| rmLB_sm {SMFilter} | R Documentation |
Sample from the matrix Langevin-Bingham on the Stiefel manifold.
Description
This function draws a sample from the matrix Langevin-Bingham on the Stiefel manifold.
Usage
rmLB_sm(num, mJ, mH, mC, mX, ir)
Arguments
num |
number of observations or sample size. |
mJ |
symmetric ip*ip matrix |
mH |
symmetric ir*ir matrix |
mC |
ip*ir matrix |
mX |
ip*ir matrix, the initial value |
ir |
ir |
Details
The matrix Langevin-Bingham distribution on the Stiefel manifold has the density kernel:
f(X) \propto \mathrm{etr}\{ H X' J X + C' X \}
where X satisfies X'X = I_r, and H and J are symmetric matrices.
Value
an array containing a sample of draws from the matrix Langevin-Bingham on the Stiefel manifold.
Author(s)
Yukai Yang, yukai.yang@statistik.uu.se
#' @section References: Hoff, P. D. (2009) "Simulation of the Matrix Bingham—von Mises—Fisher Distribution, With Applications to Multivariate and Relational Data", Journal of Computational and Graphical Statistics, Vol. 18, pp. 438-456.