tests/testthat/test-dfba_beta_bayes_factor.R

In danbarch/dfba: What the Package Does (One Line, Title Case)

test_beta_bayes_factor<-function(dfba_BF=rep(0,8)){
  ## The function does eight tests of dfba_beta_bayes_factor
  # The first four tests check the values for an interval
  # type null hypothesis for canned values for a beta.
  # Tests 1 and 2 check the values for pH1 and postH1
  # Tests 3 and 4 check the values for BF10 and BF01
  ## The last four tests check the values for a point
  # null hypothesis. Tests 5 and 6 check the values
  # for the prior and posterior probability density
  # at the null point. Tests 7 and 8 check on BF10 and BF01
  ## The output of the function is a vector of eight
  # values which are either a zero or one. These values
  # correspond to the eight tests. If a test passes, it
  # is a 0, and if a test fails, it is 1.
  ## The output vector is dfba_BF.

  dfba_BF=rep(0,8)
  ABF<-dfba_beta_bayes_factor(a = 17,b = 5,method = "interval",H0 = c(0, .5))

  pH1_test=abs(ABF$pH1-.5)
  if (pH1_test>3e-05){dfba_BF[1]=1}

  postH1_test=abs(ABF$postH1-.9964013)
  if (postH1_test>3e-05){dfba_BF[2]=1}

  BF10_test=abs(ABF$BF10-276.8789)
  if (BF10_test>2.4){dfba_BF[3]=1}

  BF01_test=abs(ABF$BF01-0.003611687)
  if (BF01_test>4e-05){dfba_BF[4]=1}

  BBF<-dfba_beta_bayes_factor(a = 377,b = 123,method = "point",H0 = .75)

  dpriorH0_test=abs(BBF$dpriorH0-1)
  if (dpriorH0_test>3e-05){dfba_BF[5]=1}

  dpostH0_test=abs(BBF$dpostH0-20.04153)
  if (dpostH0_test>3e-05){dfba_BF[6]=1}

  BF10_test2=abs(BBF$BF10-0.04989638)
  if (BF10_test2>3e-05){dfba_BF[7]=1}

  BF01_test2=abs(BBF$BF01-20.04153)
  if (BF01_test2>5e-04){dfba_BF[8]=1}

  failtot=sum(dfba_BF)
  if (failtot==0){
    cat("dfba_beta_bayes_factor passes", " ","\n") } else {
      cat("dfba_beta_bayes_factor fails. There are","\n")
      cat(failtot," failed subtests of the function.","\n")
    }
  beta_BF_list=list(dfba_BF=dfba_BF)

}