R/tkmeans.R

In tclust: Robust Trimmed Clustering

.tkmeans.old <-
function (x, k = 3, alpha = 0.05, nstart = 50, iter.max = 20, center = 0, scale = 1,
          store.x = TRUE, drop.empty.clust = TRUE, trace = 0, warnings = 2, 
          zero.tol = 1e-16)
{
  ret <- tclust (x = x, k = k, restr.fact = 1, restr = "eigen", alpha = alpha, 
                  nstart = nstart, iter.max = iter.max, equal.weights = TRUE, 
                  store.x = store.x, center = center, scale = scale, 
                  drop.empty.clust = drop.empty.clust, trace = trace, 
                  zero.tol = zero.tol, warnings = min (warnings, 2))

  ret$obj = ret$cov[1,1,1]
  ret
}