Error occured in "Clip_roi"

[taeyoonlee87]

Hi,

I run the same codes that I used to it. but I encountered an error in my last code.

"Error in bboxes[[i]] : subscript out of bounds"

###READ PACKAGES
library(lidR)
library(rgl)
library(raster)
library(sp)
library(sf)
library(terra)
library(rgdal)
library(future)
#library(rLiDAR)

##use Parallel threads/workers 

####DEFINE YOUR WORKING DIRECTORY
setwd("E:/Lidar_2022/R/trials")
plan(multicore, workers = 20)

las_cat<-readLAScatalog("E:/Lidar_2022/R/trials")

###Creat DTM using tin (res=5m)
opt_output_files(las_cat) <- "E:/Lidar_2022/R/trials/trial_dtm/DTM_{XLEFT}_{YBOTTOM}"
DTM <- rasterize_terrain(las_cat, res = 1, algorithm = tin(), keep_lowest = FALSE)

##Nomalized the DTM .. it will produce Canpoy Surface Model ..
#When it is filtered by 90% of up returns using "filter_poi"  it is a Canopy Hiehgt Model
opt_output_files(las_cat) <-"E:/Lidar_2022/R/trials/trial_dtm/norm_{XLEFT}_{YBOTTOM}" 
lasnorm <- normalize_height(las_cat, DTM)  

#####

#Read shapefile 
Buffer  <- shapefile("E:/Lidar_2022/Data/Boundary/Talladega_Field_Buffer.shp")


##filter  # Ignore points with elevations below 0
opt_filter(lasnorm) <- "-drop_z_below 0"

# filter to remove points above the 95th percentile of height 
filter_noise = function(las, sensitivity)
{
  if (is(las, "LAS"))
  {
    p95 <- pixel_metrics(las, ~quantile(Z, probs = 0.95), 10)
    las <- merge_spatial(las, p95, "p95")
    las <- filter_poi(las, Z < p95*sensitivity)
    las$p95 <- NULL
    return(las)
  }
  
  if (is(las, "LAScatalog"))
  {
    options <- list(
      need_output_file = TRUE)  # Throw an error if no output template is provided
    res <- catalog_map(las, filter_noise, sensitivity = sensitivity, .options = options)
    return(res)
  }
}
opt_output_files(lasnorm) <- "E:/Lidar_2022/R/trials/trial_dtm/denoised_{XLEFT}_{YBOTTOM}"
denoised_lasnorm<-filter_noise(lasnorm, sensitivity = 1.2)

#Clip with buffer 
opt_output_files(denoised_lasnorm) <- "E:/Lidar_2022/R/trials/trial_dtm/plot/plot_{PlotID}"
plot <- clip_roi(denoised_lasnorm, Buffer)

## function for getting the metrics 
routine <- function(chunk){ 
  las <- readLAS(chunk)
  if (is.empty(las)) return(NULL)
  data.frame(cloud_metrics(las, ~stdmetrics(X,Y,Z,Intensity,ReturnNumber,Classification,dz=1, zmin=1)))
}  

#getting the metrics 
opt_output_files(plot) <- "E:/Lidar_2022/R/trials/trial_dtm/plot/{ORIGINALFILENAME}"
catalog_apply(plot, routine)

I think the output of "clip_roi" doesn't include any data.

I checked the coordinate of the shapefile that they used the same coordinate system.

I had no problems when I running this code with same dataset.

Can you please look at it?

las.zip
las.zip

[taeyoonlee87]

Boundary.zip

[Jean-Romain]

Please provide a minimal reproducible exemple

[taeyoonlee87]

Thanks for replying to me.
I attached a shapefile of sampling plots, a las file, and codes that I used above.

[Jean-Romain]

Please see the definition of a MRE. Here you have 8 libraries loaded. How much are actually needed reproduce? Probably 2 or 3. rasterize_terrain and normalize_height are not relevant here since the error seems to be with clip. Is filter_noise really involved here ? Probably not. I see only one call to clip before routine so I guess routine is not relevant here either. So I think a 10 lines script with proper data should allow to reproduce.

[taeyoonlee87]

###READ PACKAGES
library(lidR)

###DEFINE YOUR WORKING DIRECTORY
setwd("E:/Lidar_2022/R/trials")
las_cat<-readLAScatalog("E:/Lidar_2022/R/trials")
#Read shapefile 
Buffer  <- shapefile("E:/Lidar_2022/Data/Boundary/Talladega_Field_Buffer.shp")

#Clip with buffer 
opt_output_files(las_cat) <- "E:/Lidar_2022/R/trials/trial_dtm/plot/plot_{PlotID}"
plot <- clip_roi(las_cat, Buffer)

## function for getting the metrics 
routine <- function(chunk){ 
  las <- readLAS(chunk)
  if (is.empty(las)) return(NULL)
  data.frame(cloud_metrics(las, ~stdmetrics(X,Y,Z,Intensity,ReturnNumber,Classification,dz=1, zmin=1)))
}  
#getting the metrics 
opt_output_files(plot) <- "E:/Lidar_2022/R/trials/trial_dtm/plot/{ORIGINALFILENAME}"
catalog_apply(plot, routine)

[taeyoonlee87]

Oh, I am sorry,

Yes you are right.

THe error is clearly related to the "Clip_roi".

To check this, I tested with/without the codes for filtering, normalized_hiehgt, and raterize_terrain.

Therefore, the codes are

###READ PACKAGES
library(lidR)

###DEFINE YOUR WORKING DIRECTORY
setwd("E:/Lidar_2022/R/trials")
las_cat<-readLAScatalog("E:/Lidar_2022/R/trials")
#Read shapefile 
Buffer  <- shapefile("E:/Lidar_2022/Data/Boundary/Talladega_Field_Buffer.shp")

#Clip with buffer 
opt_output_files(las_cat) <- "E:/Lidar_2022/R/trials/trial_dtm/plot/plot_{PlotID}"
plot <- clip_roi(las_cat, Buffer)

## function for getting the metrics 
routine <- function(chunk){ 
  las <- readLAS(chunk)
  if (is.empty(las)) return(NULL)
  data.frame(cloud_metrics(las, ~stdmetrics(X,Y,Z,Intensity,ReturnNumber,Classification,dz=1, zmin=1)))
}  
#getting the metrics 
opt_output_files(plot) <- "E:/Lidar_2022/R/trials/trial_dtm/plot/{ORIGINALFILENAME}"
catalog_apply(plot, routine)

And getting error "Error in bboxes[[i]] : subscript out of bounds"
I checked the coordinate system for two files but those used same coordinates.
and I used this code previously without this problems.

Thanks again