tests/testthat/test_edata_replace.R

In pmartR/pmartR: Panomics Marketplace - Quality Control and Statistical Analysis for Panomics Data

context('edata replace')

test_that('edata_replace correctly replaces one value with another', {
  # Load: isobaricpepData ------------------------------------------------------

  load(system.file('testdata',
    'little_isodata.RData',
    package = 'pmartR'
  ))

  # Test: incorrect class ------------------------------------------------------

  err_1 <- "omicsData must be of class 'pepData', 'proData', 'metabData', 'lipidData', or 'nmrData'. "
  testthat::expect_error(edata_replace(60, 0, NA), err_1) ## error

  # Test: isobaricpepData ------------------------------------------------------

  # Fabricate an isobaricpepData object.
  isodata <- as.isobaricpepData(
    e_data = edata,
    f_data = fdata,
    e_meta = emeta,
    edata_cname = 'Peptide',
    fdata_cname = 'Sample',
    emeta_cname = 'Protein'
  )


  # Replace NAs with 0s in isodata.
  expect_message(
    isodata2 <- edata_replace(
      omicsData = isodata,
      x = NA,
      y = 0
    ),
    "400 instances of NA have been replaced with 0"
  )

  expect_message(
    edata_replace(
      omicsData = isodata,
      x = 589.917555116065,
      y = 0
    ),
    "1 instances of 589.917555116065 have been replaced with 0"
  )

  # Check that the elements of the data_info attribute are all correct.
  expect_equal(
    attr(isodata2, "data_info"),
    list(
      data_scale_orig = "abundance",
      data_scale = "abundance",
      norm_info = list(is_normalized = FALSE),
      num_edata = length(unique(isodata2$e_data[, 1])),
      num_miss_obs = sum(is.na(isodata2$e_data)),
      prop_missing = (sum(is.na(isodata2$e_data)) /
        prod(dim(isodata2$e_data[, -1]))),
      num_samps = ncol(isodata2$e_data[, -1]),
      data_types = NULL,
      batch_info = list(is_bc = FALSE)
    )
  )

  # Make sure all other attributes have not changed.
  expect_equal(
    attr(isodata, 'cnames'),
    attr(isodata2, 'cnames')
  )
  expect_equal(
    attr(isodata, 'isobaric_info'),
    attr(isodata2, 'isobaric_info')
  )
  expect_equal(
    attr(isodata, 'meta_info'),
    attr(isodata2, 'meta_info')
  )
  expect_equal(
    attr(isodata, 'group_DF'),
    attr(isodata2, 'group_DF')
  )
  expect_equal(
    attr(isodata, 'filters'),
    attr(isodata2, 'filters')
  )

  # Ensure the number of replaced elements is correct.
  expect_equal(
    sum(isodata2$e_data == 0),
    400
  )

  # Examine the class of the new isodata object.
  expect_s3_class(
    isodata2,
    c('isobaricpepData', 'pepData')
  )

  # Use a threshold
  thresh = quantile(isodata$e_data[,-1], na.rm=TRUE)[2]
  num_thresholded <- sum(isodata$e_data[,-1] < thresh, na.rm = TRUE)

  expect_message(
    isodata2 <- edata_replace(
      omicsData = isodata,
      x = NA,
      y = 0,
      threshold = thresh
    ),
    sprintf("%s values below the threshold %s have been replaced with %s", num_thresholded, thresh, 0)
  )

  expect_error(
    edata_replace(
      omicsData = isodata,
      x = NA,
      y = 0,
      threshold = min(isodata$e_data[,-1], na.rm=TRUE) - 1
    ),
    'outside the range of the data'
  )

  expect_equal(sum(isodata2$e_data == 0), 400 + num_thresholded)

  # Load: lipidData ------------------------------------------------------------

  load(system.file('testdata',
    'lipidData.RData',
    package = 'pmartR'
  ))

  # Test: lipidData ------------------------------------------------------------

  # Produce a lipidData object.
  ldata <- as.lipidData(
    e_data = edata,
    f_data = fdata,
    e_meta = emeta,
    edata_cname = 'LipidCommonName',
    fdata_cname = 'Sample_Name',
    emeta_cname = 'LipidClass'
  )

  # Replace NAs with 0s in ldata.
  expect_message(
    ldata2 <- edata_replace(
      omicsData = ldata,
      x = NA,
      y = 0
    ),
    "884 instances of NA have been replaced with 0"
  )

  # Check that the elements of the data_info attribute are all correct.
  expect_equal(
    attr(ldata2, "data_info"),
    list(
      data_scale_orig = "abundance",
      data_scale = "abundance",
      norm_info = list(is_normalized = FALSE),
      num_edata = length(unique(ldata2$e_data[, 1])),
      num_miss_obs = sum(is.na(ldata2$e_data)),
      prop_missing = (sum(is.na(ldata2$e_data)) /
        prod(dim(ldata2$e_data[, -1]))),
      num_samps = ncol(ldata2$e_data[, -1]),
      data_types = NULL,
      batch_info = list(is_bc = FALSE)
    )
  )

  # Make sure all other attributes have not changed.
  expect_equal(
    attr(ldata, 'cnames'),
    attr(ldata2, 'cnames')
  )
  expect_equal(
    attr(ldata, 'meta_info'),
    attr(ldata2, 'meta_info')
  )
  expect_equal(
    attr(ldata, 'group_DF'),
    attr(ldata2, 'group_DF')
  )
  expect_equal(
    attr(ldata, 'filters'),
    attr(ldata2, 'filters')
  )

  # Ensure the number of replaced elements is correct.
  expect_equal(
    sum(ldata2$e_data == 0),
    884
  )

  # Examine the class of the new ldata object.
  expect_s3_class(
    ldata2,
    'lipidData'
  )

  # Use a threshold
  thresh = quantile(ldata$e_data[,-1], na.rm=TRUE)[2]
  num_thresholded <- sum(ldata$e_data[,-1] < thresh, na.rm = TRUE)

  expect_message(
    ldata2 <- edata_replace(
      omicsData = ldata,
      x = NA,
      y = 0,
      threshold = thresh
    ),
    sprintf("%s values below the threshold %s have been replaced with %s", num_thresholded, thresh, 0)
  )

  expect_error(
    edata_replace(
      omicsData = ldata,
      x = NA,
      y = 0,
      threshold = max(ldata$e_data[,-1], na.rm=TRUE) + 1
    ),
    'outside the range of the data'
  )

  expect_equal(sum(ldata2$e_data == 0), 884 + num_thresholded)

  # Load: metabData ------------------------------------------------------------

  load(system.file('testdata',
    'metaboliteData.RData',
    package = 'pmartR'
  ))

  # Test: metabData ------------------------------------------------------------

  # Forge a metabData object.
  mdata <- as.metabData(
    e_data = edata,
    f_data = fdata,
    e_meta = emeta,
    edata_cname = 'Metabolite',
    fdata_cname = 'SampleID',
    emeta_cname = 'MClass'
  )

  # Replace NAs with 0s in mdata.
  expect_message(
    mdata2 <- edata_replace(
      omicsData = mdata,
      x = NA,
      y = 0
    ),
    "148 instances of NA have been replaced with 0"
  )

  # Check that the elements of the data_info attribute are all correct.
  expect_equal(
    attr(mdata2, "data_info"),
    list(
      data_scale_orig = "abundance",
      data_scale = "abundance",
      norm_info = list(is_normalized = FALSE),
      num_edata = length(unique(mdata2$e_data[, 1])),
      num_miss_obs = sum(is.na(mdata2$e_data)),
      prop_missing = (sum(is.na(mdata2$e_data)) /
        prod(dim(mdata2$e_data[, -1]))),
      num_samps = ncol(mdata2$e_data[, -1]),
      data_types = NULL,
      batch_info = list(is_bc = FALSE)
    )
  )

  # Make sure all other attributes have not changed.
  expect_equal(
    attr(mdata, 'cnames'),
    attr(mdata2, 'cnames')
  )
  expect_equal(
    attr(mdata, 'meta_info'),
    attr(mdata2, 'meta_info')
  )
  expect_equal(
    attr(mdata, 'group_DF'),
    attr(mdata2, 'group_DF')
  )
  expect_equal(
    attr(mdata, 'filters'),
    attr(mdata2, 'filters')
  )

  # Ensure the number of replaced elements is correct.
  expect_equal(
    sum(mdata2$e_data == 0),
    148
  )

  # Examine the class of the new mdata object.
  expect_s3_class(
    mdata2,
    'metabData'
  )

  # Use a threshold
  thresh = quantile(mdata$e_data[,-1], na.rm=TRUE)[2]
  num_thresholded <- sum(mdata$e_data[,-1] < thresh, na.rm = TRUE)

  expect_message(
    mdata2 <- edata_replace(
      omicsData = mdata,
      x = NA,
      y = 0,
      threshold = thresh
    ),
    sprintf("%s values below the threshold %s have been replaced with %s", num_thresholded, thresh, 0)
  )

  expect_equal(sum(mdata2$e_data == 0), 148 + num_thresholded)

  # Load: nmrData --------------------------------------------------------------

  load(system.file('testdata',
    'nmrData.RData',
    package = 'pmartR'
  ))

  # Alter the original nmr data frame to contain NAs ---------------------------

  set.seed(22)

  # Select row indices.
  rows <- sample(1:38, 338, replace = TRUE)

  # Select column indices.
  columns <- sample(2:42, 338, replace = TRUE)

  # Loop through each row column pair.
  for (e in 1:338) {
    # Replace current value in edata with NA.
    edata[rows[[e]], columns[[e]]] <- NA
  }

  # Test: nmrData --------------------------------------------------------------

  # Fabricate an nmrData object.
  nmrdata <- as.nmrData(
    e_data = edata,
    f_data = fdata,
    e_meta = emeta,
    edata_cname = 'Metabolite',
    fdata_cname = 'SampleID',
    emeta_cname = 'nmrClass'
  )

  # Replace NAs with 0s in nmrdata.
  expect_message(
    nmrdata2 <- edata_replace(
      omicsData = nmrdata,
      x = NA,
      y = 0
    ),
    "302 instances of NA have been replaced with 0"
  )

  # Check that the elements of the data_info attribute are all correct.
  expect_equal(
    attr(nmrdata2, "data_info"),
    list(
      data_scale_orig = "abundance",
      data_scale = "abundance",
      norm_info = list(is_normalized = FALSE),
      num_edata = length(unique(nmrdata2$e_data[, 1])),
      num_miss_obs = sum(is.na(nmrdata2$e_data)),
      prop_missing = (sum(is.na(nmrdata2$e_data)) /
        prod(dim(nmrdata2$e_data[, -1]))),
      num_samps = ncol(nmrdata2$e_data[, -1]),
      data_types = NULL,
      batch_info = list(is_bc = FALSE)
    )
  )

  # Make sure all other attributes have not changed.
  expect_equal(
    attr(nmrdata, 'cnames'),
    attr(nmrdata2, 'cnames')
  )
  expect_equal(
    attr(nmrdata, 'nmr_info'),
    attr(nmrdata2, 'nmr_info')
  )
  expect_equal(
    attr(nmrdata, 'meta_info'),
    attr(nmrdata2, 'meta_info')
  )
  expect_equal(
    attr(nmrdata, 'group_DF'),
    attr(nmrdata2, 'group_DF')
  )
  expect_equal(
    attr(nmrdata, 'filters'),
    attr(nmrdata2, 'filters')
  )

  # Ensure the number of replaced elements is correct.
  expect_equal(
    sum(nmrdata2$e_data == 0),
    302
  )

  # Examine the class of the new nmrdata object.
  expect_s3_class(
    nmrdata,
    'nmrData'
  )

  # Use a threshold
  thresh = quantile(nmrdata$e_data[,-1], na.rm=TRUE)[2]
  num_thresholded <- sum(nmrdata$e_data[,-1] < thresh, na.rm = TRUE)

  expect_message(
    nmrdata2 <- edata_replace(
      omicsData = nmrdata,
      x = NA,
      y = 0,
      threshold = thresh
    ),
    sprintf("%s values below the threshold %s have been replaced with %s", num_thresholded, thresh, 0)
  )

  expect_error(
    edata_replace(
      omicsData = nmrdata,
      x = NA,
      y = 0,
      threshold = max(nmrdata$e_data[,-1], na.rm=TRUE) + 1
    ),
    'outside the range of the data'
  )

  expect_equal(sum(nmrdata2$e_data == 0), 302 + num_thresholded)

  # Load: pepData --------------------------------------------------------------

  load(system.file('testdata',
    'little_pdata.RData',
    package = 'pmartR'
  ))

  # Test: pepData --------------------------------------------------------------

  # Construct a pepData object.
  pdata <- as.pepData(
    e_data = edata,
    f_data = fdata,
    e_meta = emeta,
    edata_cname = 'Mass_Tag_ID',
    fdata_cname = 'SampleID',
    emeta_cname = 'Protein'
  )

  # Replace NAs with 0s in pdata.
  expect_message(
    pdata2 <- edata_replace(
      omicsData = pdata,
      x = NA,
      y = 0
    ),
    "341 instances of NA have been replaced with 0"
  )

  # Check that the elements of the data_info attribute are all correct.
  expect_equal(
    attr(pdata2, "data_info"),
    list(
      data_scale_orig = "abundance",
      data_scale = "abundance",
      norm_info = list(is_normalized = FALSE),
      num_edata = length(unique(pdata2$e_data[, 1])),
      num_miss_obs = sum(is.na(pdata2$e_data)),
      prop_missing = (sum(is.na(pdata2$e_data)) /
        prod(dim(pdata2$e_data[, -1]))),
      num_samps = ncol(pdata2$e_data[, -1]),
      data_types = NULL,
      batch_info = list(is_bc = FALSE)
    )
  )

  # Make sure all other attributes have not changed.
  expect_equal(
    attr(pdata, 'cnames'),
    attr(pdata2, 'cnames')
  )
  expect_equal(
    attr(pdata, 'meta_info'),
    attr(pdata2, 'meta_info')
  )
  expect_equal(
    attr(pdata, 'group_DF'),
    attr(pdata2, 'group_DF')
  )
  expect_equal(
    attr(pdata, 'filters'),
    attr(pdata2, 'filters')
  )

  # Ensure the number of replaced elements is correct.
  expect_equal(
    sum(pdata2$e_data == 0),
    341
  )

  # Examine the class of the new pdata object.
  expect_s3_class(
    pdata,
    'pepData'
  )

  # Use a threshold
  thresh = quantile(pdata$e_data[,-1], na.rm=TRUE)[2]
  num_thresholded <- sum(pdata$e_data[,-1] < thresh, na.rm = TRUE)

  expect_message(
    pdata2 <- edata_replace(
      omicsData = pdata,
      x = NA,
      y = 0,
      threshold = thresh
    ),
    sprintf("%s values below the threshold %s have been replaced with %s", num_thresholded, thresh, 0)
  )

  expect_error(
    edata_replace(
      omicsData = pdata,
      x = NA,
      y = 0,
      threshold = min(pdata$e_data[,-1], na.rm=TRUE) - 1
    ),
    'outside the range of the data'
  )

  expect_equal(sum(pdata2$e_data == 0), 341 + num_thresholded)

  # Load: proData --------------------------------------------------------------

  load(system.file('testdata',
    'little_prdata.RData',
    package = 'pmartR'
  ))

  # Test: proData --------------------------------------------------------------

  # Generate a proData object.
  prdata <- as.proData(
    e_data = edata,
    f_data = fdata,
    e_meta = emeta,
    edata_cname = 'Reference',
    fdata_cname = 'SampleID',
    emeta_cname = 'PClass'
  )

  # Replace NAs with 0s in prdata.
  expect_message(
    prdata2 <- edata_replace(
      omicsData = prdata,
      x = NA,
      y = 0
    ),
    "234 instances of NA have been replaced with 0"
  )

  # Check that the elements of the data_info attribute are all correct.
  expect_equal(
    attr(prdata2, "data_info"),
    list(
      data_scale_orig = "abundance",
      data_scale = "abundance",
      norm_info = list(is_normalized = FALSE),
      num_edata = length(unique(prdata2$e_data[, 1])),
      num_miss_obs = sum(is.na(prdata2$e_data)),
      prop_missing = (sum(is.na(prdata2$e_data)) /
        prod(dim(prdata2$e_data[, -1]))),
      num_samps = ncol(prdata2$e_data[, -1]),
      data_types = NULL,
      batch_info = list(is_bc = FALSE)
    )
  )

  # Make sure all other attributes have not changed.
  expect_equal(
    attr(prdata, 'cnames'),
    attr(prdata2, 'cnames')
  )
  expect_equal(
    attr(prdata, 'meta_info'),
    attr(prdata2, 'meta_info')
  )
  expect_equal(
    attr(prdata, 'group_DF'),
    attr(prdata2, 'group_DF')
  )
  expect_equal(
    attr(prdata, 'filters'),
    attr(prdata2, 'filters')
  )

  # Ensure the number of replaced elements is correct.
  expect_equal(
    sum(prdata2$e_data == 0),
    234
  )

  # Examine the class of the new prdata object.
  expect_s3_class(
    prdata,
    'proData'
  )

  # Use a threshold
  thresh = quantile(prdata$e_data[,-1], na.rm=TRUE)[2]
  num_thresholded <- sum(prdata$e_data[,-1] < thresh, na.rm = TRUE)

  expect_message(
    prdata2 <- edata_replace(
      omicsData = prdata,
      x = NA,
      y = 0,
      threshold = thresh
    ),
    sprintf("%s values below the threshold %s have been replaced with %s", num_thresholded, thresh, 0)
  )

  expect_equal(sum(prdata2$e_data == 0), 234 + num_thresholded)

  # Test: seqData ------------------------------------------------------
  load(system.file('testdata',
    'little_seqdata.RData',
    package = 'pmartR'
  ))

  # Construct a seqData object with the edata, fdata, and emeta data frames.
  seqdata <- as.seqData(
    e_data = edata,
    f_data = fdata,
    edata_cname = 'ID_REF',
    fdata_cname = 'Samples'
  )

  err_1 <- "omicsData must be of class 'pepData', 'proData', 'metabData', 'lipidData', or 'nmrData'. "

  # Not allowed
  testthat::expect_error(edata_replace(seqdata, 0, NA), err_1) ## error
})