torch_topological Logo
latest

Getting Started

  • Installing and using torch_topological
  • Topological machine learning in a nutshell
  • Comparison with other packages
  • Examples

Modules

  • torch_topological.data
  • torch_topological.nn
  • torch_topological.utils
torch_topological
  • torch_topological – Topological Machine Learning with pytorch
  • Edit on GitHub

torch_topological – Topological Machine Learning with pytorch

pytorch-topological, also known as torch_topological, brings the power of topological methods to your machine learning project. torch_topological is specifically geared towards working well with other PyTorch projects, so if you are already familiar with this framework you should feel right at home.

Getting Started

  • Installing and using torch_topological
    • Requirements
    • Installation via pip
    • Installation from source
  • Topological machine learning in a nutshell
    • High-level View
    • Additional Resources
  • Comparison with other packages
    • giotto-tda
    • Teaspoon
    • TopologyLayer
  • Examples
    • Autoencoders with geometrical–topological losses
    • Point cloud optimisation with summary statistics

Modules

  • torch_topological.data
    • sample_from_annulus()
    • sample_from_disk()
    • sample_from_sphere()
    • sample_from_torus()
    • sample_from_unit_cube()
  • torch_topological.nn
    • AlphaComplex
    • CubicalComplex
    • EulerDistance
    • MultiScaleKernel
    • PersistenceInformation
    • SignatureLoss
    • SlicedWassersteinDistance
    • SlicedWassersteinKernel
    • SummaryStatisticLoss
    • VietorisRipsComplex
    • WassersteinDistance
    • WeightedEulerCurve
  • torch_topological.utils
    • is_iterable()
    • nesting_level()
    • pairwise()
    • persistent_entropy()
    • polynomial_function()
    • total_persistence()
    • wrap_if_not_iterable()
    • SelectByDimension

Indices and tables

  • Index

  • Module Index

  • Search Page

Next

© Copyright 2022, Bastian Rieck. Revision 82248dd5.

Built with Sphinx using a theme provided by Read the Docs.