#  File tests/testthat/test-gflomiss.R in package ergm, part of the
#  Statnet suite of packages for network analysis, https://statnet.org .
#
#  This software is distributed under the GPL-3 license.  It is free,
#  open source, and has the attribution requirements (GPL Section 7) at
#  https://statnet.org/attribution .
#
#  Copyright 2003-2024 Statnet Commons
################################################################################

data(florentine)

run.test <- function() {
  ##
  ## Create random 5% missing
  ##
  #mflomarriage <- rergm(network.size(flomarriage),prob=0.05, directed=FALSE)
  #flomarriage <- set.graph.attribute(flomarriage, "design", mflomarriage)
  #summary(flomarriage)

  #
  # Create random 2 nodes who are non-respondents
  #
  #respondent <- rmultinom(n=1, size=network.size(flomarriage)-2,
  #                        prob=rep(1,network.size(flomarriage)))
  #respondent

  respondent <- rep(FALSE,network.size(flomarriage))
  respondent[sample(1:network.size(flomarriage), size=network.size(flomarriage)-2,replace=FALSE)] <- TRUE
  respondent

  #
  #one <- matrix(1,ncol=1,nrow=network.size(flomarriage))
  #orespondent <- one %*% t(respondent) + respondent %*% t(one) - respondent %*% t(respondent)
  #orespondent <- 1-orespondent
  #diag(orespondent) <- 0
  ##
  #mflomarriage <- graph(orespondent, directed=FALSE)
  #summary(mflomarriage)
  #sociomatrix(mflomarriage)
  #flomarriage <- set.graph.attribute(flomarriage, "design", mflomarriage)

  #efit <- ergm(flomarriage ~ edges + kstar(2), control=control.ergm(MCMC.samplesize=1000, MCMC.interval=1000))
  efit <- ergm(flomarriage ~ edges + kstar(2), estimate="MPLE")
  summary(efit)

  flomarriage <- set.vertex.attribute(flomarriage, "respondent", respondent)
  rm(respondent)
  summary(flomarriage)

  efit <- ergm(flomarriage ~ edges + kstar(2), estimate="MPLE")
  summary(efit)

  efit <- ergm(flomarriage ~ edges + kstar(2), control=control.ergm(MCMC.samplesize=1000, MCMC.interval=1000))
  summary(efit)

  efit <- ergm(flomarriage ~ edges + kstar(2), control=control.ergm(init=c(-1.6,0)))

  #
  # edges  -1.6     -1.74242 0.8557   0.044   0.041373
  # kstar2  0.0      0.02746 0.1774   0.877   0.009076
  #
  summary(efit)
}

test_that("undirected network with missing data and dyadic dependence", {
  expect_error(run.test(), NA)
})