1: \begin{abstract}

2: Several applications in optimization, image and signal processing deal with data that belong to the Stiefel manifold $ \Stnp $, that is, the set of $n\times p$ matrices with orthonormal columns. Some applications, like the Riemannian center of mass, require evaluating the Riemannian distance between two arbitrary points on $ \Stnp $. This can be done by explicitly constructing the geodesic connecting these two points.

3: An existing method for finding geodesics is the leapfrog algorithm of J. L. Noakes. This algorithm is related to the Gauss--Seidel method, a classical iterative method for solving a linear system of equations that can be extended to nonlinear systems. We propose a convergence proof of leapfrog as a nonlinear Gauss--Seidel method. Our discussion is limited to the case of the Stiefel manifold, however it may be generalized to other embedded submanifolds.

4: We discuss other aspects of leapfrog and present some numerical experiments.

5: 

6: 

7: %Include keywords and mathematical subject classification numbers as needed.

8: \bigskip

9: \textbf{Key words.} Riemannian manifolds, geodesics, Stiefel manifold, nonlinear Gauss--Seidel

10: 

11: \medskip

12: \textbf{AMS subject classifications.} 65L10, 49Q12, 65K10

13: 

14: \end{abstract}

15: