README.md

In bingx1990/STRS: Self-tuning rank selection in multivariate response regression

Self-tuning Rank Selection In Multivariate Response Regression

This package implements a self-tuning procedure of selecting the rank of the regression coefficient matrix in multivariate response regression models. The procedure can be used to select the rank of the factor components under a factor model.

Example: Let Y and X be a n by m response matrix and a n by p feature design matrix generated from the model Y = XA + E. STRS(Y, X) returns the estimated rank of the coefficient matrix A via regressing Y on X.

The following commands install the STRS package in R (require the installation of devtools package). library(devtools) install_github("bingx1990/STRS") Example: library(STRS) est_rank <- STRS(Y, X)

bingx1990/STRS documentation built on Jan. 22, 2022, 8:46 a.m.