extract: extract elements from objects of various classes

Description

A generic function to extract elements from objects of class GDINA, itemfit, modelcomp, Qval or simGDINA. This page gives the elements that can be extracted from the class GDINA. To see what can be extracted from itemfit, modelcomp, and Qval, go to the corresponding function help page. Objects which can be extracted from GDINA objects include:

AIC: AIC
att.prior: attribute prior weights for calculating marginalized likelihood in the last iteration
att.str: argument att.str
BIC: BIC
call: function call
catprob.parm: category success probability for each latent group; the same as itemprob.parm for dichotomous response items.
catprob.se: SE associated with the category success probability for each latent group.
catprob.cov: variance-covariance matrix of item endorsement probabilities for all items
conv.crit: argument conv.crit
dat: item responses analyzed
delta.parm: delta parameters for each category
delta.cov: Convariance matrix associated with the delta parameters.
delta.se: SE associated with the delta parameters for each latent group.
deviance: deviance: -2 times observed log-likelihood value
dif.LL: absolute change in deviance in the last EM iteration
dif.p: max absolute change in success probabilities in the last EM iteration
digits: argument digits
discrim: GDINA discrimination index
empirical: argument empirical
end.time: end time
expectedCorrect: expected # of examinees in each latent group answering item correctly
expectedTotal: expected # of examinees in each latent group
higher.order: higher-order model specifications
higher.order.method: argument higher.order$method
higher.order.model: argument higher.order$model
HO.parm.history: HO.parm.history in diagnosis mode
initial.catprob: initial item category probability parameters
iter.history: iter.history in diagnosis mode
item.names: argument item.names
itemprob.history: itemprob.history in diagnosis mode
itemprob.parm: item success probability for each latent group
itemprob.se: SE associated with the item success probability for each latent group.
LCprob.parm: category success probability for each latent class
logLik: observed log-likelihood value
loglikelihood.i: log-likelihood for each examinee
likepost.history: likepost.history in diagnosis mode
logposterior.i: log-posteriori for each examinee
maxitr: argument maxitr
models: fitted CDMs for each item/category
mono.constraint: argument mono.constraint
natt: number of attributes
ncat: number of categories excluding category zero
ngroup: number of groups
nitem: number of items
nitr: number of iterations
nobs: number of individuals
npar: number of parameters
npar.item: number of item parameters
npar.att: number of attribute parameters
nstarts: argument nstarts
prevalence: prevalence of each attribute
posterior.prob: posterior weights for each latent class
Q: Q-matrix
Qc: Qc-matrix
RN.history: RN.history in diagnosis mode
start.time: starting time
sequential: argument sequential
seq.dat: data for sequential models
time: time used
verbose: argument verbose

Usage

extract(object, what, ...)

Arguments

object

objects from class GDINA,itemfit, modelcomp, Qval or simGDINA

what

what to extract

...

additional arguments