Functions and C++ Header Files for Computation on Manifolds

We provide a number of algorithms to estimate fundamental statistics including Fréchet mean and geometric median for manifold-valued data. Also, C++ header files are contained that implement elementary operations on manifolds such as Sphere, Grassmann, and others. See Bhattacharya and Bhattacharya (2012) if you are interested in statistics on manifolds, and Absil et al (2007, ISBN:9780691132983) on computational aspects of optimization on matrix manifolds.

Installation

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

install.packages("RiemBase")

or the development version from github:

# > install.packages("devtools")
# if not installed.
library(devtools)
devtools::install_github("kisungyou/RiemBase")

News

News for Package RiemBase

Changes in version 0.2.2

Function rbase.curvedist is added to compute L2 distance between two curves on manifolds.
Vector-valued data handling is fixed. Bug reports are welcome.

Changes in version 0.2.1

NEWS initialized. Updates will be recorded here.
Computation of (intrinsic) Frechet means are improved:
- parallel option provided using OpenMP support.
- Wolfe condition becomes obsolete. Faster convergence is still achieved from fixed step size.