Usage
fillmiss(X = X, Z = NULL, mixmod = FALSE, B = NULL, Bt = NULL,
res_mixmod = NULL, nbclustmax = 10, X1 = FALSE)
Arguments
X
the dataset (matrix) with missing values
Z
the structure associated to X. Can be a matrix of zeros if non structure.
mixmod
boolean to say if the function has to use mixture model hypothesis or just the observed mean.
B
the matrix of the coefficients for sub-regressions
Bt
the matrix used for X1 if X1=TRUE
res_mixmod
the best results found by mixmod if already computed
nbclustmax
the max number of cluster for mixmod.
X1
boolean to say if dependent variables on the right will be filled based on the structure