R/searchZ_sparse.R

In CorReg: Linear Regression Based on Linear Structure Between Variables

# ' Sparse structure research
# ' @param X the dataset
# ' @param Zi indices of the rows of the 1
# ' @param Zj indices of the columns of the 1
# ' @param Si rowSums vector
# ' @param Sj colSums vector
# ' @param Bic_null_vect the BIC of the null hypothesis (used for independent variables)
# ' @param candidates 0: row and column,-1: column only, int>0: random int candidates, -2: all (but the diag), -3: non-zeros
# ' @param methode  parameter for OLS (matrix inversion) 1:householderQr, 2:colPivHouseholderQr
# ' @param p1max maximum complexity for a regression
# ' @param Maxiter number of steps
# ' @param plot TRUE: returns for each step the type of move, complexity and BIC
# ' @param best TRUE: systematically jumps to the best BIC seen ever when seen (it is stored even if best=FALSE)
# ' @param better TRUE: systematically jumps to the best candidate if better than stationnarity (random wheighted jump otherwise)
# ' @param random if FALSE:moves only to improve and only to the best
# ' @param verbose 0:none, 1:BIC,step and complexity when best BIC found 2:BIC, step, complexity, nb candidates and best candidate when best BIC found
# ' @param nb_opt_max stop criterion defining how many times the chain can walk (or stay) on the max found
# ' @param Mixmod
# ' @export
# '@return step 0:delete, 1: add, 2: stationnarity
# '
searchZ_sparse <- function(X = X, Zi = NULL, Zj = NULL, Si = NULL, Sj = NULL, Bic_null_vect = NULL, candidates = 2, methode = 1, p1max = 5, 
                           Maxiter = 1, plot = FALSE, best = TRUE, better = FALSE, random = TRUE, verbose = 1, nb_opt_max = NULL) {
  X = 1 * as.matrix(X)
  if (is.null(nb_opt_max)) {
    nb_opt_max = Maxiter
  }
  if (is.null(Bic_null_vect)) {
    Bic_null_vect = density_estimation(X = X)$BIC_vect
  }
  if (is.null(Zi) | is.null(Zj)) {
    Zi = as.vector(0)
    Zj = as.vector(0)
    p = ncol(X)
    Si = rep(0, times = p)
    Sj = rep(0, times = p)
  }
  if (is.null(Si) | is.null(Sj)) {
    p = ncol(X)
    Si = rep(0, times = p)
    Sj = rep(0, times = p)
    for (i in 1:p) {
      Si[i] = length(which(Zi == i))
      Sj[i] = length(which(Zj == i))
    }
  }
  res = .Call("rechercheZ_sparse_relax", X, Zi, Zj, Si, Sj, Bic_null_vect, candidates, methode, p1max, Maxiter, plot, best, better, random, verbose, nb_opt_max, PACKAGE = "CorReg")
  return(res)
}