tests/testthat/test-estimateColorMixture.R

In soilDB: Soil Database Interface

context("estimating color mixtures")

## mixing done in CIE LAB by default now


## example data
# http://ncss-tech.github.io/AQP/soilDB/mixing-soil-color-data.html
colors <- c('10YR 5/6', '7.5YR 6/2')
weights <- c(80, 20)

# re-make the data expected by estimateColorMixture()
x <- data.frame(
  parseMunsell(colors, convertColors=FALSE),
  parseMunsell(colors, return_triplets=TRUE, returnLAB=TRUE),
  pct = weights,
  col = parseMunsell(colors, convertColors=TRUE)
)

names(x)[1:3] <- paste0('color', names(x)[1:3])
x$id <- 1
x$colormoistst <- "Dry"


# setting a baseline here, there is nothing behind the results apart from anecdotal evidence
test_that("estimateColorMixture basic functionality", {

  skip_if_not_installed("farver")

  # there is no back-transformation to Munsell by default, request it
  res <- estimateColorMixture(x, backTransform = TRUE)

  # should be a single row
  expect_equal(nrow(res), 1)

  # ensure values aren't NA
  expect_true(all(! sapply(res, is.na)))

  # check sRGB coordinates
  # using wide tolerance, because changes in the Munsell LUT can create errors
  expect_equal(res$r, 0.6, tolerance = 0.1)
  expect_equal(res$g, 0.4, tolerance = 0.1)
  expect_equal(res$b, 0.2, tolerance = 0.1)

})

## TODO: many more examples of colors + resulting mixtures

## TODO: test mixing in the absence of color percentages