holq.From the output of holq, this function will calculate the
MLEs for the component covariance matrices and for the total
variation parameter.
mle_from_holq(holq_obj)The output returned from holq.
cov_mle A list of positive definite matrices. These
are the MLEs for the component covariance matrices.
sig_mle A numeric. This is an estimate of the "standard
deviation" form of the total variation parameter.
The function simply takes the A[[i]] output of holq
and returs A[[i]] %*% t(A[[i]]). The estimate of the total
variation parameter is sqrt(sig ^ 2 / prod{p}), whre p is the
vector of dimensions of the data array and sig is the output
from holq.
Gerard, D., & Hoff, P. (2016). A higher-order LQ decomposition for separable covariance models. Linear Algebra and its Applications, 505, 57-84. https://doi.org/10.1016/j.laa.2016.04.033 http://arxiv.org/pdf/1410.1094v1.pdf
holq.