tests/testthat/testthat_wsValidate.R

In droglenc/FSAWs: Construct and Validate Standard Weight (Ws) Equations for Fish

## Test Messages


## Test Output Types


## Validate Results
test_that("wsValidate() match Ogle & Winfield (2009) results for Ruffe", {
  source_test_helpers()

  ## Compare 75th percentile RLP results to Table 2 of Ogle & Winfield
  ### Willis method
  res <- wsValidate(ruf75.rlp,rWs.v,"regrnum","tl","wt",min=60,max=210,w=10,
                    type="Willis")
  tmp <- summary(res)
  expect_identical(tmp$statistic,c("number of successes"=15))
  expect_identical(tmp$parameter,c("number of trials"=20))
  expect_equal(tmp$p.value,c("Negative"=0.0414),tolerance=0.0001)
  ### Weighted EmpQ method
  res <- wsValidate(ruf75.rlp,rWs.v,"regrnum","tl","wt",min=60,max=220,w=10,
                    type="EmpQ",weighted=TRUE)
  tmp <- anova(res)
  expect_identical(tmp$Df,c(1L,1L,10L))
  expect_equal(tmp$`Pr(>F)`,c(0.3620,0.0034,NA),tolerance=0.0001)
  
  ## Compare 50th percentile RLP results to Table 2 of Ogle & Winfield
  ### Willis method
  res <- wsValidate(ruf50.rlp,rWs.v,"regrnum","tl","wt",min=60,max=210,w=10,
                    type="Willis")
  tmp <- summary(res)
  expect_identical(tmp$statistic,c("number of successes"=14))
  expect_identical(tmp$parameter,c("number of trials"=19))
  expect_equal(tmp$p.value,c("Negative"=0.0636),tolerance=0.0001)
  ### Weighted EmpQ method
  res <- wsValidate(ruf50.rlp,rWs.v,"regrnum","tl","wt",min=60,max=220,w=10,
                    type="EmpQ",weighted=TRUE,probs=0.5)
  tmp <- anova(res)
  expect_identical(tmp$Df,c(1L,1L,10L))
  expect_equal(tmp$`Pr(>F)`,c(0.2272,0.0750,NA),tolerance=0.0001)
  
  ## Compare 75th percentile results EMP results to Table 2 of Ogle & Winfield
  ### Willis method
  res <- wsValidate(ruf75.emp,rWs.v,"regrnum","tl","wt",min=60,max=210,w=10,
                    type="Willis")
  tmp <- summary(res)
  expect_identical(tmp$statistic,c("number of successes"=15))
  expect_identical(tmp$parameter,c("number of trials"=22))
  expect_equal(tmp$p.value,c("Negative"=0.1338),tolerance=0.0001)
  ### Weighted EmpQ method
  res <- wsValidate(ruf75.emp,rWs.v,"regrnum","tl","wt",min=60,max=220,w=10,
                    type="EmpQ",weighted=TRUE)
  tmp <- anova(res)
  expect_identical(tmp$Df,c(1L,1L,10L))
  expect_equal(tmp$`Pr(>F)`,c(0.4401,0.9174,NA),tolerance=0.0001)

  ## Compare 50th percentile results EMP results to Table 2 of Ogle & Winfield
  ### Willis method
  res <- wsValidate(ruf50.emp,rWs.v,"regrnum","tl","wt",min=60,max=210,w=10,
                    type="Willis")
  tmp <- summary(res)
  expect_identical(tmp$statistic,c("number of successes"=12))
  expect_identical(tmp$parameter,c("number of trials"=20))
  expect_equal(tmp$p.value,c("Negative"=0.5034),tolerance=0.0001)
  ### Weighted EmpQ method
  res <- wsValidate(ruf50.emp,rWs.v,"regrnum","tl","wt",min=60,max=220,w=10,
                    type="EmpQ",weighted=TRUE,probs=0.5)
  tmp <- anova(res)
  expect_identical(tmp$Df,c(1L,1L,10L))
  expect_equal(tmp$`Pr(>F)`,c(0.4838,0.9368,NA),tolerance=0.0001)
  
  ## Compare 75th percentile (no quad) EMP results to Table 2 of Ogle & Winfield
  res <- wsValidate(ruf75nq.emp,rWs.v,"regrnum","tl","wt",min=60,max=210,w=10,
                    type="Willis")
  tmp <- summary(res)
  expect_identical(tmp$statistic,c("number of successes"=16))
  expect_identical(tmp$parameter,c("number of trials"=17))
  expect_equal(tmp$p.value,c("Negative"=0.0003),tolerance=0.0001)
  ### Weighted EmpQ method
  res <- wsValidate(ruf75nq.emp,rWs.v,"regrnum","tl","wt",min=60,max=220,w=10,
                    type="EmpQ",weighted=TRUE)
  tmp <- anova(res)
  expect_identical(tmp$Df,c(1L,1L,10L))
  expect_equal(tmp$`Pr(>F)`,c(0.5308,0.0024,NA),tolerance=0.0001)
  
  ## Compare 50th percentile (no quad) EMP results to Table 2 of Ogle & Winfield
  ### Willis method
  res <- wsValidate(ruf50nq.emp,rWs.v,"regrnum","tl","wt",min=60,max=210,w=10,
                    type="Willis")
  tmp <- summary(res)
  expect_identical(tmp$statistic,c("number of successes"=14))
  expect_identical(tmp$parameter,c("number of trials"=19))
  expect_equal(tmp$p.value,c("Negative"=0.0636),tolerance=0.0001)
  ### Weighted EmpQ method
  res <- wsValidate(ruf50nq.emp,rWs.v,"regrnum","tl","wt",min=60,max=220,w=10,
                    type="EmpQ",weighted=TRUE,probs=0.5)
  tmp <- anova(res)
  expect_identical(tmp$Df,c(1L,1L,10L))
  expect_equal(tmp$`Pr(>F)`,c(0.3844,0.0685,NA),tolerance=0.0001)
  
  ## Compare Froese results to Table 2 of Ogle & Winfield
  ### Willis method
  res <- wsValidate(ruf.Froese,rWs.v,"regrnum","tl","wt",min=60,max=210,w=10,
                    type="Willis")
  tmp <- summary(res)
  expect_identical(tmp$statistic,c("number of successes"=13))
  expect_identical(tmp$parameter,c("number of trials"=19))
  expect_equal(tmp$p.value,c("Negative"=0.1671),tolerance=0.0001)
  ### Weighted EmpQ method
  res <- wsValidate(ruf.Froese,rWs.v,"regrnum","tl","wt",min=60,max=220,w=10,
                    type="EmpQ",weighted=TRUE,use.means=TRUE)
  tmp <- anova(res)
  expect_identical(tmp$Df,c(1L,1L,10L))
  expect_equal(tmp$`Pr(>F)`,c(0.4151,0.0011,NA),tolerance=0.0001)
})