Re: [Swarm-Modelling] status_access_violation question

[Paul E. Johnson]

Use gdb.

pj

Steve Railsback wrote:

> address@hidden wrote:
>
> > Hello,
> >
> > I am currently in the midst of coding my first involved swarm model,
> > which I am going to use in my dissertation research on early hominid
> > food sharing.  I've just run across a problem that I'm unable to solve
> > with my limited experience.  After I compile the executable (share.exe)
> > and run it at the command prompt, the simulation runs successfully for
> > anywhere from 50 to 700 time steps.  However, though it runs as it
> > should, the model always bails with the following error message at some
> > (seemingly random) point:
> >
> > 0 [main] share 1000 handle_exceptions: Exception:
> > STATUS_ACCESS_VIOLATION
>
>
> Yeah, these kind of errors are not so clear to find. In general, it 
> means you are trying to use memory that isn't allocated--trying to 
> store something in an array at a location bigger than the array size, 
> trying to access a location in a Grid2D object that's beyond its 
> boundaries, etc. When it happens after a model has run for a while, 
> perhaps one of your agents is wandering off the edge of the world, etc.
>
> You could try putting print statements wherever you use an array or 
> space object, to print out the location you're trying to access before 
> you actually do access it. Check carefully for logic errors that could 
> cause the code to try to read a variable or array location that 
> doesn't exist.
>
> Steve R.


--
Paul E. Johnson                       email: address@hidden
Dept. of Political Science            http://lark.cc.ukans.edu/~pauljohn
University of Kansas                  Office: (785) 864-9086
Lawrence, Kansas 66045                FAX: (785) 864-5700