Re: PLEASE READ: use of random

[Ginger Booth]

Sven,
    As a "heavy user", I'm not sure what you're talking about.  But I'll try....

> a) what is the *average* number of random variates needed for a single run
> of your model 

Hellifiknow.  Not bounded, somewhere in the zillions.  For a modest single
run, 5000 timesteps, maybe 1000 agents avg. population, assuming all of them
have some random component in their decisions (not true, but probably averages
out thataway), let's say on the order of 10 million?

This question would probably make more sense if I had a bounded number of
agents.  Mine are neither bounded nor constant in number.

> (keeping in mind that any call to a distribution that returns
> a double now uses 2 random variates);

Huh? 

> b) what is the *largest* number of random variates you imagine needing for
> a single run?

Well, 2 orders of magnitude more than for a modest run?  A billion?

> c) how many independent runs do *you* consider sufficient for a single
> experiment?

Depends on the experiment.  Somewhere in the 10-1000 range.  Typically 20.
Sometimes you're scouting terrain, sometimes zeroing in on a benchmark....

> d) I also have the following question for y'all regarding PMMLCG:
> Does anyone have strong objections to my fixing this generator in this way?

Fix it.

Sorry my opinions are under-researched....

Ciao,
    Ginger