### library(poolr); library(testthat); Sys.setenv(NOT_CRAN="true")

source("tolerances.r")

context("Checking bonferroni() function")

test_that("bonferroni() works correctly under independence.", {

  res <- bonferroni(grid2ip.p)
  out <- capture.output(print(res))

  expect_equivalent(c(res$p), 0.03881585, tolerance = p_tol)
  expect_equivalent(c(res$statistic), 0.001687646, tolerance = stat_tol)

})

test_that("bonferroni() works correctly with effective number of tests.", {

  res_nyh <- bonferroni(grid2ip.p, adjust = "nyholt", R = mvnconv(grid2ip.ld, target = "p", cov2cor = TRUE))
  res_lj  <- bonferroni(grid2ip.p, adjust = "liji", R = mvnconv(grid2ip.ld, target = "p", cov2cor = TRUE))
  res_gao <- bonferroni(grid2ip.p, adjust = "gao", R = mvnconv(grid2ip.ld, target = "p", cov2cor = TRUE))
  res_gal <- bonferroni(grid2ip.p, adjust = "galwey", R = mvnconv(grid2ip.ld, target = "p", cov2cor = TRUE))
  res_che <- bonferroni(grid2ip.p, adjust = "chen", R = mvnconv(grid2ip.ld, target = "p", cov2cor = TRUE))
  res_user <- bonferroni(grid2ip.p, m = 18)

  out <- capture.output(print(res_nyh))
  out <- capture.output(print(res_lj))
  out <- capture.output(print(res_gao))
  out <- capture.output(print(res_gal))
  out <- capture.output(print(res_che))
  out <- capture.output(print(res_user))

  expect_equivalent(c(res_nyh$p), 0.0371282, tolerance = p_tol)
  expect_equivalent(c(res_nyh$statistic), 0.001687646, tolerance = stat_tol)

  expect_equivalent(c(res_lj$p), 0.03544056, tolerance = p_tol)
  expect_equivalent(c(res_lj$statistic), 0.001687646, tolerance = stat_tol)

  expect_equivalent(c(res_gao$p), 0.03881585, tolerance = p_tol)
  expect_equivalent(c(res_gao$statistic), 0.001687646, tolerance = stat_tol)

  expect_equivalent(c(res_gal$p), 0.03375291, tolerance = p_tol)
  expect_equivalent(c(res_gal$statistic), 0.001687646, tolerance = stat_tol)

  expect_equivalent(c(res_che$p), 0.0371282, tolerance = p_tol)
  expect_equivalent(c(res_che$statistic), 0.001687646, tolerance = stat_tol)

  expect_equivalent(c(res_user$p), 0.03037762, tolerance = p_tol)
  expect_equivalent(c(res_user$statistic), 0.001687646, tolerance = stat_tol)

})

test_that("bonferroni() works correctly with empirically-derived null distributions.", {

  set.seed(1234)
  res <- bonferroni(grid2ip.p, adjust = "empirical", R = grid2ip.ld)
  out <- capture.output(print(res))

  expect_equivalent(c(res$p), 0.03229677, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$statistic), 0.001687646, tolerance = stat_tol * emp_sca)
  expect_equivalent(c(res$ci[1]), 0.02891875, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$ci[2]), 0.0359506, tolerance = p_tol * emp_sca)

  set.seed(1234)
  res <- bonferroni(grid2ip.p, adjust = "empirical", R = grid2ip.ld, size = 100000)
  out <- capture.output(print(res))

  expect_equivalent(c(res$p), 0.03065969, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$statistic), 0.001687646, tolerance = stat_tol * emp_sca)
  expect_equivalent(c(res$ci[1]), 0.02959984, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$ci[2]), 0.03174688, tolerance = p_tol * emp_sca)

  set.seed(1234)
  res <- bonferroni(grid2ip.p, adjust = "empirical", R = grid2ip.ld, size = 1000000, batchsize = 1000)
  out <- capture.output(print(res))

  expect_equivalent(c(res$p), 0.03024897, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$statistic), 0.001687646, tolerance = stat_tol * emp_sca)
  expect_equivalent(c(res$ci[1]), 0.02991414, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$ci[2]), 0.03058652, tolerance = p_tol * emp_sca)

  set.seed(1234)
  res <- bonferroni(grid2ip.p, adjust = "empirical", R = grid2ip.ld, size = c(1000, 10000, 100000), threshold = c(0.10, 0.01))
  out <- capture.output(print(res))

  expect_equivalent(c(res$p), 0.03139686, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$statistic), 0.001687646, tolerance = stat_tol * emp_sca)
  expect_equivalent(c(res$ci[1]), 0.02806613, tolerance = p_tol * emp_sca)
  expect_equivalent(c(res$ci[2]), 0.035004, tolerance = p_tol * emp_sca)

})