% Generated by roxygen2: do not edit by hand % Please edit documentation in R/simulate.R \name{simulate_correlation} \alias{simulate_correlation} \title{Simulate n points of dimension p correlated with a reference matrix.} \usage{ simulate_correlation(reference, p, r2, equal.var = TRUE) } \arguments{ \item{reference}{a numeric matrix to which the simulated data will be correlated} \item{p}{an \code{int} value indicating the number of dimensions (variables) simulated} \item{r2}{the fraction of variation shared between the \code{reference} and the simulated data} \item{equal.var}{a \code{logical} value indicating if the dimensions must be scaled to force \code{sd=1}. \code{TRUE} by default.} } \value{ a numeric matrix of \code{nrow(reference)} rows and \code{p} columns } \description{ Simulates a set of point correlated to another set according to the procrustean correlation definition. Points are simulated by drawing values of each dimension from a normal distribution of mean 0 and standard deviation equals to 1. The mean of each dimension is forced to 0 (data are centred). By default variable are also scaled to enforce a strandard deviation strictly equal to 1. Covariances between dimensions are not controled. Therefore they are expected to be equal to 0 and reflect only the random distribution of the covariance between two random vectors. The intensity of the correlation is determined by the \code{r2} parameter. } \examples{ sim1 <- simulate_matrix(25,10) class(sim1) dim(sim1) sim2 <- simulate_correlation(sim1,20,0.8) corls(sim1, sim2)^2 } \author{ Eric Coissac Christelle Gonindard-Melodelima }