All Tools
G
DataFreeOpen Source
GEOMSTATS
Statistics and ML on Riemannian manifolds
MIT
ABOUT
Shape, pose, and covariance data live on manifolds, but Euclidean ML ops ignore that geometry. geomstats provides Riemannian metrics, geodesics, and stats so pipelines run PCA and clustering on spheres and SPD matrices correctly.
INSTALL
pip install geomstatsINTEGRATION GUIDE
1. Run Frechet mean and PCA on a sphere of unit directions
2. Cluster covariance matrices with an SPD Riemannian metric
3. Compute geodesics between poses in a robotics dataset
4. Fit a manifold-aware regression on shape landmarks
TAGS
pythongeometrymanifoldsriemannianscientificstatistics