R/spatialContMat.R

In rhobis/FPdust: An Implementation of the FPDUST and PPS FPDUST Spatial Sampling Designs for Finite Populations

#' Create a spatial contingency matrix from a list of contingencies
#'
#' Create a matrix of local contingency matrix of size \code{NxN} starting from
#' a list.
#'
#'
#' @param clist list of length N in which each element is a vector of id's continguous units
#'
#'
#' @details
#' The input argument \code{clist} should be a list of lenght N. Each element of
#' this list must be a numeric vector, including the indices of the units that
#' are contiguous to the unit corresponding to the list element.
#' Function \code{spatialContingencyMatrix} returns a matrix whose values are 1's (contiguous units)
#' and 0's (non-contiguous units).
#'
#'
#' @examples
#' # List of contiguous units ---
#' cl <- list( c(2, 3), c(1,4), c(1,6), c(2,6), NULL, c(3,4))
#'
#' # Generate spatial contingency matrix ---
#' cm <- spatialContingencyMatrix(cl)
#'
#'
#' @export
#'


spatialContingencyMatrix <- function(clist){

    if(!is.list(clist)) stop('the argument should be a list of length N')
    N <- length(clist)

    out <- lapply(seq_len(N), function(i) {
        v <- rep(0, N)
        v[ clist[[i]] ] <- 1
        return(v)
    } )
    out <- Reduce(rbind, out)
    diag(out) <- 0

    if( !isSymmetric(out, check.attributes=FALSE) )
        warning("The spatial contingency matrix is not symmetric, please check your input list!")

    return(out)
}