh swap the first line with

git clone git@github.com:jwcalder/GraphLearning.git

Documentation and Examples

Full documentation for the package is available here. The documentation includes examples of how to use the package. All example scripts linked from the documentation can be found in the examples folder.

Older versions of GraphLearning

This repository hosts the current version of the package, which is numbered >=1.0.0. This version is not backwards compatible with earlier versions of the package. The old version is archived here and can be installed with

pip install graphlearning==0.0.3

To make sure you will load the old version when running import graphlearning, it may be necessary to uninstall all existing versions pip uninstall graphlearning before running the installation command above.

Citations

If you use this package in your research, please cite the package with the bibtex entry below.

@software{graphlearning,
  author       = {Jeff Calder},
  title        = {GraphLearning Python Package},
  month        = jan,
  year         = 2022,
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.5850940},
  url          = {https://doi.org/10.5281/zenodo.5850940}
}

Contact and questions

Email jwcalder@umn.edu with any questions or comments.

Acknowledgments

Several people have contributed to the development of this software:

  1. Mauricio Rios Flores (Machine Learning Researcher, Amazon)
  2. Brendan Cook (PhD Candidate in Mathematics, University of Minnesota)
  3. Matt Jacobs (Postdoc, UCLA)
  4. Mahmood Ettehad (Postdoc, IMA)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

graphlearning-1.1.0.tar.gz (65.4 kB view hashes)

Uploaded Source

Built Distributions

graphlearning-1.1.0-pp38-pypy38_pp73-win_amd64.whl (76.4 kB view hashes)

Uploaded PyPy Windows x86-64

graphlearning-1.1.0-pp38-pypy38_pp73-macosx_10_7_x86_64.whl (83.5 kB view hashes)

Uploaded PyPy macOS 10.7+ x86-64

graphlearning-1.1.0-pp37-pypy37_pp73-win_amd64.whl (76.4 kB view hashes)

Uploaded PyPy Windows x86-64

graphlearning-1.1.0-pp37-pypy37_pp73-macosx_10_7_x86_64.whl (83.5 kB view hashes)

Uploaded PyPy macOS 10.7+ x86-64

graphlearning-1.1.0-cp310-cp310-win_amd64.whl (76.3 kB view hashes)

Uploaded CPython 3.10 Windows x86-64

graphlearning-1.1.0-cp310-cp310-macosx_10_15_x86_64.whl (84.3 kB view hashes)

Uploaded CPython 3.10 macOS 10.15+ x86-64

graphlearning-1.1.0-cp39-cp39-win_amd64.whl (76.3 kB view hashes)

Uploaded CPython 3.9 Windows x86-64

graphlearning-1.1.0-cp39-cp39-macosx_10_15_x86_64.whl (84.3 kB view hashes)

Uploaded CPython 3.9 macOS 10.15+ x86-64

graphlearning-1.1.0-cp38-cp38-win_amd64.whl (76.3 kB view hashes)

Uploaded CPython 3.8 Windows x86-64

graphlearning-1.1.0-cp38-cp38-macosx_10_14_x86_64.whl (84.1 kB view hashes)

Uploaded CPython 3.8 macOS 10.14+ x86-64

graphlearning-1.1.0-cp37-cp37m-win_amd64.whl (76.3 kB view hashes)

Uploaded CPython 3.7m Windows x86-64

graphlearning-1.1.0-cp37-cp37m-macosx_10_14_x86_64.whl (84.1 kB view hashes)

Uploaded CPython 3.7m macOS 10.14+ x86-64

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page