[R] can R scripts detect signals sent by the task scheduler ?

[Barry Rowlingson]

On Fri, Nov 20, 2009 at 1:25 PM,  <mauede at alice.it> wrote:
> In general, is it possible to run R scripts through cron jobs ?

 Yes, the only problem might be if you use anything that needs a
graphics window. In the old days you needed an X11 display to create
png graphics with the png() function, but not any more. I'm not sure
if yo need an X11 system for anything apart from showing graphs. But I
don't know everything.

> Is it possible to make  the script detect the system interrupt, save its current status and then exit so that next time it is rescheduled it can pick up from where it left ?

?Signals:

Interrupting Execution of R
Description:
     On receiving ‘SIGUSR1’ R will save the workspace and quit.
     ‘SIGUSR2’ has the same result except that the ‘.Last’ function and
     ‘on.exit’ expressions will not be called.

 So if you can send that signal to your process you might be in luck.
However if by 'system interrupt' you mean the SIGQUIT signal, then I'm
not sure that R can trap interrupts to that extent. You might want to
see if your operating system supports 'checkpointing', in which case
any process can be saved and restarted.

 If by 'system interrupt' you mean SIGKILL (signal 9) then you are very stuck.

 What is going on in your system? Are you trying to checkpoint if the
machine is shut down?

 Barry