R/ghg.R

In ellessenne/mla: An Algorithm for Least-squares Curve Fitting

#' @keywords internal
ghg <- function(m, v, fu) {
  ghg <- 0
  inf <- NULL
  sup <- NULL
  for (i in 1:m) {
    for (j in 1:m) {
      if (j >= i) {
        ij <- (j - 1) * j / 2 + i
        inf <- c(inf, ij)
      } else {
        ij <- (i - 1) * i / 2 + j
        sup <- c(sup, ij)
      }
      ghg <- ghg + v[m * (m + 1) / 2 + i] * fu[ij] * v[m * (m + 1) / 2 + j]
    }
  }
  return(list(ghg = ghg, inf = inf, sup = sup))
}