tests/testthat/test_PEARSON.R

In NVE/fitdistrib: A list a functions to fit data to probability distributions

# Testing the pearson functions
# Problem: fitting functions for pearson do not return the standard error
# So the 95% confidence interval was replaced by +-20% accuracy

param <- c(9.5, 6, 2.5)  # Approximate parameters taken from the fit of the Narsjo "2.11" station data
CI <- 2  # +-2 times the Std_error to get 95% confidence interval

random_distrib <- nsRFA::rand.gamma(1000, param[1],param[2],param[3])

#####
test <- pearson_mle(random_distrib)

test_that("pearson_mle return list of correct length", {
  expect_length(test$estimate, 3)
  expect_length(test$se, 3)
})

test_that("pearson_mle returns reasonable estimates", {

#   expected_min <- param * (1 - CI * test$se)
#   expected_max <- param * (1 + CI * test$se)

  expected_min <- param * (1 - 0.2)
  expected_max <- param * (1 + 0.2)

  expect_true(test$estimate[1] < expected_max[1]  && test$estimate[1] > expected_min[1])
  expect_true(test$estimate[2] < expected_max[2]  && test$estimate[2] > expected_min[2])
  expect_true(test$estimate[3] < expected_max[3]  && test$estimate[3] > expected_min[3])
})

#####
test <- pearson_Lmom(random_distrib)

test_that("pearson_Lmom return list of correct length", {
  expect_length(test$estimate, 3)
  expect_length(test$se, 3)
})


test_that("pearson_Lmom returns reasonable estimates", {

#   expected_min <- param * (1 - CI * test$se)
#   expected_max <- param * (1 + CI * test$se)

  expected_min <- param * (1 - 0.2)
  expected_max <- param * (1 + 0.2)

  expect_true(test$estimate[1] < expected_max[1]  && test$estimate[1] > expected_min[1])
  expect_true(test$estimate[2] < expected_max[2]  && test$estimate[2] > expected_min[2])
  expect_true(test$estimate[3] < expected_max[3]  && test$estimate[3] > expected_min[3])
})

####
test <- pearson_mom(random_distrib)

test_that("pearson_mom return list of correct length", {
  expect_length(test$estimate, 3)
  expect_length(test$se, 3)
})


test_that("pearson_mom returns reasonable estimates", {

  #   expected_min <- param * (1 - CI * test$se)
  #   expected_max <- param * (1 + CI * test$se)

  expected_min <- param * (1 - 0.2)
  expected_max <- param * (1 + 0.2)

  expect_true(test$estimate[1] < expected_max[1]  && test$estimate[1] > expected_min[1])
  expect_true(test$estimate[2] < expected_max[2]  && test$estimate[2] > expected_min[2])
  # expect_true(test$estimate[3] < expected_max[3]  && test$estimate[3] > expected_min[3])
})