R/lin1_theta.R

In ChrisKust/rexpar: Simplicial Depth for Explosive Autoregressive Processes

lin1_theta <- function(dat)
{
  dat1 <- dat[1:(length(dat) - 1)]
  dat2 <- dat[2:length(dat)]
  t1 <- 0
  cand_t1 <- numeric(length(dat1))
  for(i in 1:(length(dat1)))
  {
    cand_t1[i] <- (dat2[i] - dat1[i]) / (dat1[i])#-(dat2[i]-dat1[i])/(dat1[i]))/2
  }
  cand_t1S <- sort(cand_t1)
  for(i in 1:(length(cand_t1S) - 1))
  {
    for(j in (i + 1):length(cand_t1S))
    {
      t1 <- c(t1,(cand_t1S[j] + cand_t1S[i]) / 2)
    }
  }
  t1 <- c(t1) + 1
  list(t1 = t1)
}