mcem_mult: Estimate mean and Variance by MCEM multivariate

In JoaoPedro2536/univ.em:

Description

This is the function of M-step, to iterate over the results for mean and variance.

Usage

mcem_mult(data, mu_init, sigma_init, maxit = 1000, tol1 = 0.001, tol2 = 0.001)

Arguments

data	Data in the form of multivariated grouped data.
mu_init	The initialized mean vector.
sigma_init	The initialize covariance matrix.
maxit	Integer, number of max iterations .
tol1	The stopping criterias for means vectors.
tol2	The stopping criterias for covariance matrix.

Value

Returns a list with the mu estimate and the variance matrix.

Examples

library(MASS)
library(em.univ)
library(tmvtnorm)

simulateddata = em.univ::mult_simul(mm = c(68,68),
                          ss = base::matrix(c(3,2,2,6),2,2) ,
                          n_data_sets = 1,
                          breaks_x = c(-Inf,64,65,66,67,68,69,70,71,72,Inf),
                          breaks_y = c(-Inf,64.2,65.2,66.2,
                                       67.2,68.2,69.2,70.2,
                                       71.2,72.2,Inf),
                          lower_x = base::rep(c(-Inf,64,65,66,67,
                                          68,69,70,71,72),10),
                          lower_y = c(base::rep(-Inf,10),
                                      base::rep(64.2,10),
                                      base::rep(65.2,10),
                                      base::rep(66.2,10),
                                      base::rep(67.2,10),
                                      base::rep(68.2,10),
                                      base::rep(69.2,10),
                                      base::rep(70.2,10),
                                      base::rep(71.2,10),
                                      base::rep(72.2,10)),
                          upper_x = base::rep(c(64,65,66,67,68,69,70,71,72,Inf),10),
                          upper_y = c(base::rep(64.2,10),
                                      base::rep(65.2,10),
                                      base::rep(66.2,10),
                                      base::rep(67.2,10),
                                      base::rep(68.2,10),
                                      base::rep(69.2,10),
                                      base::rep(70.2,10),
                                      base::rep(71.2,10),
                                      base::rep(72,2,10),
                                      base::rep(Inf,10))
)

mu2<- c(67,67)
sigma2<- base::matrix(c(3.1,2.2,2.2,4.3),2,2)


out<- em.univ::mcem_mult(data=simulateddata[,,1],
              mu_init=mu2,
              sigma_init=sigma2,
              maxit=1000,
              tol1=1e-4,
              tol2=1e-3)

out

JoaoPedro2536/univ.em documentation built on Dec. 18, 2021, 1:38 a.m.