Topometry Versions Save

Systematically learn and evaluate manifolds from high-dimensional data

v0.0.1

3 years ago

The initial release of TopOMetry. As of now, users can:

  • Learn topological metrics with diffusion harmonics and continuous k-nearest-neighbors
  • Learn topological basis with multiscale Diffusion Maps, or with diffuse or continuous Laplacian Eigenmaps
  • Learn topological graphs from the learned basis
  • Optimize the graph layout with an adapted UMAP implementation
  • Use pyMDE within TopOMetry