R/array2longitudinal.r

Defines functions array2longitudinal

Documented in array2longitudinal

array2longitudinal <- function(Y, 
                               keepMissings=TRUE){

	########################################################################
	#
	# DESCRIPTION:
	# Converts a 3-dim array (containing time-series data of multiple 
	# individuals) to an object of which can directly by converted to 
	# the 'longitudinal' class.
	#
	# ARGUMENTS:
	# -> Y			: Three-dimensional array containing the 
	#                         data. The first, second and third dimensions 
	#                         correspond to covariates, time and 
	#			  samples, respectively.
	# -> keepMissings	: The 'array'-format assumes a balanced 
	#                         layout of the time-course experiment. The 
	#                         experiment may have failed for some design 
	#                         points and no data is available. In the 
	#                         'longitudinal'-format these design points 
	#                         may be left out. This 'logical'
	# 		          indicates whether they should be kept 
	#                         in (or not).
	#
	#
	# NOTES:
	# ...
	#
	########################################################################

	# make rownames for longitudinal object
	rNames <- paste(sort(rep(1:dim(Y)[2], dim(Y)[3])), 
			rep(1:dim(Y)[3], dim(Y)[2]), sep="-")

	# fill matrix with longitudinal data
	Ymatrix <- matrix(NA, ncol=dim(Y)[1], nrow=dim(Y)[2] * dim(Y)[3])
	for (t in 1:dim(Y)[2]){
		Ymatrix[(t-1)*dim(Y)[3] + 1:dim(Y)[3], ] <- t(Y[,t,])
	}
	
	# assign rownames and attributes
	rownames(Ymatrix) <- rNames
	attr(Ymatrix, "time") <- 1:dim(Y)[2]
    	attr(Ymatrix, "repeats") <- rep(dim(Y)[3], dim(Y)[2])

	# return object
	return(Ymatrix)
}

Try the ragt2ridges package in your browser

Any scripts or data that you put into this service are public.

ragt2ridges documentation built on Nov. 21, 2017, 5:06 p.m.