tests/testthat/test-predict_race_2010.R

In wru: Who are You? Bayesian Prediction of Racial Category Using Surname, First Name, Middle Name, and Geolocation

# Note: must provide a valid U.S. Census API key for test cases that use U.S. Census statistics
# > usethis::edit_r_profile
# Sys.setenv("CENSUS_API_KEY" = "yourkey")
options("piggyback.verbose" = FALSE)
options("wru_data_wd" = TRUE)

test_that("Tests surname only predictions", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  # Prediction using surname only
  x <- suppressMessages(predict_race(
    voter.file = voters,
    year = 2010,
    surname.only = TRUE))
  # Test and confirm prediction output is as expected
  expect_equal(dim(x), c(10, 20))
  expect_equal(sum(is.na(x)), 0)
  expect_equal(round(x[x$surname == "Khanna", "pred.whi"], 4), 0.045, tolerance = 0.01)
  expect_equal(round(x[x$surname == "Johnson", "pred.his"], 4), 0.0272, tolerance = 0.01)
})

test_that("Test BISG NJ at county level", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  census <- readRDS(test_path("data/census_test_nj_block_2010.rds"))
  x <- suppressMessages(predict_race(
      voter.file = voters[voters$state == "NJ",],
      year = 2010,
      census.geo = "county",
      census.data = census
    ))

  expect_equal(as.character(x$VoterID), as.character(c(1, 2, 4, 5, 6, 8, 9)))
  expect_equal(dim(x), c(7, 20))
  expect_equal(sum(is.na(x)), 0L)
  expect_equal(sum(x$surname == "Johnson"), 0)
  expect_equal(round(x[x$surname == "Khanna", "pred.whi"], 4), 0.0314, tolerance = 0.01)
  expect_equal(round(x[x$surname == "Khanna", "pred.asi"], 4), 0.9367, tolerance = 0.01)
  expect_equal(round(x[x$surname == "Fifield", "pred.whi"], 4), 0.9230, tolerance = 0.01)
  expect_equal(round(x[x$surname == "Lopez", "pred.his"], 4), 0.9178, tolerance = 0.01)
})

test_that("Test fBISG NJ at tract level", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  
  census <- readRDS(test_path("data/census_test_nj_block_2010.rds"))

  x <- suppressMessages(predict_race(
    voter.file = voters[voters$state == "NJ",],
    year = 2010,
    census.geo = "tract",
    census.data = census,
    model = "fBISG",
    control=list(verbose=FALSE)
  ))
  
  expect_equal(as.character(x$VoterID), as.character(c(1, 2, 4, 5, 6, 8, 9)))
  expect_equal(dim(x), c(7, 20))
  expect_equal(sum(is.na(x)), 0L)
  expect_equal(sum(x$surname == "Johnson"), 0)
  expect_equal(round(x[x$surname == "Khanna", "pred.whi"], 4), 0.063, tolerance = 0.01) # 0.0644
  expect_equal(round(x[x$surname == "Lopez", "pred.his"], 4), 0.78, tolerance = 0.01) # 0.0644
})

test_that("BISG NJ at block level", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  census <- readRDS(test_path("data/census_test_nj_block_2010.rds"))
  voters[voters$surname=="Ratkovic", "block"] <- "3001"
  
  x <- suppressMessages(predict_race(
    voter.file = voters[voters$state == "NJ", ],
    year = 2010,
    census.geo = "block", 
    census.key = NULL, 
    census.data = census, 
    use.counties = TRUE)
  )
  
  expect_equal(dim(x), c(7, 20))
  expect_equal(sum(is.na(x$pred.asi)), 0L)
  expect_true(!any(duplicated(x$surname)))
  expect_equal(x[x$surname == "Khanna", "pred.asi"], 0.7640, tolerance = 0.01)
  expect_equal(x[x$surname == "Zhou", "pred.asi"], 1.0, tolerance = 0.1)
  expect_equal(x[x$surname == "Lopez", "pred.his"], 0.7, tolerance = 0.1)
})

test_that("BISG NJ at block_group level", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  census <- readRDS(test_path("data/census_test_nj_block_2010.rds"))
  
  voters <- voters[voters$state == "NJ", ]
  voters$block_group <- "1"
  
  x <- suppressMessages(predict_race(
    voter.file = voters,
    year = 2010,
    census.geo = "block_group", 
    census.key = NULL, 
    census.data = census, 
    use.counties = TRUE)
  )
  
  expect_equal(dim(x), c(7, 21))
  expect_equal(sum(is.na(x$pred.asi)), 0)
  expect_true(!any(duplicated(x$surname)))
  expect_equal(x[x$surname == "Khanna", "pred.asi"], 0.9183, tolerance = 0.01)
  expect_equal(x[x$surname == "Zhou", "pred.asi"], 1.0, tolerance = 0.01)
  expect_equal(x[x$surname == "Lopez", "pred.his"], 0.75, tolerance = 0.01)
})

test_that("Fails on territories", {
  data(voters)
  voters$state <- "GU"
  expect_error(
    predict_race(voter.file = voters, year = 2010),
    "The wru package does not support US territories"
  )
}) 

test_that("Fails on missing geolocation if skip_bad_geos default is used", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  census <- readRDS(test_path("data/census_test_nj_block_2010.rds"))
  expect_error(suppressMessages(predict_race(
    voter.file = voters[voters$state == "NJ", ],
    year = 2010,
    census.geo = "block", 
    census.key = NULL, 
    census.data = census, 
    use.counties = TRUE)
  ),
  "Stopping predictions. Please revise"
  )
})

test_that("Skip_bad_geos option successfully returns working geolocations", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  census <- readRDS(test_path("data/census_test_nj_block_2010.rds"))
  test_drop <- suppressMessages(predict_race(
    voter.file = voters[voters$state == "NJ", ], 
    year = 2010,
    census.geo = "block", 
    census.key = NULL, 
    census.data = census,
    skip_bad_geos = TRUE,
    use.counties = TRUE)
  )
  expect_equal(nrow(test_drop), 6)
})

test_that("Handles zero-pop. geolocations", {
  skip_on_cran()
  set.seed(42)
  data(voters)
  census <- readRDS(test_path("data/census_test_nj_block_2010.rds"))
  census$NJ$county[6,grep("P005",colnames(census$NJ$county))] <- 0
  x <- suppressMessages(predict_race(
    voter.file = voters[voters$state == "NJ", ],
    year = 2010,
    census.geo = "county", 
    census.key = NULL, 
    census.data = census, 
    use.counties = TRUE)
  )
  expect_equal(dim(x), c(7, 20))
  expect_equal(sum(is.na(x$pred.asi)), 0)
  expect_true(!any(duplicated(x$surname)))
  expect_equal(x[x$surname == "Khanna", "pred.asi"], 0.91, tolerance = 0.01)
  expect_equal(x[x$surname == "Zhou", "pred.asi"], 0.99, tolerance = 0.01)
  expect_equal(x[x$surname == "Lopez", "pred.his"], 0.92, tolerance = 0.01)
})

test_that("Fixes for issue #68 work as expected", {
  # skip_on_cran()
  set.seed(42)
  surname <- c("SULLIVAN")
  one <- predict_race(voter.file=data.frame(surname), year = 2010, surname.only=TRUE)
  
  surname <- c("SULLIVAN", "SULLIVAN")
  two <- predict_race(voter.file=data.frame(surname), year = 2010, surname.only=TRUE)
  
  surname <- c("SULLIVAN", "SULLIVAN", "SULLIVAN")
  three <- predict_race(voter.file=data.frame(surname), year = 2010, surname.only=TRUE)
  
  expect_equal(one$pred.whi, 0.8397254)
  expect_equal(two$pred.whi[1], 0.8397254)
  expect_equal(two$pred.whi[2], 0.8397254)
  
  expect_equal(three$pred.whi[1], 0.8397254)
  expect_equal(three$pred.whi[2], 0.8397254)
  expect_equal(three$pred.whi[3], 0.8397254)
})