normalize_intensity - error with LAScatalog

[lucas-johnson]

Current Behavior

normalize_intensity works with a single las file, but fails with a las catalog. The error indicates that the algorithm provided (range_correction(sensor_positions, Rs = 2000)) is invalid, but it is the same for the single las file and the las catalog.

normalize_intensity(readLAS("~/Downloads/some_las/u_5550087800_2015.las"), range_correction(sensor_positions, Rs = 2000))
#> class        : LAS (v1.4 format 6)
#> memory       : 1.2 Gb 
#> extent       : 555000, 556500, 4878000, 4879500 (xmin, xmax, ymin, ymax)
#> coord. ref.  : +proj=utm +zone=18 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs 
#> area         : 2.25 km²
#> points       : 12.96 million points
#> density      : 5.76 points/m²
normalize_intensity(ctg, range_correction(sensor_positions, Rs = 2000))
#> Processing [=>------------------------------------------------------------------------------]   3% (1/38) eta:  2m
#> Error: Invalid function provided as algorithm.
plot(ctg)
range_correction(sensor_positions, Rs = 2000)
#> Object of class lidR algorithm
#> Algorithm for: intensity normalisation 
#> Designed to be used with: normalize_intensity 
#> Native C++ parallelization: no 
#> Parameters: 
#>  - elevation = Z <character>
#>  - f = 2.3 <numeric>
#>  - gpstime = gpstime <character>
#>  - Rs = 2000 <numeric>
#>  - sensor <SpatialPointsDataFrame>

Desired Behavior

normalize_intensity() should work on a las catalog just like it works on a single las file. Or, since the algorithm used in both cases is the same, the error should describe a different issue (perhaps with my las catalog).

[Jean-Romain]

Minimal reproducible example

LASfile <- system.file("extdata", "Topography.laz", package="lidR")
las <- readLAS(LASfile)
sensor <- track_sensor(las, Roussel2020(pmin = 15))

ctg = readLAScatalog(LASfile)
opt_output_files(ctg) <- tempfile()
ctg <- normalize_intensity(ctg, range_correction(sensor, Rs = 2000))
#> Erreur : Invalid function provided as algorithm.

[Jean-Romain]

After investigation, what is received is an Extent from raster and the method normalize_intensity.LAScluster is by-passed. This is weird. I'll investigate later. This will be fixed on Monday.

[lucas-johnson]

Thank you for looking into it so quickly! Looking forward to hearing back from you.

[Jean-Romain]

By the way, the methods does not require a buffer around the processed regions. Unless you want to process small chunks, you can write a simple for loop to process each file independently. Just saying in case you want work today. You are not necessarily limited by this bug.

[lucas-johnson]

Yes - thanks for the advice. I've written my own .LAScatalog and .LAScluster functions to work with catalog_apply(). This is working fine for now.

[Jean-Romain]

Fixed. Thank you for reporting