density_estimation: BIC of estimated marginal gaussian mixture densities

In CorReg: Linear Regression Based on Linear Structure Between Variables

Description

Estimates the density of each covariates with gaussian mixture models and then gives the associated BIC.

Usage

density_estimation(
  X = X,
  nbclustmax = 10,
  nbclustmin = 1,
  verbose = FALSE,
  detailed = FALSE,
  max = TRUE,
  package = c("mclust", "Rmixmod"),
  nbini = 20,
  matshape = FALSE,
  ...
)

Arguments

X	the dataset (matrix)
nbclustmax	max number of clusters in the gaussian mixtures
nbclustmin	min number of clusters in the gaussian mixtures
verbose	verbose or not
detailed	boolean to give the details of the mixtures found
max	boolean. Use an heuristic to shrink nbclustmax according to the number of individuals in the dataset
package	package to use ("Rmixmod", "mclust")
nbini	number of initial points for Rmixmod
matshape	boolean to give the detail in matricial shape
...	additional parameters

Value

a list that contains:

BIC_vect	vector of the BIC (one per variable)
BIC	global value of the BIC (=sum(BIC_vect))
nbclust	vector of the numbers of components
details	list of matrices that describe each Gaussian Mixture (proportions, means and variances)

Examples

# dataset generation
base = mixture_generator(n = 150, p = 10, valid = 0, ratio = 0.4, tp1 = 1, tp2 = 1, tp3 = 1, 
                         positive = 0.5, R2Y = 0.8, R2 = 0.9, scale = TRUE, max_compl = 3, 
                         lambda = 1)
X_appr = base$X_appr # learning sample
density = density_estimation(X = X_appr, detailed = TRUE) # estimation of the marginal densities
density$BIC_vect # vector of the BIC (one per variable)
density$BIC # global value of the BIC (sum of the BICs)
density$nbclust # vector of the numbers of components.
density$details # matrices that describe each Gaussian Mixture (proportions, means and variances)

CorReg documentation built on Feb. 20, 2020, 5:07 p.m.