R/design.R

In mihaiconstantin/netTinker: Simulations for the master's thesis

#' @title Build the factorial design of the study.
#'
#' @description
#' This function builds the factorial design of the study based on the elements of the
#' vectors specified as arguments. Each argument specified (i.e., vector) is regarded
#' as a factor, and each value of the vectors is regarded as a level.
#'
#
#' @usage build_design(participants, nodes, connectedness, architectures)
#'
#' @param participants (int vector) Varying number of participants, e.g., \code{c(10, 50, 100)}.
#' @param nodes        (int vector) Varying number of nodes (i.e., variables), e.g., \code{c(10, 15, 30)}.
#' @param connectedness    (double vector) Varying number of connectivity, e.g., \code{c(1, 2, 3)}.
#' @param architecture (int vector) Network architectures, e.g., \code{c(random = 1, small_world = 2, scale_free = 3)}.
#' @param models (int vector) Network models, e.g., \code{c(ising = 1, ggm = 2)}.
#'
#' @return A matrix representing the factorial design of the study, where each row is a design cell.
#'
#' @export
#'
build_design <- function(participants, nodes, architectures, connectedness, models, labels = c('participants', 'nodes', 'architectures', 'connectedness', 'models'))
{
	# options(warn=-1)
	
	# The design matrix.
	design = as.data.frame(matrix(NA, 0, 5))

	for (a in 1:length(participants))
	{
		for (b in 1:length(nodes))
		{
			for(c in 1:length(architectures))
			{
				for (d in 1:length(connectedness))
				{
					for (e in 1:length(models)) {
    	            	design = rbind(design, c(participants[a], nodes[b], architectures[c], connectedness[d], models[e]))
					}
				}
			}
		}
	}
	colnames(design) <- labels

	# User feedback.
	cat(
		'-> Design generated:',
		nrow(design), 'cells',
		'(each row in the matrix stands for one design cell).',
		'\n\n'
	)

	return(design)
}