Skip to content

aywagner/DIPOLE

Repository files navigation

DIPOLE

DIstributed Persistence-Optimized Local Embeddings

RUNME.py will download/process the datasets, create the visualizations, and run the analysis found in the manuscript.

brain1.npy is derived from data made available at https://www.insight-journal.org/midas/community/view/21 by Bullitt, Elizabeth; Smith, J Keith; Lin, Weili.

Mammoth data downloaded from https://github.com/PAIR-code/understanding-umap and originally from https://3d.si.edu/object/3d/mammuthus-primigenius-blumbach:341c96cd-f967-4540-8ed1-d3fc56d31f12.

Stanford faces dataset face_data.mat downloaded from http://isomap.stanford.edu/datasets.html.

About

DIstributed Persistence-Optimized Local Embeddings

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages