R/valse.R

   1 #' Main function
   2 #'
   3 #' @param X matrix of covariates (of size n*p)
   4 #' @param Y matrix of responses (of size n*m)
   5 #' @param procedure among 'LassoMLE' or 'LassoRank'
   6 #' @param selecMod method to select a model among 'SlopeHeuristic', 'BIC', 'AIC'
   7 #' @param gamma integer for the power in the penaly, by default = 1
   8 #' @param mini integer, minimum number of iterations in the EM algorithm, by default = 10
   9 #' @param maxi integer, maximum number of iterations in the EM algorithm, by default = 100
  10 #' @param eps real, threshold to say the EM algorithm converges, by default = 1e-4
  11 #' @param kmin integer, minimum number of clusters, by default = 2
  12 #' @param kmax integer, maximum number of clusters, by default = 10
  13 #' @param rang.min integer, minimum rank in the low rank procedure, by default = 1
  14 #' @param rang.max integer, maximum rank in the
  15 #' @return a list with estimators of parameters
  16 #' @export
  17 #-----------------------------------------------------------------------
  18 valse = function(X,Y,procedure,selecMod,gamma = 1,mini = 10,
  19                  maxi = 100,eps = 1e-4,kmin = 2,kmax = 10,
  20                  rang.min = 1,rang.max = 10) {
  21   ##################################
  22   #core workflow: compute all models
  23   ##################################
  24
  25   p = dim(phiInit)[1]
  26   m = dim(phiInit)[2]
  27
  28   print("main loop: over all k and all lambda")
  29   for (k in kmin:kmax)
  30   {
  31     print(k)
  32
  33     print("Parameters initialization")
  34     #smallEM initializes parameters by k-means and regression model in each component,
  35     #doing this 20 times, and keeping the values maximizing the likelihood after 10
  36     #iterations of the EM algorithm.
  37     init = initSmallEM(k, X, Y)
  38     phiInit <<- init$phiInit
  39     rhoInit <<- init$rhoInit
  40     piInit  <<- init$piInit
  41     gamInit <<- init$gamInit
  42
  43     gridLambda <<- gridLambda(phiInit, rhoInit, piInit, tauInit, X, Y, gamma, mini, maxi, eps)
  44
  45     print("Compute relevant parameters")
  46     #select variables according to each regularization parameter
  47     #from the grid: A1 corresponding to selected variables, and
  48     #A2 corresponding to unselected variables.
  49     params = selectiontotale(phiInit,rhoInit,piInit,tauInit,
  50                              mini,maxi,gamma,gridLambda,
  51                              X,Y,thresh,eps)
  52     A1 <<- params$A1
  53     A2 <<- params$A2
  54     Rho <<- params$Rho
  55     Pi <<- params$Pi
  56
  57     if (procedure == 'LassoMLE') {
  58       print('run the procedure Lasso-MLE')
  59       #compute parameter estimations, with the Maximum Likelihood
  60       #Estimator, restricted on selected variables.
  61       model = constructionModelesLassoMLE(
  62         phiInit, rhoInit,piInit,tauInit,mini,maxi,
  63         gamma,gridLambda,X,Y,thresh,eps,A1,A2)
  64       ################################################
  65       ### Regarder la SUITE
  66       r1 = runProcedure1()
  67       Phi2 = Phi
  68       Rho2 = Rho
  69       Pi2 = Pi
  70
  71       if (is.null(dim(Phi2)))
  72         #test was: size(Phi2) == 0
  73       {
  74         Phi[, , 1:k] <<- r1$phi
  75         Rho[, , 1:k] <<- r1$rho
  76         Pi[1:k,] <<- r1$pi
  77       } else
  78       {
  79         Phi <<-
  80           array(0., dim = c(p, m, kmax, dim(Phi2)[4] + dim(r1$phi)[4]))
  81         Phi[, , 1:(dim(Phi2)[3]), 1:(dim(Phi2)[4])] <<- Phi2
  82         Phi[, , 1:k, dim(Phi2)[4] + 1] <<- r1$phi
  83         Rho <<-
  84           array(0., dim = c(m, m, kmax, dim(Rho2)[4] + dim(r1$rho)[4]))
  85         Rho[, , 1:(dim(Rho2)[3]), 1:(dim(Rho2)[4])] <<- Rho2
  86         Rho[, , 1:k, dim(Rho2)[4] + 1] <<- r1$rho
  87         Pi <<- array(0., dim = c(kmax, dim(Pi2)[2] + dim(r1$pi)[2]))
  88         Pi[1:nrow(Pi2), 1:ncol(Pi2)] <<- Pi2
  89         Pi[1:k, ncol(Pi2) + 1] <<- r1$pi
  90       }
  91     } else {
  92       print('run the procedure Lasso-Rank')
  93       #compute parameter estimations, with the Low Rank
  94       #Estimator, restricted on selected variables.
  95       model = constructionModelesLassoRank(Pi, Rho, mini, maxi, X, Y, eps,
  96                                            A1, rank.min, rank.max)
  97
  98       ################################################
  99       ### Regarder la SUITE
 100       phi = runProcedure2()$phi
 101       Phi2 = Phi
 102       if (dim(Phi2)[1] == 0)
 103       {
 104         Phi[, , 1:k,] <<- phi
 105       } else
 106       {
 107         Phi <<- array(0, dim = c(p, m, kmax, dim(Phi2)[4] + dim(phi)[4]))
 108         Phi[, , 1:(dim(Phi2)[3]), 1:(dim(Phi2)[4])] <<- Phi2
 109         Phi[, , 1:k,-(1:(dim(Phi2)[4]))] <<- phi
 110       }
 111     }
 112   }
 113   print('Model selection')
 114   if (selecMod == 'SlopeHeuristic') {
 115
 116   } else if (selecMod == 'BIC') {
 117
 118   } else if (selecMod == 'AIC') {
 119
 120   }
 121 }