Load or install multiple R packages

This function attempts to load multiple R packages specified by the user. If a package is not installed, the function can optionally install it before loading. It also provides an option to print the names and versions of the loaded packages.

Usage

load_packages(
  ...,
  package_list = NULL,
  verbose = FALSE,
  install_missing = FALSE,
  n_cpus = getOption("Ncpus", 1L)
)

Arguments

...

Character. Names of the packages to be loaded or installed.

package_list

Character vector. An alternative or additional way to specify package names as a vector.

verbose

Logical. If TRUE, prints the names and versions of the loaded packages. Defaults to FALSE.

install_missing

Logical. If TRUE, missing packages are automatically installed and then loaded. Defaults to FALSE.

n_cpus

Integer. Number of CPUs to use for parallel installation of packages. Defaults to the value of the Ncpus option. This is only valid if install_missing is TRUE.

Value

This function is used for its side effects (loading/installing packages) and does not return any value.

Details

On HPC systems using network filesystems (e.g. Lustre) with renv, utils::installed.packages() can be slow due to high metadata I/O latency across large library trees. This function mitigates that by calling installed.packages() once with fields = "Version" only, caching the result, and reading package versions from the cached matrix rather than re-reading individual DESCRIPTION files via packageDescription().

Author

Ahmed El-Gabbas

Examples

# Currently loaded packages
(P1 <- ecokit::loaded_packages())
#>  [1] "qs2"       "stringr"   "tools"     "future"    "car"       "carData"  
#>  [7] "purrr"     "archive"   "Hmsc"      "coda"      "rworldmap" "arrow"    
#> [13] "dismo"     "raster"    "sp"        "terra"     "fs"        "tidyr"    
#> [19] "tibble"    "png"       "sf"        "ggplot2"   "dplyr"     "ecokit"   
#> [25] "magrittr"  "stats"     "graphics"  "grDevices" "utils"     "datasets" 
#> [31] "methods"   "base"     

# Load tidyr
load_packages(tidyr, raster, ggplot2, nnet, verbose = TRUE)
#> The following packages were already loaded:
#>  >> ggplot2 (4.0.3)
#>  >> raster (3.6-32)
#>  >> tidyr (1.3.2)
#> Loading packages:
#>  >> nnet (7.3-20)

# Loaded packages after implementing the function
(P2 <- ecokit::loaded_packages())
#>  [1] "nnet"      "qs2"       "stringr"   "tools"     "future"    "car"      
#>  [7] "carData"   "purrr"     "archive"   "Hmsc"      "coda"      "rworldmap"
#> [13] "arrow"     "dismo"     "raster"    "sp"        "terra"     "fs"       
#> [19] "tidyr"     "tibble"    "png"       "sf"        "ggplot2"   "dplyr"    
#> [25] "ecokit"    "magrittr"  "stats"     "graphics"  "grDevices" "utils"    
#> [31] "datasets"  "methods"   "base"     

# Which packages were loaded?
setdiff(P2, P1)
#> [1] "nnet"

# verbose = FALSE (default)
load_packages(tidyterra, verbose = FALSE)

# load already loaded packages
load_packages(tidyr, tidyterra, verbose = TRUE)
#> The following packages were already loaded:
#>  >> tidyr (1.3.2)
#>  >> tidyterra (1.2.0)

# non-existent package
load_packages("non_existent")
#> The following packages are neither available nor installed as install_missing = FALSE:
#>  >>  non_existent