Tools for Inference, Learning, and Optimization on Stiefel Manifold

Stiefel manifold is a set of orthonormal frames in Euclidean space. We provide algorithms for statistical inference, optimization, and learning over the Stiefel manifold.


You can install the released version of RiemStiefel from CRAN with:


And the development version from GitHub with:

# install.packages("devtools")

Available Functions

function description
st.mean Frechet mean computation.
st.runif random sample generation from uniform distribution.
st.utestR test of uniformity.
stopt.SA Simulated Annealing algorithm for optimization.