tests/testthat/test-bp-output.R

In cmlopera/olsrr: Tools for Building OLS Regression Models

context("breusch pagan test output")

model <- lm(mpg ~ disp + hp + wt + drat + qsec, data = mtcars)

test_that("when fitted.values == TRUE, fitted values from the regression\n\tare used for the test", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n             Data               \n -------------------------------\n Response : mpg \n Variables: fitted values of mpg \n\n        Test Summary         \n ----------------------------\n DF            =    1 \n Chi2          =    1.255517 \n Prob > Chi2   =    0.2625014 ")

  expect_output(print(ols_test_breusch_pagan(model)), x)
})

test_that("when rhs == TRUE, predictors from the regression\n\tare used for the test", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n             Data               \n -------------------------------\n Response : mpg \n Variables: disp hp wt drat qsec \n\n        Test Summary         \n ----------------------------\n DF            =    5 \n Chi2          =    2.489028 \n Prob > Chi2   =    0.7781466")

  expect_output(print(ols_test_breusch_pagan(model, rhs = TRUE)), x)
})

test_that("when rhs == TRUE and multiple == TRUE, multiple p values are\n\treturned", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n             Data               \n -------------------------------\n Response : mpg \n Variables: disp hp wt drat qsec \n\n        Test Summary (Unadjusted p values)       \n ----------------------------------------------\n  Variable           chi2       df        p     \n ----------------------------------------------\n  disp             0.9237291     1    0.3364977 \n  hp               0.7652006     1    0.3817059 \n  wt               0.7748714     1    0.3787143 \n  drat             0.7751270     1    0.3786356 \n  qsec             1.2902861     1    0.2559952 \n ----------------------------------------------\n  simultaneous     2.4890277     5    0.7781466 \n ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE)), x)
})


test_that("when rhs == TRUE, multiple == TRUE and p.adj == 'bonferroni'
	bonferroni adjusted p values are returned", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n             Data               \n -------------------------------\n Response : mpg \n Variables: disp hp wt drat qsec \n\n        Test Summary (Bonferroni p values)       \n ----------------------------------------------\n  Variable           chi2       df        p     \n ----------------------------------------------\n  disp             0.9237291     1    1.0000000 \n  hp               0.7652006     1    1.0000000 \n  wt               0.7748714     1    1.0000000 \n  drat             0.7751270     1    1.0000000 \n  qsec             1.2902861     1    1.0000000 \n ----------------------------------------------\n  simultaneous     2.4890277     5    0.7781466 \n ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = "bonferroni")), x)
})


test_that("when rhs == TRUE, multiple == TRUE and p.adj == 'holm',
	bonferroni adjusted p values are returned", {
  x <- cat("
 Breusch Pagan Test for Heteroskedasticity
 -----------------------------------------
 Ho: the variance is constant
 Ha: the variance is not constant

             Data
 -------------------------------
 Response : mpg
 Variables: disp hp wt drat qsec

          Test Summary (Holm's p values)
 ----------------------------------------------
  Variable           chi2       df        p
 ----------------------------------------------
  disp             0.9237291     1    1.0000000
  hp               0.7652006     1    0.3817059
  wt               0.7748714     1    0.7574285
  drat             0.7751270     1    1.0000000
  qsec             1.2902861     1    1.0000000
 ----------------------------------------------
  simultaneous     2.4890277     5    0.7781466
 ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = "holm")), x)
})


test_that("when rhs == TRUE, multiple == TRUE and p.adj == 'sidak',
	bonferroni adjusted p values are returned", {
  x <- cat("
 Breusch Pagan Test for Heteroskedasticity
 -----------------------------------------
 Ho: the variance is constant
 Ha: the variance is not constant

             Data
 -------------------------------
 Response : mpg
 Variables: disp hp wt drat qsec

          Test Summary (Sidak p values)
 ----------------------------------------------
  Variable           chi2       df        p
 ----------------------------------------------
  disp             0.9237291     1    0.8714086
  hp               0.7652006     1    0.9096401
  wt               0.7748714     1    0.9074328
  drat             0.7751270     1    0.9073743
  qsec             1.2902861     1    0.7720295
 ----------------------------------------------
  simultaneous     2.4890277     5    0.7781466
 ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = "sidak")), x)
})

test_that("when vars != NA, variables specified are used for the test", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n     Data       \n ---------------\n Response : mpg \n Variables: disp \n\n        Test Summary         \n ----------------------------\n DF            =    1 \n Chi2          =    0.9237291 \n Prob > Chi2   =    0.3364977")

  expect_output(print(ols_test_breusch_pagan(model, vars = c("disp"))), x)
})


test_that("when rhs == FALSE, multiple == TRUE and vars != NA,\n\tvariables specified are used for the test", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n       Data        \n ------------------\n Response : mpg \n Variables: disp hp \n\n        Test Summary (Unadjusted p values)       \n ----------------------------------------------\n  Variable           chi2       df        p     \n ----------------------------------------------\n  disp             0.9237291     1    0.3364977 \n  hp               0.7652006     1    0.3817059 \n ----------------------------------------------\n  simultaneous     0.9587887     2    0.6191583 \n ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(model, multiple = TRUE, rhs = FALSE, vars = c("disp", "hp"))), x)
})


test_that("when multiple == TRUE and vars != NA and p.adj == 'bonferroni',
	variables specified are used for the test", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n       Data        \n ------------------\n Response : mpg \n Variables: disp hp \n\n        Test Summary (Bonferroni p values)       \n ----------------------------------------------\n  Variable           chi2       df        p     \n ----------------------------------------------\n  disp             0.9237291     1    0.6729955 \n  hp               0.7652006     1    0.7634118 \n ----------------------------------------------\n  simultaneous     0.9587887     2    0.6191583 \n ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(
    model, multiple = TRUE, vars = c("disp", "hp"),
    p.adj = "bonferroni"
  )), x)
})


test_that("when multiple == TRUE and vars != NA and p.adj == 'sidak',
	variables specified are used for the test", {
  x <- cat("\n Breusch Pagan Test for Heteroskedasticity\n -----------------------------------------\n Ho: the variance is constant            \n Ha: the variance is not constant        \n\n       Data        \n ------------------\n Response : mpg \n Variables: disp hp \n\n          Test Summary (Sidak p values)          \n ----------------------------------------------\n  Variable           chi2       df        p     \n ----------------------------------------------\n  disp             0.9237291     1    0.5597648 \n  hp               0.7652006     1    0.6177124 \n ----------------------------------------------\n  simultaneous     0.9587887     2    0.6191583 \n ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(model, multiple = TRUE, vars = c("disp", "hp"), p.adj = "sidak")), x)
})


test_that("when multiple == TRUE and vars != NA and p.adj == 'holm',
	variables specified are used for the test", {
  x <- cat("
 Breusch Pagan Test for Heteroskedasticity
 -----------------------------------------
 Ho: the variance is constant
 Ha: the variance is not constant

       Data
 ------------------
 Response : mpg
 Variables: disp hp

          Test Summary (Holm's p values)
 ----------------------------------------------
  Variable           chi2       df        p
 ----------------------------------------------
  disp             0.9237291     1    0.6729955
  hp               0.7652006     1    0.3817059
 ----------------------------------------------
  simultaneous     0.9587887     2    0.6191583
 ----------------------------------------------")

  expect_output(print(ols_test_breusch_pagan(
    model, multiple = TRUE,
    vars = c("disp", "hp"), p.adj = "holm"
  )), x)
})


test_that("when multiple == TRUE, rhs == FALSE and one variable is specified", {
  x <- cat("
 Breusch Pagan Test for Heteroskedasticity
 -----------------------------------------
 Ho: the variance is constant
 Ha: the variance is not constant

     Data
 ---------------
 Response : mpg
 Variables: disp

        Test Summary
 ----------------------------
 DF            =    1
 Chi2          =    0.9237291
 Prob > Chi2   =    0.3364977")

  expect_output(print(ols_test_breusch_pagan(model, multiple = TRUE, rhs = FALSE, vars = c("disp"))), x)
})