ctglm.model: ctglm model definition

In ctglm: ctglm

Description

define ctglm model

Usage

ctglm.model( d, id="id", time="time", person.var=c("b"=TRUE,"mu.t1"=TRUE),
track.person.par=NULL, Lambda=NULL, measurement.model=binomial(link="logit"),
engine=c("jags","ctstan","ctsem"), ..., priors=NULL, verbose=TRUE )

Arguments

d	data frame
id	name of id column in data frame
time	name of time column in data frame
person.var	parameters that vary over persons
track.person.par	track person parameters (in jags/ctstan), may very much slow down run time, possible values: c("bj","mu.t1.j","theta"), "theta" only implemented in jags right now
Lambda	loading matrix of items and latent variables
measurement.model	measurement model
engine	software used for estimation
...	additional parameters, see details
priors	priors for jags
verbose	output messages to console

Details

additionally parameters: – delete.persons without varying responses (IRT) from data set, options: "persons_without_response_variance_at_at_least_one_timepoint" "persons_without_response_variance_at_all_timepoints" – novar.out.dir: directory where "no variance" descriptives (as list in Rdata file) is outputted, path is <novar.out.dir>/<model.name>_novar.Rdata – model.name: name of model

Value

returns a ctglm model list

Warning

This version is alpha. Use with care.

Author(s)

Martin Hecht

ctglm documentation built on May 31, 2017, 1:54 a.m.