R/crit_statisWj_fast.R

In ClustBlock: Clustering of Datasets

.crit_statisWj_fast=function(Wj, index, RVtot){
  #Wj list

  n=nrow(Wj[[1]])
  nblo=length(Wj)

  RV=RVtot[index, index]

  if (length(index)>1)
  {
    ressvd=svd(RV)
    lambda=ressvd$d[1]
  }else{
    lambda=1
  }


  Q=nblo-lambda

  return(Q)

}