R-bloggers

Parse arguments of an R script

(This article was first published on Gregor Gorjanc (gg), and kindly contributed to R-bloggers)

R can be used also as a scripting tool. We just need to add shebang in the first line of a file (script):

#!/usr/bin/Rscript

and then the R code should follow.

Often we want to pass arguments to such a script, which can be collected in the script by the commandArgs() function. Then we need to parse the arguments and conditional on them do something. I came with a rather general way of parsing these arguments using simply these few lines:
## Collect arguments
args <- commandArgs(TRUE)
 
## Default setting when no arguments passed
if(length(args) < 1) {
args <- c("--help")
}
 
## Help section
if("--help" %in% args) {
cat("
The R Script
 
Arguments:
--arg1=someValue - numeric, blah blah
--arg2=someValue - character, blah blah
--arg3=someValue - logical, blah blah
--help - print this text
 
Example:
./test.R --arg1=1 --arg2="
output.txt" --arg3=TRUE \n\n")
 
q(save="no")
}
 
## Parse arguments (we expect the form --arg=value)
parseArgs <- function(x) strsplit(sub("^--", "", x), "=")
argsDF <- as.data.frame(do.call("rbind", parseArgs(args)))
argsL <- as.list(as.character(argsDF$V2))
names(argsL) <- argsDF$V1
 
## Arg1 default
if(is.null(args$arg1)) {
## do something
}
 
## Arg2 default
if(is.null(args$arg2)) {
## do something
}
 
## Arg3 default
if(is.null(args$arg3)) {
## do something
}
 
## ... your code here ...

It is some work, but I find it pretty neat and use it for quite a while now. I do wonder what others have come up for this task. I hope I did not miss some very general solution.

To leave a comment for the author, please follow the link and comment on their blog: Gregor Gorjanc (gg).

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...