tensorIA

Integrative analysis based on tensor modelling.

For a grouped multivariates gression model, with or without aparsity assumptions, treating the coefficients as a third-order tensor and borrowing Tucker decomposition to reduce the number of parameters.

Installation

#install Rtools 3.5 (http://cran.r-project.org/bin/windows/Rtools)
#install.packages("devtools")
#install.packages("Rcpp")
library(devtools)
install_github("xliusufe/tensorIA")

Usage

• tensorIA-manual ------------ Details of the usage of the package.

Example

library(tensorIA)

n <- 200
p <- 5
q <- 5
g <- 5
r10 <- 2
r20 <- 2
r30 <- 2
S3 <- matrix(runif(r10*r20*r30,3,7),nrow = r30)
T1 <- matrix(rnorm(p*r10),nrow = p)
U1 <- qr.Q(qr(T1))
T1 <- matrix(rnorm(g*r20),nrow = g)
U2 <- qr.Q(qr(T1))  
T1 <- matrix(rnorm(q*r30),nrow = q)
U3 <- qr.Q(qr(T1))
D3 <- U3%*%S3%*%t(kronecker(U2,U1))
X <- matrix(rnorm(n*p*g), nrow = n)
eps <- matrix(rnorm(n*q),n,q)
Y <- X%*%t(D3)  + eps

fit <- integ_dr(Y, X, g, intercept=FALSE)
D3hat <- fit$Dnew
D2hat <- TransferModalUnfoldings(D3hat,3,2,p,g,q)
opt <- fit$rk_opt	

References

Integrative analysis based on tensor modelling. Manuscript.

Development

The R-package is developed by Xu Liu (liu.xu@sufe.edu.cn).