tests/testthat/test-svygini.R

In DjalmaPessoa/convey: Income Concentration Analysis with Complex Survey Samples

# compares with vardpoor output

# load libraries
library(laeken)
library(survey)
library(vardpoor)

# test context
context("svygini comparison with vardpoor")

# collect and format data
data(eusilc)
names(eusilc) <- tolower(names(eusilc))

# create new data.frame
dati = data.frame(IDd = seq(10000 , 10000 + nrow(eusilc) - 1) , eusilc)

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

### convey calculations

# build survey design objects
des_eusilc <-
  svydesign(
    ids = ~ rb030 ,
    strata =  ~ db040 ,
    weights = ~ rb050 ,
    data = eusilc
  )
des_eusilc_rep <- as.svrepdesign(des_eusilc , type = "bootstrap")

# prepare survey design objects for convey
des_eusilc <- convey_prep(des_eusilc)
des_eusilc_rep <- convey_prep(des_eusilc_rep)

# calculate estimates using convey
fun_giniw <- svygini(~ eqincome , des_eusilc)
fun_giniw_rep <- svygini(~ eqincome , des_eusilc_rep)

# collect point estimates from convey object
convest <- coef(fun_giniw)
attributes(convest) <- NULL

# collect SE estimates from convey object
convse <- SE(fun_giniw)
attributes(convse) <- NULL

### vardpoor calculations

# calculate estimates with vardpoor
vardpoor_giniw <-
  lingini(
    Y = "eqincome",
    id = "IDd",
    weight = "rb050",
    Dom = NULL,
    dataset = dati
  )

# set up coefficients object
vardest <- vardpoor_giniw$value
attributes(vardest) <- NULL
vardest <- unlist(vardest)

# calculate SE using vardpoor linearization, but survey's variance estimation function
varse <- SE_lin2(vardpoor_giniw$lin$lin_gini , des_eusilc)
attributes(varse) <- NULL

##### domain estimation

### using vardpoor

# domain esitmation using vardpoor
vardpoor_ginid <-
  lingini(
    Y = "eqincome" ,
    id = "IDd" ,
    weight = "rb050" ,
    Dom = "hsize" ,
    dataset = dati
  )

# colect point estimates
vardestd <- unlist(vardpoor_ginid$value$Gini)

# calculate SE using vardpoor linearization, but survey's variance estimation function
varsed <-
  sapply(data.frame(vardpoor_ginid$lin)[, 2:10] , function(t)
    SE_lin2(t , des_eusilc))
attributes(varsed) <- NULL

### using convey

# calculate estimates
fun_ginid <- svyby(~ eqincome , ~ hsize , des_eusilc , svygini)

# collect point estimates
convestd <- coef(fun_ginid)
attributes(convestd) <- NULL

# collect SE estimates
convsed <- SE(fun_ginid)

# perform tests
test_that("compare results convey vs vardpoor", {
  # compare point estimates
  expect_equal(vardest[[1]] , convest * 100)

  # compare point estimates on domains
  expect_equal(vardestd , convestd * 100)

  # compare SE estimates
  expect_equal(varse , convse * 100)
  expect_equal(varsed , convsed * 100)

})