R/LADVAR.R

In ConnectednessApproach: Connectedness Approach

#' @title Least absolute deviation vector autoregression
#' @description Estimation of a LAD VAR using equation-by-equation LAD regressions.
#' @param x zoo data matrix
#' @param nlag Lag length
#' @param configuration model configuration
#' @return Estimate LAD VAR model
#' @examples
#' data("dy2012")
#' fit = LADVAR(dy2012, configuration=list(nlag=1))
#' @author David Gabauer
#' @importFrom L1pack lad
#' @importFrom stats lm
#' @importFrom stats embed
#' @export
LADVAR = function (x, configuration = list(nlag = 1)) {
  if (!is(x, "zoo")) {
    stop("Data needs to be of type 'zoo'")
  }
  k = ncol(x)
  NAMES = colnames(x)
  if (is.null(NAMES)) {
    NAMES = 1:k
  }
  nlag = as.numeric(configuration$nlag)
  if (nlag <= 0) {
    stop("nlag needs to be a positive integer")
  }
  Res = B = se = NULL
  for (i in 1:k) {
    z = stats::embed(x, nlag + 1)
    fit = summary(lad(z[, i] ~ z[, -c(1:k)]))
    B = rbind(B, fit$coefficients[-1, 1])
    se = rbind(se, fit$coefficients[-1, 2])
    Res = cbind(Res, fit$residuals)
  }
  Q = array(t(Res) %*% Res/nrow(Res), c(k, k, 1),
            dimnames = list(NAMES, NAMES, tail(as.character(zoo::index(x)), 1)))
  results = list(B=B, Q=Q, se=se)
}