context("test scan_test related functions")
data(nydf)
cases <- floor(nydf$cases)
pop <- nydf$pop
ty <- sum(cases)
tpop <- sum(pop)
ex <- ty / tpop * pop

# number of cases and population
# poisson test statistic
zone <- c(
  52, 50, 53, 38, 49, 48, 15, 39, 37, 1, 16, 44,
  47, 40, 14, 2, 51, 13, 43, 45, 17, 55, 11, 3,
  12, 46, 36, 35, 54, 10, 5
)
yin <- sum(cases[zone])
ein <- sum(ex[zone])
yout <- ty - yin
eout <- ty - ein
t1 <- scan.stat(yin = yin, ty = ty, ein = ein)
t1_ <- scan_stat(yin = yin, ty = ty, ein = ein)
t2 <- stat.poisson(yin = yin, yout = yout, ein = ein, eout = eout)
t2_ <- stat_poisson(yin = yin, yout = yout, ein = ein, eout = eout)
t2_adj <- stat_poisson_adj(
  yin = yin, ty = ty,
  logein = log(ein),
  logeout = log(eout)
)

zone <- c(1, 2, 3, 12, 13, 14, 15, 35, 47, 49)
yin <- sum(cases[zone])
popin <- sum(pop[zone])
yout <- ty - yin
popout <- tpop - popin
t3 <- scan.stat(
  yin = yin, ty = ty, popin = popin, tpop = tpop,
  type = "binomial"
)
t3_ <- scan_stat(
  yin = yin, ty = ty, popin = popin, tpop = tpop,
  type = "binomial"
)

t4 <- stat.binom(
  yin = yin, yout = yout, ty = ty,
  popin = popin, popout = popout, tpop = tpop
)
t4_ <- stat_binom(
  yin = yin, yout = yout, ty = ty,
  popin = popin, popout = popout, tpop = tpop
)
t4_adj <- stat_binom_adj(
  yin = yin, ty = ty,
  popin = popin, popout = popout,
  logpopin = log(popin),
  logpopout = log(popout),
  tpop = tpop
)

test_that("check accuracy for scan.stat for NY data", {
  # taken from satscannyoutpoisson
  expect_equal(round(t1, 6), 14.780276)
  expect_equal(round(t2, 6), 14.780276)
  # taken from scan_binomial
  expect_equal(round(t3, 5), 8.47836)
  expect_equal(round(t4, 5), 8.47836)

  # compare stat.poisson, stat_poisson, etc
  expect_equal(t1, t1_)
  expect_equal(t2, t2_)
  expect_equal(t3, t3_)
  expect_equal(t4, t4_)
})

test_that("check accuracy for stat_poisson_adj and stat_binomial_adj", {

  # compare stat_poisson, stat_poisson_adj, etc
  expect_equal(t2, t2_adj)
  expect_equal(t4, t4_adj)
})