R/logmap_vecs_spd.R

In mrparker909/MGLMRiem: Multivariate Generalized Linear Models on Riemannian Manifolds

# function V = logmap_vecs_spd(X,Y)
# %LOGMAP_VECS_SPD returns logmap(X,Y) for SPD manifolds.
# %
# %    V = LOGMAP_VECS_SPD(P,X)
# %
# %    X, Y is a set of SPD matrices (dimX x dimX x N, where dimX = size(X,1)).
# %    V is a set of symmetric matrices.
# %
# %   See also LOGMAP_SPD, LOGMAP_PT2ARRAY_SPD, EXPMAP_SPD
# 
# %   Hyunwoo J. Kim
# %   $Revision: 0.1 $  $Date: 2014/06/23 15:42:13 $
#   
#   
#   V = zeros(size(Y));
# if size(X,3) ==1 
# for i = 1:size(Y,3)
# yi = Y(:,:,i);
# V(:,:,i) = logmap_spd(X,yi);
# end
# else
#   for i = 1:size(X,3)
# xi = X(:,:,i);
# yi = Y(:,:,i);
# V(:,:,i) = logmap_spd(xi,yi);
# end
# end
# 

#' @export
logmap_vecs_spd <- function(X,Y){
#LOGMAP_VECS_SPD returns logmap(X,Y) for SPD manifolds.
#
#    V = LOGMAP_VECS_SPD(P,X)
#
#    X, Y is a set of SPD matrices (dimX x dimX x N, where dimX = size(X,1)).
#    V is a set of symmetric matrices.
#
#   See also LOGMAP_SPD, LOGMAP_PT2ARRAY_SPD, EXPMAP_SPD

#   Hyunwoo J. Kim
#   $Revision: 0.1 $  $Date: 2014/06/23 15:42:13 $

#   Migrated to R by Matthew RP Parker
#   $Revision: 0.2 $  $Date: 2019/06/07 $   
  if(is.na(dim(Y)[3])) { Y <- aug3(Y) }
  if(is.na(dim(X)[3])) { X <- aug3(X) }
  
  V = array(0, dim=sizeR(Y))#zeros(size(Y));
  if(sizeR(X,3)==1) { 
    for(i in 1:sizeR(Y,3)) {
      yi = Y[,,i]
      V[,,i] = logmap_spd(drop(X),yi)
    }
  } else {
    for(i in 1:sizeR(X,3)) {
      xi = X[,,i]
      yi = Y[,,i]
      V[,,i] = logmap_spd(xi,yi)
    }
  }
  return(V)
}