LMest (version 3.1.2)

est_mc_cov: Estimate Markov chain (MC) model with covariates

Description

Main function for estimating the MC model with covariates.

The function is no longer maintained. Please look at lmestMc function.

Usage

est_mc_cov(S, X1 = NULL, X2 = NULL, yv = rep(1,nrow(S)), start = 0, tol = 10^-8,
	   maxit = 1000, out_se = FALSE, output = FALSE, fort = TRUE)

Value

lk

maximum log-likelihood

Be

estimated array of the parameters affecting the logit for the initial probabilities

Ga

estimated array of the parameters affecting the logit for the transition probabilities

np

number of free parameters

aic

value of AIC for model selection

bic

value of BIC for model selection

seBe

standard errors for Be

seGa

standard errors for Ga

Piv

estimate of initial probability matrix

PI

estimate of transition probability matrices

call

command used to call the function

Arguments

S

matrix of available configurations of the response variable (n x TT) with categories starting from 0

X1

matrix of covariates affecting the initial probabilities (n x nc1)

X2

array of covariates affecting the transition probabilities (n x TT-1 x nc2)

yv

vector of frequencies of the available configurations

start

type of starting values (0 = deterministic, 1 = random)

tol

tolerance level for checking convergence of the algorithm

maxit

maximum number of iterations of the algorithm

out_se

to compute the information matrix and standard errors

output

to return additional output (PI,Piv)

fort

to use fortran routine when possible (FALSE for not use fortran)

Author

Francesco Bartolucci, Silvia Pandolfi, University of Perugia, http://www.stat.unipg.it/bartolucci

References

Bartolucci, F., Farcomeni, A. and Pennoni, F. (2013) Latent Markov Models for Longitudinal Data, Chapman and Hall/CRC press.

Examples

if (FALSE) {

# Example based on criminal data

# load criminal data
data(data_criminal_sim)

#We consider the response variable referring of crime of type 5

out <- long2wide(data_criminal_sim, "id", "time", "sex",
"y5", aggr = T, full = 999)
XX <- out$XX-1
YY <- out$YY
freq <- out$freq
TT <- 6

X1 <- as.matrix(XX[,1])
X2 <- as.matrix(XX[,2:TT])
# estimate the model
res <- est_mc_cov(S = YY, yv = freq, X1 = X1, X2 = X2, output = TRUE)
summary(res)

# Initial probability for female
Piv0 <- round(colMeans(res$Piv[X1 == 0,]), 4)

# Initial probability for male
Piv1 <- round(colMeans(res$Piv[X1 == 1,]), 4)

}