I would argue the opposite.  While I agree with Doug that you need good  
RNGs (though not necessarily true RNGs) in order to avoid bias, the  
problem with good pseudo- or true- RNGs is that they have order N^2  
convergence for Monte Carlo simulations.  Quasi-random number generators  
on the other hand (such as multiples of an irrational square root, or a  
Peano tiling) converge in order N.  If you can trust the results, faster  
conergence lets you simulate more.
-Roger

On Sat, 21 Jul 2007 23:18:36 -0600, Douglas Roberts <[EMAIL PROTECTED]>  
wrote:

> Simulations of stochastic processes also require good RN generators,
> especially for simulations of large systems with (I hate to use this  
> word)
> emergent behavioral properties.  A bad RN generator will introduce  
> emergent
> behavior that will be "flavored" by a bad random sequences.
>
>



============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org

Reply via email to