R/validate_array.R

In AlessandroDeCarlo27/mvlognCorrEst: Sampling from a multivariate Log-Normal distribution using Log-Normal parameters and Estimation of Indirect Correlations

#' @title Validation of vectors of statistical parameters (mean, sd, cv)
#' @description This function is used to validate vectors of statistical parameters (mean, sd, cv) which are typically
#' the input of the functions in this package. They must be array objects without \code{NA}, \code{NaN} or infinite values and
#' their length must be equal to the size of the associated covariance matrix/correlation matrix (square matrix).
#' If these requisites are not respected, execution will be stopped.
#' @param name_arr string with name of the object that represents input array to validate
#' @param arr object that represents input array to validate
#' @param matRef matrix object that represents correlation matrix/covariance matrix
#' @param name_mat string with name of the object that represents correlation matrix/covariance matrix
#' @author Alessandro De Carlo \email{alessandro.decarlo01@@universitadipavia.it}
#' @noRd


validate_array <- function(name_arr,arr,matRef,name_mat){
    if(!is.array(arr)){
        stop(paste(name_arr,"must be an array object",sep=" "))
    }

    if(length(arr)!=ncol(matRef)|length(arr)!=nrow(matRef)){
        stop(paste("Wrong dimensions: number of elements in", name_arr,"must match the dimensions of", name_mat,".",sep=" "))
    }

    if(any(is.na(arr))||any(is.nan(arr))||any(is.infinite(arr))){
        stop(paste(name_arr,"contains illegal values (NA,NaN Inf)."))
    }


}