Covariance Inference and Decompositions for Tensor Datasets
-mode product.
Array normal conditional distributions.
Calculate the AIC and BIC.
Array indices.
Tucker product.
Collapse multiple modes into one mode.
Convert the output from equi_mcmc to component covariance matrices.
Demeans array data.
Gibbs sampler using an invariant prior.
Frobenius norm of an array.
Get the Bayes rule under multiway Stein's loss.
Calculate the incredible SVD (ISVD).
Calculate the incredible higher-order LQ decomposition (HOLQ).
Calculate the higher-order orthogonal iteration (HOOI).
Calculate the (truncated) higher-order SVD (HOSVD).
The incredible higher-order polar decomposition (IHOP).
Kendall's tau measure of association.
Commutation matrix.
Log-likelihood of array normal model.
Element-wise matrix products between two lists.
LQ decomposition.
Draw from null distribution of likelihood ratio test statistic.
Calculate the likelihood ratio test statistic.
Unfold a matrix.
The symmetric square root of a positive definite matrix.
Get MLE from output of holq.
Calculate multiway Stein's loss from square root matrices.
Calculate multiway Stein's loss from component covariance matrices.
Calculate a truncated multiway Takemura estimator.
The left polar decomposition.
QR Decomposition.
Generate a list of orthogonal matrices drawn from Haar distribution.
Sample from the mirror-Wishart distribution.
Multivariate normal simulation.
Standard normal array.
Wishart simulation.
Gibbs update of Phi_inv.
Update for total variation parameter in equi_mcmc.
Get list of identity matrices.
Sample covariance matrices for each mode.
tensr: A package for Kronecker structured covariance inference.
Top K elements of a vector.
Trace of a matrix.
Truncates small numbers to 0.
Tucker sum.
Normal scores.
A collection of functions for Kronecker structured covariance estimation and testing under the array normal model. For estimation, maximum likelihood and Bayesian equivariant estimation procedures are implemented. For testing, a likelihood ratio testing procedure is available. This package also contains additional functions for manipulating and decomposing tensor data sets. This work was partially supported by NSF grant DMS-1505136. Details of the methods are described in Gerard and Hoff (2015) <doi:10.1016/j.jmva.2015.01.020> and Gerard and Hoff (2016) <doi:10.1016/j.laa.2016.04.033>.