tests/testthat/test-cc_brightness.R

In rorynolan/nandb: Number and Brightness Image Analysis

test_that("cc_brightness() works", {
  img <- ijtiff::read_tif(system.file("extdata", "two_ch.tif",
    package = "nandb"
  ), msg = FALSE)
  set.seed(1)
  cc_b <- cc_brightness(img, thresh = "Huang", detrend = TRUE, filt = "median")
  expect_equal(median(cc_b, na.rm = TRUE), 0.014, tolerance = 0.2)
  cc_b <- cc_brightness(img, thresh = "Huang", filt = "smooth")
  expect_equal(median(cc_b, na.rm = TRUE), 0.014, tolerance = 0.2)
  expect_error(
    cc_brightness(img, ch2 = 3),
    paste0(
      "You have requested to use channel 3, but your image\\s?",
      "has only 2.+channels in total.+This is not possible.+\\s?",
      "Please retry with valid channel.+numbers."
    )
  )
  expect_error(
    cc_brightness(
      img %T>% {
        .[, , 1, ] <- NA
      }
    ),
    paste0(
      "The first channel is all NAs.+Can't compute on an\\s?",
      "array of all NAs."
    )
  )
  expect_error(
    cc_brightness(
      img %T>% {
        .[, , 2, ] <- NA
      }
    ),
    paste0(
      "The second channel is all NAs.+Can't compute on an\\s?",
      "array of all NAs."
    )
  )
  expect_error(
    cc_brightness(img, thresh = max(img) + 1),
    paste0(
      "After thresholding, the first channel is all\\s?",
      "NAs.+Can't compute on an array of all NAs.+You need to\\s?",
      "choose a less severe threshold for this.+channel."
    )
  )
  expect_error(
    cc_brightness(img, thresh = c(0, max(img) + 1)),
    paste0(
      "After thresholding, the second channel is all\\s?",
      "NAs.+Can't compute on an array of all NAs.+You need to\\s?",
      "choose a less severe threshold for this.+channel."
    )
  )
})

test_that("cc_brightness_timeseries() works", {
  img <- ijtiff::read_tif(system.file("extdata", "two_ch.tif",
    package = "nandb"
  ), msg = FALSE)
  set.seed(1)
  cc_b_ts <- cc_brightness_timeseries(img, 10,
    thresh = "Huang", detrend = TRUE, filt = "median"
  )
  expect_equal(median(cc_b_ts, na.rm = TRUE), 0, tolerance = 0.01)
  cc_b_ts <- cc_brightness_timeseries(img, 10,
    thresh = "Huang", filt = "smooth"
  )
  expect_equal(median(cc_b_ts, na.rm = TRUE), 0.0013, tolerance = 0.003)
  cc_b_ts_overlapped <- cc_brightness_timeseries(img, 10,
    overlap = TRUE,
    thresh = "H", filt = "s"
  )
  common_frames <- which(seq_len(dim(img)[4]) %% 10 == 1)
  expect_equal(
    cc_b_ts_overlapped[, , , common_frames, drop = FALSE] %>% {
      list(dim(.), as.vector(.))
    },
    cc_b_ts %>% {
      list(dim(.), as.vector(.))
    }
  )
  cc_b <- cc_brightness(img)
  cc_b_ts_overlapped <- cc_brightness_timeseries(img, dim(img)[4],
    overlap = TRUE
  )
  expect_equal(
    cc_b_ts_overlapped %>% {
      list(dim(.), as.vector(.))
    },
    cc_b %>% {
      list(dim(.), as.vector(.))
    }
  )
  expect_error(
    cc_brightness_timeseries(img, 9999),
    paste0(
      "You have selected 9999 frames per set, but there are only\\s?",
      "100,.+frames in total.+Please select less than 100 frames\\s?",
      "per set"
    )
  )
  expect_error(
    cc_brightness_timeseries(
      img %T>% {
        .[, , 1, ] <- NA
      }, 20
    ),
    paste0(
      "The first channel is all NAs.+Can't compute on an\\s?",
      "array of all NAs."
    )
  )
  expect_error(
    cc_brightness_timeseries(
      img %T>% {
        .[, , 2, ] <- NA
      }, 20
    ),
    paste0(
      "The second channel is all NAs.+Can't compute on an\\s?",
      "array of all NAs."
    )
  )
  expect_error(
    cc_brightness_timeseries(img, 20, thresh = max(img) + 1),
    paste0(
      "After thresholding, the first channel is all\\s?",
      "NAs.+Can't compute on an array of all NAs.+You need to\\s?",
      "choose a less severe threshold for this.+channel."
    )
  )
  expect_error(
    cc_brightness_timeseries(img, 20, thresh = c(0, max(img) + 1)),
    paste0(
      "After thresholding, the second channel is all\\s?",
      "NAs.+Can't compute on an array of all NAs.+You need to\\s?",
      "choose a less severe threshold for this.+channel."
    )
  )
})

test_that("cc_brightness_folder() works", {
  img <- ijtiff::read_tif(system.file("extdata", "two_ch.tif",
    package = "nandb"
  ), msg = FALSE)
  cwd <- setwd(tempdir())
  on.exit(setwd(cwd))
  ijtiff::write_tif(img, "a.tif", msg = FALSE)
  set.seed(1)
  cc_brightness_folder(thresh = "Huang", detrend = FALSE)
  cc_b <- dir(pattern = "cc_brightness", recursive = TRUE) %>%
    ijtiff::read_tif(msg = FALSE)
  expect_equal(median(cc_b, na.rm = TRUE), 0.02, tolerance = 0.2)
  suppressMessages(filesstrings::dir.remove("cc_brightness"))
  dir.create("dir")
  ijtiff::write_tif(img, "dir/a.tif", msg = FALSE)
  set.seed(1)
  cc_brightness_file("dir/a.tif", thresh = "Huang", detrend = FALSE)
  cc_b <- dir(pattern = "cc_brightness", recursive = TRUE) %>%
    ijtiff::read_tif(msg = FALSE)
  expect_equal(median(cc_b, na.rm = TRUE), 0.02, tolerance = 0.2)
  suppressMessages(filesstrings::dir.remove("dir"))
  file.remove("a.tif")
  setwd(cwd)
})

test_that("cc_brightness_timeseries_folder() works", {
  img <- ijtiff::read_tif(system.file("extdata", "two_ch.tif",
    package = "nandb"
  ), msg = FALSE)
  cwd <- setwd(tempdir())
  on.exit(setwd(cwd))
  ijtiff::write_tif(img, "a.tif", msg = FALSE)
  set.seed(1)
  cc_brightness_timeseries_folder(
    thresh = "Huang", frames_per_set = 10, detrend = FALSE,
    filt = "median", parallel = 2
  )
  cc_b_ts <- c(
    "cc_brightness_contiguous_timeseries",
    "cc_brightness_overlapped_timeseries"
  ) %>%
    purrr::map(~ dir(pattern = paste0(., ".*tif$"), recursive = TRUE)) %>%
    unlist() %>%
    ijtiff::read_tif(msg = FALSE)
  expect_equal(median(cc_b_ts, na.rm = TRUE), 0, tolerance = 0.001)
  suppressMessages(filesstrings::dir.remove("cc_brightness_timeseries"))
  dir.create("dir")
  ijtiff::write_tif(img, "dir/a.tif", msg = FALSE)
  set.seed(1)
  cc_brightness_timeseries_file("dir/a.tif",
    thresh = "Huang", detrend = FALSE,
    frames_per_set = 10,
    filt = "median", parallel = 2
  )
  cc_b_ts <- c(
    "cc_brightness_contiguous_timeseries",
    "cc_brightness_overlapped_timeseries"
  ) %>%
    purrr::map(~ dir(pattern = paste0(., ".*tif$"), recursive = TRUE)) %>%
    unlist() %>%
    ijtiff::read_tif(msg = FALSE)
  expect_equal(median(cc_b_ts, na.rm = TRUE), 0, tolerance = 0.001)
  suppressMessages(filesstrings::dir.remove("dir", "cc_brightness_timeseries"))
  file.remove(dir(pattern = "tif$"))
  setwd(cwd)
})