LMest (version 3.1.2)

draw_lm_cov_latent: Draw samples from LM model with covariaates in the latent model

Description

Function that draws samples from the LM model with individual covariates with specific parameters.

The function is no longer maintained.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

Usage

draw_lm_cov_latent(X1, X2, param = "multilogit", Psi, Be, Ga, fort = TRUE)

Value

Y

matrix of response configurations unit by unit (n x TT x r)

U

matrix containing the sequence of latent states (n x TT)

Arguments

X1

desing matrix for the covariates on the initial probabilities (n x nc1)

X2

desing matrix for the covariates on the transition probabilities (n x TT-1 x nc2)

param

type of parametrization for the transition probabilities ("multilogit" = standard multinomial logit for every row of the transition matrix, "difflogit" = multinomial logit based on the difference between two sets of parameters)

Psi

array of conditional response probabilities (mb x k x r)

Be

parameters affecting the logit for the initial probabilities

Ga

parametes affecting the logit for the transition probabilities

fort

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

Author

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

Examples

if (FALSE) {

# draw a sample for 1000 units, 10 response variable and 2 covariates
n <- 1000
TT <- 5
k <- 2
nc <- 2 #number of covariates
r <- 10 #number of response variables
mb <- 2 #maximum number of response categories
fort <- TRUE


Psi <- matrix(c(0.9,0.1,0.1,0.9), mb, k)
Psi <- array(Psi, c(mb, k, r))
Ga <- matrix(c(-log(0.9/0.1),0.5,1), (nc+1)*(k-1), k)
Be <- array(c(0,0.5,1), (nc+1)*(k-1))
#Simulate covariates
X1 <- matrix(0, n, nc)
for(j in 1:nc) X1[,j] <- rnorm(n)
X2 <- array(0,c(n, TT-1, nc))
for (t in 1:(TT-1)) for(j in 1:nc){
	if(t==1){
		X2[,t,j] <- 0.5*X1[,j] + rnorm(n)
	}else{
		X2[,t,j] <- 0.5 *X2[,t-1,j] + rnorm(n)
	}
}

out <- draw_lm_cov_latent(X1, X2, Psi = Psi, Be = Be, Ga = Ga, fort = fort)
}