Robust PCA in Python! Open the notebook to see it in action I plan on getting this into sklearn TODO: Write tests, and sklearn interface Need to figure out why scipy sparse svd won't compute all the singular values of a matrix Prove it out on some real world data sets