Jeroen Jeroen - 8 days ago 6
R Question

Save traceback on error using tryCatch

When calling an R function inside my webapp, I would like to catch the stack-trace when an error occurs and present it to the user for debugging purposes. Something like the output of

traceback()
in an interactive session. However,
traceback
doesn't seem to work when it is called inside the error handler, it returns
No traceback available
:

f <- function() {
g <- function() stop("test traceback")
g()
}

errhandler <- function(e){
stacktrace <- traceback()
unlist(stacktrace);
}

out <- tryCatch(f(), error=errhandler)


Is there any way I can programatically catch the stack trace of an error? I.e. get the output that I would get when calling
traceback()
manually after the error:

f()
traceback()

Answer

It turns out that the latest version of the evaluate package has a function called try_capture_stack which is a pretty good implementation of this.