estdist: Estimation of population distribution via EM algorithm.

In takuizum/irtfun2: Some useful functions for IRT

Description

This function is new version of MML_EM_dist

Usage

estdist(
  x,
  para,
  N = 31L,
  eMLL = 1e-06,
  eDIST = 1e-04,
  fc0 = 2L,
  ng = 1L,
  gc0 = 2L,
  D = 1.702,
  fix = 1L,
  print = 0L,
  max = 4,
  min = -4,
  maxiter_em = 200L,
  rm_list = as.character(c("NONE"))
)

Arguments

x	Item response matrix.
para	item parameter data.frame estimated by estip
N	the number of nodes in integration.
eMLL	a convergence criteria(CC) for marginal log likelihood.
eDIST	a CC for population distribution.
fc0	a column of first item response.
ng	the number of groups
gc0	a column of group. the element must be integer and the minimum number must be 1.
D	factor constant.
fix	Don't use. If 1, fix population distribution mean and sigma each EM cycle.
print	How much information you want to display? from 1 to 3. The larger, more information is displayed.
max	maximum value of theta in integration.
min	minimum value of theta in integration.
maxiter_em	maximum iteration time for EM cycle.
rm_list	a vector of item U want to remove for estimation. NOT list.

takuizum/irtfun2 documentation built on May 10, 2020, 8:30 a.m.