Call R from R

It is sometimes useful to perform a computation in a separate R process, without affecting the current R process at all. This packages does exactly that.


It is sometimes useful to perform a computation in a separate R process, without affecting the current R process at all. This packages does exactly that.


source("https://install-github.me/MangoTheCat/callr")

Use r to run an R function in a new child process. The results are passed back seemlessly:

r(function() var(iris[, 1:4]))
 
#>              Sepal.Length Sepal.Width Petal.Length Petal.Width
#> Sepal.Length    0.6856935  -0.0424340    1.2743154   0.5162707
#> Sepal.Width    -0.0424340   0.1899794   -0.3296564  -0.1216394
#> Petal.Length    1.2743154  -0.3296564    3.1162779   1.2956094
#> Petal.Width     0.5162707  -0.1216394    1.2956094   0.5810063

You can pass arguments to the function, just set args to the list of arguments. This is often necessary as these arguments are explicitly passed to the child process, whereas the evaluated function cannot refer to variables in the parent. For the example the following does not work:

mycars <- cars
r(function() summary(mycars))
 
#> Error in summary(mycars) (from internal.R#90) : object 'mycars' not found

But this does:

r(function(x) summary(x), args = list(mycars))
 
#>     speed           dist
#> Min.   : 4.0   Min.   :  2.00
#> 1st Qu.:12.0   1st Qu.: 26.00
#> Median :15.0   Median : 36.00
#> Mean   :15.4   Mean   : 42.98
#> 3rd Qu.:19.0   3rd Qu.: 56.00
#> Max.   :25.0   Max.   :120.00

Note that the arguments will be serialized and save to a file, so if they are large R objects, then it might take a long time for the child process to start up.

You can use any R package in the child process, just make sure that you refer to it explicitly with the :: operator. For example the following code creates an igraph graph in the child, and calculates some metrics of it.

r(function() { g <- igraph::sample_gnp(1000, 4/1000); igraph::diameter(g) })
 
#> 12

callr provides three ways to handle errors that happen in the child process. The default is to forward any errors to the parent:

r(function() 1 + "A")
#> Error in 1 + "A" : non-numeric argument to binary operator

You can catch these error on the parent, but the context is of course lost. To get the context, you need to specify the error = "stack" option. This copies the whole stack to the parent on an error. The stack is part of the error object thrown on the parent, and you can catch it with tryCatch, and examine it. Here is an example:

tryCatch(
  r(function() { f <- function() g(); g <- function() 1 + "A"; f() },
    error = "stack"),
  error = function(e) print(e$stack)
)
 
#> $`(function () \n{\n    f <- function() g()\n    g <- function() 1 + "A"\n    f()`
#> <environment: 0x7fc1e4b61e08>
#>
#> $`#2: f()`
#> <environment: 0x7fc1e4b62150>
#>
#> $`#2: g()`
#> <environment: 0x7fc1e4b62188>
#>
#> attr(,"error.message")
#> [1] "non-numeric argument to binary operator"
#> attr(,"class")
#> [1] "dump.frames"

The third possible value for error is "debugger" which starts a debugger (see ?debugger in the call stack returned from the child:

r(function() { f <- function() g(); g <- function() 1 + "A"; f() },
  error = "debugger")
 
#> Message:  non-numeric argument to binary operator
#> Available environments had calls:
#> 1: (function ()
#> {
#>     f <- function() g()
#>     g <- function() 1 + "A"
#>     f()
#> 2: #1: f()
#> 3: #1: g()
#>
#> Enter an environment number, or 0 to exit  Selection:

By default the standard output and error of the child is lost, but you can request callr to redirect them to files, and then inspect the files in the parent:

x <- r(function() { print("hello world!"); message("hello again!") },
  stdout = "/tmp/out", stderr = "/tmp/err"
)
readLines("/tmp/out")
 
#> [1] "[1] \"hello world!\""
 
readLines("/tmp/err")
 
#> [1] "hello again!"

With the stdout option, the standard output is collected and can be examined once the child process finished. The show = TRUE options will also show the output of the child, as it is printed, on the console of the parent.

It is good practice to create an anonymous function for the r() call, instead of passing a function from a package to r() directly. This is because callr resets the environment of the function, and some functions will not work then. Here is an example:

r(praise::praise)
 
#> Error: could not find function "is_template"

But with an anonymous function this works fine:

r(function() praise::praise())
 
#> [1] "You are outstanding!"

The rcmd() function calls an R CMD command. For example you can call R CMD INSTALL, R CMD check or R CMD config this way:

rcmd("config", "CC")
 
#>$stdout
#>[1] "clang\n"
#>
#>$stderr
#>[1] ""
#>
#>$status
#>[1] 0

It returns a list with three components: the standard output, the standard error and the exit (status) code of the R CMD command.

MIT © Mango Solutions

News

1.0.0

First public release.

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("callr")

1.0.0 by Gábor Csárdi, a year ago


https://github.com/MangoTheCat/callr


Report a bug at https://github.com/MangoTheCat/callr/issues


Browse source code at https://github.com/cran/callr


Authors: Gábor Csárdi


Documentation:   PDF Manual  


MIT + file LICENSE license


Suggests covr, testthat


Imported by cyclocomp, document, rcmdcheck, reprex, rhub.

Suggested by MonetDBLite.


See at CRAN