error in segment_trees()

[Jean-Romain]

Discussed in https://github.com/r-lidar/lidR/discussions/693

^{Originally posted by **taeyoonlee87** July 24, 2023} Hello, I tried to segment individual trees. I created CHM using 'rasterize_canopy' and created the tree tops using 'locate_trees'. I applied `st_as_sf()` to convert tree tops to SF objects. Like below ```r opt_output_files(las_cat) <- "" ttops <- locate_trees(las_cat, lmf(5), uniqueness = "bitmerge") ttops2 <- st_as_sf(ttops) ``` Then, I segment it using `segement_trees()` ```r opt_output_files(las_cat) <- "/scratch/tl96575/mapping/chm/segmented_{XLEFT}_{YBOTTOM}" algo <- silva2016(chm, ttops2) ctg_segmented <- segment_trees(las_cat, algo) ``` but it resulted in the error message ```r #Chunk 1 of 635 (0.2%): state ✗ #Error: On-disk rasters not supported in 'raster_as_dataframe()' ``` When I test my code with a small portion of data, it worked without errors. But when I ran my code for the whole dataset, it gave me this error. Did you ever see this error? Or could you please advise me on what point I need to check.?

[taeyoonlee87]

want to update you.
When I run my code for the whole dataset (625 las files), it gave me an error

> crown_metrics(ctg_segmented, NULL, geom = "convex")
Error in lazyeval::f_capture(func) : Promise has already been forced
Calls: crown_metrics -> crown_metrics.LAScatalog -> <Anonymous>

When I ran my code for partial data (20~30 las files), it gave the results that I wanted.

However, I also found that some las files were missing.

Do you think does it relate to the bug?

Bests,

[Jean-Romain]

Not related at all for sure. Open a new issue with a reproducible exemple

[Jean-Romain]

MRE

library(lidR)
LASfile <- system.file("extdata", "MixedConifer.laz", package="lidR")
las <- readLAS(LASfile)
col <- pastel.colors(200)

chm <- rasterize_canopy(las, res = 0.5, p2r(0.3))
f = tempfile(fileext = ".tif")
terra::writeRaster(chm, f)
chm = terra::rast(f)

ttops <- locate_trees(chm, lmf(4, 2))
las   <- segment_trees(las, silva2016(chm, ttops))

[Jean-Romain]

Fixed

[taeyoonlee87]

Thanks !!

Do I need to update the package to apply it?

[Jean-Romain]

You need to install version 4.0.4 from github. It has not been released on CRAN yet