Re: [Swarm-Modelling] Non-integer timesteps

[Paul Johnson]

This is not an easy one, but it has been done.

The person that can show exactly how to do this is Doug Donalson, an 
ecologist.  He was doing that kind of scheduling when I started with 
Swarm before the turn of the century.  At the time, I could not quite 
understand it, but my recollection is that it worked basically like 
this.  Take a floating point number, say

7.124

and look at it as two integers 7 and 124.  Then build a schedule in 
which, for each integer valued timestep, there is an integer valued 
schedule inside it. So if you want something at 7.124, you grab the 
schedule at time 7, and then throw something at time 124 inside that 
schedule.

I'm sorry I don't have a working code example.  I never did have a full 
working model showing this, but Doug had posted a snippit to the swarm 
list that showed how it was done. But can't find it. I've changed 
computers 2 or 3 times since then, I suppose.  But I'm googling right 
now and I see he made a Swarmfest presentation on this in  2001

http://www.swarm.org/pipermail/modelling/2001-March/002843.html

I know for sure that Dough described it to the support email list, but 
without a search engine, I can't find it.

http://www.swarm.org/pipermail/support/

pj

address@hidden wrote:
> Hello,
>
> I sent a message to the support mailing list a few days ago asking about
> asynchronous schedules.
>
> I have a much more specific question now: is it possible to schedule
> events at non-integer timesteps in swarm? The schedule method
> at$createActionTo$message only accepts integer values... I know I could
> just convert everything to integers and forget about what shows up in the
> x-axis of the EZgraph, but that would complicate things and cost a lot of
> computation time. Is there some simple way to do this?
>
> Thank you,
>
> Dinis Gökaydin
>
> P.S.-Thank you for the answers the last time, they were very helpful, I've
> studied the Mousetrap code and I think I understand roughly how dynamic
> schedule works.
>
> _______________________________________________
> Modelling mailing list
> address@hidden
> http://www.swarm.org/mailman/listinfo/modelling


--
Paul E. Johnson                       email: address@hidden
Dept. of Political Science            http://lark.cc.ku.edu/~pauljohn
1541 Lilac Lane, Rm 504
University of Kansas                  Office: (785) 864-9086
Lawrence, Kansas 66044-3177           FAX: (785) 864-5700