R/groups.R

In seagull: Lasso, Group Lasso, and Sparse-Group Lasso for Mixed Models

#' @title Group indices for explanatory variables
#' 
#' @description A vector of integers which assigns each variable (genotype
#' marker) to a particular group. The clustering was performed via the R package
#' \href{http://www.math-evry.cnrs.fr/publications/logiciels}{BALD}. This
#' package uses linkage disequilibrium as a measure of proximity. In total, 98
#' groups are available. Group sizes vary from 1 to 23. The median of group
#' sizes is equal to 3. For more details about the distribution of group sizes,
#' please check out the example below.
#' 
#' @name groups
#' 
#' @docType data
#' 
#' @author Jan Klosa \email{klosa@fbn-dummerstorf.de}
#' 
#' @format A vector consisting of 466 integer values.
#' 
#' @seealso \code{\link{plot}}
#' 
#' @examples
#' data(seagull_data)
#' 
#' ## Create a vector with group sizes:
#' sizes <- rep(as.integer(NA), max(groups))
#' for (i in 1:98) {
#'   sizes[i] = max(which(groups==i)) - min(which(groups==i)) + 1
#' }
#' 
#' ## Plot the sorted group sizes:
#' plot(x = seq(1, max(groups), 1), y = sort(sizes))
"groups"