mi.lori {lori}

The mi.lori performs M multiple imputations using the lori method. Multiple imputation allows to produce estimates of missing values, as well as intervals of variability. The classical procedure is to perform M multiple imputations using the mi.lori method, and to aggregate them using the pool.lori method.

Description

The mi.lori performs M multiple imputations using the lori method. Multiple imputation allows to produce estimates of missing values, as well as intervals of variability. The classical procedure is to perform M multiple imputations using the mi.lori method, and to aggregate them using the pool.lori method.

Usage

mi.lori(
  Y,
  cov = NULL,
  lambda1 = NULL,
  lambda2 = NULL,
  M = 25,
  intercept = T,
  reff = T,
  ceff = T,
  rank.max = 5,
  algo = c("alt", "mcgd"),
  thresh = 1e-05,
  maxit = 1000,
  trace.it = F
)

Arguments

Y

[matrix, data.frame] count table (nxp).

cov

[matrix, data.frame] design matrix (np*q) in order row1xcol1,row2xcol2,..,rownxcol1,row1xcol2,row2xcol2,...,...,rownxcolp

lambda1

[positive number] the regularization parameter for the interaction matrix.

lambda2

[positive number] the regularization parameter for the covariate effects.

M

[integer] the number of multiple imputations to perform

intercept

[boolean] whether an intercept should be fitted, default value is FALSE

reff

[boolean] whether row effects should be fitted, default value is TRUE

ceff

[boolean] whether column effects should be fitted, default value is TRUE

rank.max

[integer] maximum rank of interaction matrix (smaller than min(n-1,p-1))

algo

type of algorithm to use, either one of "mcgd" (mixed coordinate gradient descent, adapted to large dimensions) or "alt" (alternating minimization, adapted to small dimensions)

thresh

[positive number] convergence tolerance of algorithm, by default 1e-6.

maxit

[integer] maximum allowed number of iterations.

trace.it

[boolean] whether convergence information should be printed

Value

mi.imputed

a list of length M containing the imputed count tables

mi.alpha

a (Mxn) matrix containing in rows the estimated row effects (one row corresponds to one single imputation)

mi.beta

a (Mxp) matrix containing in rows the estimated column effects (one row corresponds to one single imputation)

mi.epsilon

a (Mxq) matrix containing in rows the estimated effects of covariates (one row corresponds to one single imputation)

mi.theta

a list of length M containing the estimated interaction matrices

mi.mu

a list of length M containing the estimated Poisson means

mi.y

list of bootstrapped count tables used fot multiple imputation

Y

original incomplete count table

Examples

X <- matrix(rnorm(50), 25)
Y <- matrix(rpois(25, 1:25), 5)
res <- mi.lori(Y, X, 10, 10, 2)