This repository contains an implementation of Trimmed Grassmann Average (TGA) by Hauberg S et al. in Python. The implementation follows the same logic as the original Matlab implementation provided by the authors at http://ps.is.tue.mpg.de/research_projects/robust-pca

It is implemented with scikit-learn API.

Add tests

Hauberg, Soren, Aasa Feragen, and Michael J. Black. "Grassmann averages for scalable robust PCA." Computer Vision and Pattern Recognition (CVPR), 2014 IEEE Conference on. IEEE, 2014.