[starlogo-users] Random seed

[Andrew Begel]

This is disturbing. There are only two random number generators in 
StarLogo, both controllable by set-random-seed and 
set-scheduler-random-seed (only active if you're using the randomized 
scheduler algorithm). Otherwise, as long as your simulation does not 
rely on any human intervention (e.g. pushing/popping buttons other than 
to initially start it up), it should run deterministically.

Can you send us a copy of your project that doesn't work with 
instructions on how to run it with two different seeds? And what to 
look for to see that the results are not the same?

Thanks,

Andrew

On Aug 24, 2004, at 3:13 PM, C. Athena Aktipis wrote:

> I'm trying to run several simulations with the same random seed and I 
> find
> that they do not run exactly the same even when I set the random seeds 
> to
> identical numbers (and I've verified that the seeds are actually 
> getting
> set).  Are there other factors that can effect the random actions of 
> the
> turtles (their movement or how their variables are initially set) 
> aside from
> the random seed?
>   I appreciate any insight anyone might have on this topic.
> Thank you,
> C.Athena Aktipis
>
> _______________________________________________
> starlogo-users mailing list
> starlogo-users at media.mit.edu
> http://education.mit.edu/mailman/listinfo/starlogo-users