Services & ResourcesWolfram ForumsMathGroup Archive

[Date Index] [Thread Index] [Author Index]

Re: spurious $Aborted messages. How to track down cause?

• To: mathgroup at smc.vnet.net
• Subject: [mg79524] Re: spurious $Aborted messages. How to track down cause?
• From: Nasser Abbasi <nma at 12000.org>
• Date: Sat, 28 Jul 2007 05:39:17 -0400 (EDT)
• References: <f8cgnd$44d$1@smc.vnet.net>

On Jul 27, 3:17 am, Nasser Abbasi <n... at 12000.org> wrote:
> Hello;
>
> One of the most annoying things I find is that sometimes I get an
> $Aborted message from the kernel during long computation, but no idea
> why and from where. There are no error messages in the 'message'
> window either.
>
> This happens in this context: I use Manipulate, and in my code (which
> Manipulate calls), I do long computation, then once in a while I see
> an $Aborted message show up where the plot should show up.  This
> happens when I set the length of computation to be larger. when 'n' is
> small, this does not seem to happen.
>
> I do not think this is a memory problem.  The code runs correctly, but
> once 'n' becomes large I see this message pop up, and it happens at
> different stages.
>
> Any idea what could be going on? Could Manipulate have some timing
> limitation on how long the kernel is kept busy while the interface is
> waiting for the result back?
>
> I can put link to my notebook which shows the problem.
>
> thanks,
> Nasser

I think I know now what is going on.

When doing

Manipulate[ process[n], {n,1,100,1} ]

It seems that Manipulate has time constrained set up on how long
process[] takes to complete.

This does not seem to be documented anywhere. In the above, when I
make n large, the function process[n] was taking too long to complete,
(since n makes the computation being done by process[] larger) and
Manipulate would $Aborted.

When I changed the above to the following

Manipulate[ TimeConstrained[ process[n], 500],  {n,1,100,1} ]

Now I no longer see these $Aborted messages any more!

It is just a theory, but I see no other reason to explain this.  I
think this is something that should be documented somewhere?

Nasser

• Follow-Ups:
  • Re: Re: spurious $Aborted messages. How to track down cause?
    • From: Brett Champion <brettc@wolfram.com>