R/lowD.R

In TriangularCell/PgaMsgl: Proximal Gradient Algorithm for Multi-variate Sparse Group Lasso

#' Simulated example data on relative low dimension
#'
#' This is a list includes simulated example data on relative low dimension for \code{\link{PgaMsgl}} testing.
#' @name lowD
#' @docType data
#' @format A list with 16 components:
#' \describe{
#' \item{\strong{\code{X}}}{Simulated covariate matrix \code{X} in the model \code{Y=XB} with dimension 25x50, \emph{i.e.} \code{X} is composed with values of 50 variates in 25 samples. Elements in \code{X} are independently standard normal distributed.}
#' 
#' \item{\strong{\code{Beta0}}}{Simulated coefficient matrix \code{B} in the model \code{Y=XB} with dimension 50x50. \code{B} is with 10x10 groups tiling it, and each group is a 5×5 module of \code{B}. 10 out of 100 groups were randomly selected, each with 12 elements randomly selected to be with a random value in [-1, 1].}
#' 
#' \item{\strong{\code{Y}}}{Simulated response matrix \code{Y} in the model \code{Y=XB} with dimension 25x50, i.e. \code{Y} is composed of values of 50 responses in 25 samples. It is generated by multiply \code{X} with \code{Beta0}.}
#' 
#' \item{\strong{\code{Gm}}}{Matrix of the group structure of coefficient matrix \code{Beta0}. It is a 100x4 matrix with each row indicating a group, four columns indicate the row-start, row-end, column-start and column-end of the group. The row/column index is 1-based.}
#' 
#' \item{\strong{\code{mi}}}{Maximum number of iteration, which is 1000.}
#' 
#' \item{\strong{\code{mg}}}{Maximum number of groups in matrix \code{B} to be reserved, which is 10.}
#' 
#' \item{\strong{\code{mc}}}{Maximum number of single coefficients in matrix \code{B} to be reserved, which is 120.}
#' 
#' \item{\strong{\code{B0}}}{Simulated initial matrix of \code{B}, which is composed of all zeros.}
#' 
#' \item{\strong{\code{B1}}}{Simulated initial matrix of \code{B}, which is composed of all ones.}
#'  
#' \item{\strong{\code{B2}}}{Simulated initial matrix of \code{B}, which is generated adding random noise (rnorm(1,mean=1,sd=1)) to the real solution, Beta0.}
#' 
#' \item{\strong{\code{B3}}}{Simulated initial matrix of \code{B}, which is composed of all random noises (rnorm(1,mean=1,sd=1)).}
#' 
#' \item{\strong{\code{B4}}}{Simulated initial matrix of \code{B}, which is to have all non-zero elements in Beta0 as 1.}
#' 
#' \item{\strong{\code{B5}}}{Simulated initial matrix of \code{B}, which is to have non-zero values in Beta0 has 80\% probability to be 1, and zero values has 20\% probability to be 1.}
#' 
#' \item{\strong{\code{B6}}}{Simulated initial matrix of \code{B}, which is to have non-zero values in Beta0 has 80\% probability to be 1, and zero values to be 0.}
#'
#' \item{\strong{\code{B7_01}}}{Simulated initial matrix of \code{B}. Similar to \code{B7_001}, but the Benjamini-Hochberg adjusted P-value cutoff was 0.1.}
#' 
#' \item{\strong{\code{B7_01_coeff}}}{Simulated initial matrix of \code{B}. Similar to \code{B7_001_coeff}, but the Benjamini-Hochberg adjusted P-value cutoff was 0.1.}
#' }
#' 
#' @author Yiming Qin
#' @keywords data
#' @examples 
#' data(lowD)
#' system.time(lowD_result <- PgaMsgl(lowD$X, lowD$Y, lowD$B0, model="L121", lowD$Gm, lowD$mi, lowD$mg, lowD$mc))
#' 
#' 
NULL