tests/testthat/test-quantization.R

In Jean-Romain/lidR: Airborne LiDAR Data Manipulation and Visualization for Forestry Applications

context("quantization")

x = runif(100, 1000, 2000)
y = data.table::copy(x)
scale = 0.001
offset = 500

las = megaplot

test_that("is.quantized works", {
  expect_true(!is.quantized(x, scale, offset))
})

test_that("Quantization works", {

  quantize(x, scale, offset)

  expect_equal(x, round(y, 3), tolerance = 1e-10)
  expect_equal(count_not_quantized(x, scale, offset), 0L)
  expect_equal(count_not_quantized(y, scale, offset), 100L)
})

test_that("Quantization works on las data", {

  xscale = las@header@PHB$`X scale factor`
  xoffset = las@header@PHB$`X offset`
  yscale = las@header@PHB$`Y scale factor`
  yoffset = las@header@PHB$`Y offset`
  zscale = las@header@PHB$`Z scale factor`
  zoffset = las@header@PHB$`Z offset`

  expect_equal(count_not_quantized(las$X, xscale, xoffset), 0L)
  expect_equal(count_not_quantized(las$Y, yscale, yoffset), 0L)
  expect_equal(count_not_quantized(las$Z, zscale, zoffset), 0L)

  x = las@data[["X"]][1]
  las@data[["X"]][1] = x + 0.0001

  expect_equal(count_not_quantized(las$X, xscale, xoffset), 1L)

  las_quantize(las)

  expect_equal(las@data[["X"]][1], x,  tolerance = 0)
})

test_that("Quantization works with out of 32 bits integer range values", {

  expect_error(quantize(1234567890.1, 0.01, 0), "'x' contains unquantizable values out of the storable range.")
  expect_equal(count_not_quantized(1234567890.1, 0.01, 0), 1L)
})