Skip to contents

Simple with functions. Similar to or from withr implementations.

Usage

gwith_options(new, code)

gwith_package(package, code, pos = 2L)

gwith_seed(seed = 1234, code, ..., substitute = TRUE, envir = parent.frame())

Arguments

new

new option to set and its value

code

R code to execute with temporary settings

package

character. The name of a package

pos

integer. position in search to attach at. Default = 2

...

additional params to pass. See details.

substitute

If TRUE, argument expr is substitute():ed, otherwise not.

envir

The environment in which the expression should be evaluated.

Details

gwith_seed() : ... passes to set.seed()

Functions

  • gwith_options(): Eval with temporary option setting

  • gwith_package(): Eval with temporarily attached package

  • gwith_seed(): Eval with temporary specifiable seed

Examples

# options ###################################
gwith_options(list(gutils.temp = "found"), {
    print(getOption("gutils.temp", default = "not_found"))
})
#> [1] "found"
getOption("gutils.temp", default = "not_found")
#> [1] "not_found"

# packages ##################################
# temporarily attach at end of search (right before "base")
gwith_package("data.table", print(search()), pos = length(search()))
#>  [1] ".GlobalEnv"          "package:GiottoUtils" "package:stats"      
#>  [4] "package:graphics"    "package:grDevices"   "package:utils"      
#>  [7] "package:datasets"    "package:methods"     "Autoloads"          
#> [10] "package:data.table"  "package:base"       
search()
#>  [1] ".GlobalEnv"          "package:GiottoUtils" "package:stats"      
#>  [4] "package:graphics"    "package:grDevices"   "package:utils"      
#>  [7] "package:datasets"    "package:methods"     "Autoloads"          
#> [10] "package:base"       

# seed  #####################################
start_seed <- .Random.seed

# identical generation
a <- gwith_seed(runif(10), seed = 1234)
b <- gwith_seed(runif(10), seed = 1234)
identical(a, b)
#> [1] TRUE

# does not alter pre-existing seed
end_seed <- .Random.seed
identical(start_seed, end_seed)
#> [1] TRUE