context("Qsr output")

skip_on_cran()

library(survey)
library(laeken)
library(vardpoor)
data(eusilc)
names(eusilc) <- tolower(names(eusilc))
dati = data.frame(IDd = seq(10000 , 10000 + nrow(eusilc) - 1) , eusilc)

SE_lin2 <- function(t, des) {
  variance <-
    survey::svyrecvar(t / des$prob,
                      des$cluster,
                      des$strata,
                      des$fpc,
                      postStrata = des$postStrata)
  sqrt(variance)
}

des_eusilc <-
  svydesign(
    ids = ~ rb030,
    strata =  ~ db040,
    weights = ~ rb050,
    data = eusilc
  )

des_eusilc <- convey_prep(des_eusilc)

vardpoor_qsrw <-
  linqsr(
    Y = "eqincome",
    id = "IDd",
    weight = "rb050",
    dataset = dati
  )
vardest <- vardpoor_qsrw$value$QSR
attributes(vardest) <- NULL
vardest <- unlist(vardest)
varse <- SE_lin2(vardpoor_qsrw$lin$lin_qsr, des_eusilc)
attributes(varse) <- NULL
fun_qsrw <- svyqsr( ~ eqincome, design = des_eusilc)
convest <- coef(fun_qsrw)
attributes(convest) <- NULL
convse <- SE(fun_qsrw)
attributes(convse) <- NULL
#domain
vardpoor_qsrd <-
  linqsr(
    Y = "eqincome",
    id = "IDd",
    weight = "rb050",
    Dom = c("hsize"),
    dataset = dati
  )
#  point estimates
vardestd <- unlist(vardpoor_qsrd$value$QSR)
#  se estimates
varsed <-
  sapply(data.frame(vardpoor_qsrd$lin)[, 2:10], function(t)
    SE_lin2(t, des_eusilc))
attributes (varsed) <- NULL
# library convey
fun_qsrd <- svyby(
  ~ eqincome,
  by = ~ hsize,
  design = des_eusilc,
  FUN = svyqsr,
  deff = FALSE
)
convestd <- coef(fun_qsrd)
attributes(convestd) <- NULL
convsed <- SE(fun_qsrd)

test_that("compare results convey vs vardpoor", {
  expect_equal(vardest, convest)
  expect_equal(varse, convse)
  expect_equal(vardestd, convestd)

  expect_equal(varsed, convsed)
})