Croping a stars object takes 16.89 hours while terra takes 0.16 seconds

[alexyshr]

Croping a stars object takes 16.89 hours while terra takes 0.16 seconds. I just installed sf and stars from GitHub.

library(stars)
#> Loading required package: abind
#> Loading required package: sf
#> Linking to GEOS 3.12.1, GDAL 3.8.4, PROJ 9.3.1; sf_use_s2() is TRUE

library(terra)
#> terra 1.7.78

library(sf)
myurl = "http://geocorp.co/wind/MWP_2017242_090W040N_3D3OT.tif"
the_40n_file = "./MWP_2017242_090W040N_3D3OT.tif"
download.file(myurl, the_40n_file, mode = "wb")

myurl = "http://geocorp.co/wind/MWP_2017242_090W030N_3D3OT.tif"
the_30n_file = "./MWP_2017242_090W030N_3D3OT.tif"
download.file(myurl, the_30n_file, mode = "wb")

myurl = "http://geocorp.co/wind/counties_florida.zip"
download.file(myurl, "counties_florida.zip", mode = "wb")
unzip("counties_florida.zip")
the_florida_file = "./counties_florida.shp"
#read the shapefile
sf_florida <- sf::st_read(the_florida_file, quiet = TRUE)
#transform st_florida to 4326
sf_florida <- sf::st_transform(sf_florida, 4326)

#read th files with stars
st_30n <- stars::read_stars(the_30n_file, quiet = TRUE, proxy = FALSE)
st_40n <- stars::read_stars(the_40n_file, quiet = TRUE, proxy = FALSE)

#convert the files to terra
r_30n <- terra::rast(the_30n_file)
r_40n <- terra::rast(the_40n_file)

#mosaic the files the files with stars
system.time({
  st_mosaic <- stars::st_mosaic(st_30n, st_40n)
})
#>    user  system elapsed 
#>    2.46    0.68    3.73

#plot st_mosaic
plot(st_mosaic)
#> downsample set to 10

#mosaic the files the files with terra
system.time({
  r_mosaic <- terra::mosaic(r_30n, r_40n)
})
#>    user  system elapsed 
#>    1.57    1.39    5.19

#crop the files with stars
#it is taking 16.89 hours
#system.time({
#  st_crop <- sf::st_crop(st_mosaic, sf_florida)
#})
#system.time({
#  st_crop <- st_mosaic[sf_florida, as_points = T]
#})
#    user   system  elapsed 
#21499.75 39246.67 60792.59 
#

#crop the files with terra
system.time({
  r_crop <- terra::crop(r_mosaic, sf_florida)
})
#>    user  system elapsed 
#>    0.11    0.05    0.16

plot(r_crop)

^{Created on 2024-07-04 with reprex v2.1.0}

Session info

sessionInfo()
#> R version 4.4.1 (2024-06-14 ucrt)
#> Platform: x86_64-w64-mingw32/x64
#> Running under: Windows 10 x64 (build 19045)
#> 
#> Matrix products: default
#> 
#> 
#> locale:
#> [1] LC_COLLATE=English_United States.utf8 
#> [2] LC_CTYPE=English_United States.utf8   
#> [3] LC_MONETARY=English_United States.utf8
#> [4] LC_NUMERIC=C                          
#> [5] LC_TIME=English_United States.utf8    
#> 
#> time zone: America/New_York
#> tzcode source: internal
#> 
#> attached base packages:
#> [1] stats     graphics  grDevices utils     datasets  methods   base     
#> 
#> other attached packages:
#> [1] terra_1.7-78 stars_0.6-6  sf_1.0-17    abind_1.4-5 
#> 
#> loaded via a namespace (and not attached):
#>  [1] vctrs_0.6.5        cli_3.6.3          knitr_1.47         rlang_1.1.4       
#>  [5] xfun_0.45          highr_0.11         KernSmooth_2.23-24 DBI_1.2.3         
#>  [9] purrr_1.0.2        styler_1.10.3      glue_1.7.0         htmltools_0.5.8.1 
#> [13] e1071_1.7-14       rmarkdown_2.27     grid_4.4.1         R.cache_0.16.0    
#> [17] evaluate_0.24.0    classInt_0.4-10    fastmap_1.2.0      yaml_2.3.8        
#> [21] lifecycle_1.0.4    compiler_4.4.1     codetools_0.2-20   fs_1.6.4          
#> [25] Rcpp_1.0.12        R.oo_1.26.0        R.utils_2.12.3     digest_0.6.36     
#> [29] class_7.3-22       reprex_2.1.0       curl_5.2.1         parallel_4.4.1    
#> [33] magrittr_2.0.3     R.methodsS3_1.8.2  tools_4.4.1        withr_3.0.0       
#> [37] proxy_0.4-27       xml2_1.3.6         units_0.8-5

[kadyb]

Did you try cutting to the bounding box? Then it should work fast. See also this issue.

st_crop <- st_crop(st_mosaic, st_bbox(sf_florida))

[alexyshr]

Yes! It is taking 0.12 secs (faster than 0.17 with terra)

Thank you!

[alexyshr]

In my initial code, I was comparing two different types of crops. So the time comparison was not valid. If we compare the crop to the bounding box, stars is 1.41 times faster.