CovRegRF

Covariance Regression with Random Forests ('CovRegRF') is a random forest method for estimating the covariance matrix of a multivariate response given a set of covariates. Random forest trees are built with a new splitting rule which is designed to maximize the distance between the sample covariance matrix estimates of the child nodes. The method is described in Alakus et al. (2023) doi:10.1186/s12859-023-05377-y. 'CovRegRF' uses 'randomForestSRC' package (Ishwaran and Kogalur, 2022) by freezing at the version 3.1.0. The custom splitting rule feature is utilised to apply the proposed splitting rule. 'LAPACK' and 'BLAS' libraries are used for matrix decompositions. The 'CovRegRF' package includes the header files 'lapacke.h' and 'cblas.h' from the 'LAPACK' and 'BLAS' libraries. The 'LAPACK' library is licensed under modified BSD license.