log_lik_2pl: log_lik_2pl Function

In nmolanog/bayesuirt:

Description

log likelihood function for 2pl

Usage

log_lik_2pl(y, X, psi, z, theta, nitems)

Arguments

y	vectorised test
X	design matrix indexing each observation to corresponding item
psi	vector of item parameters
z	design matrix indexing each observation to corresponding individual
theta	vector of latent traits (i.e. random effects associated to individual level)
nitems	integer especifying the number of items. It is used to separate parameters as psi[1:nitems] and psi[(nitems+1):(2*nitems)]

Value

a double atomic vector

Examples

data("test_data")
temp<-uirt_DM(test_data,"id")
nitems<-ncol(temp$X)
nind<-ncol(temp$Z)
alpha<-runif(nitems,0.6,1.5) ##simulation of discrimination parameters
beta<-runif(nitems,-2.4,2.4) ##simulation of dificulty parameters
theta<-rnorm(nind) ##simulation of random effects
psi<-c(-alpha*beta,log(alpha)) ###vector of parameters based on alpha and beta
res<-log_lik_2pl(temp$Y,temp$X,psi,temp$Z,theta,nitems) ###2pl log likelihood evaluated at given values

nmolanog/bayesuirt documentation built on May 23, 2019, 8:52 a.m.