Re: Polarized distributions...

[Paul E. Johnson]

Darren Schreiber wrote:
> 
> I have been running my voting model simulations assuming a uniform
> distribution of voters across the issue space.  I am going to do some more
> runs looking at the impact of using a normal distribution (easy enough to
> change the distribution in Swarm.)  But, I also want to represent a
> randomly generated electorate with a distribution that is polarized -- very
> few voters in the middle and increasing numbers towards the extreme.  I was
> wondering if anyone had an easy way to generate such a distribution.
> 
> Flipping through William Greene's Econometrics text I note that the Beta
> distribution can be U-shaped (which is what I am looking for).  The text
> does not say what parameters would create that shape, and I am not sure how
> to write a distribution object in Swarm.
> 
> Thanks in advance...
> 
>         Darren
>
By an incredible coincidence, I started working on the random part of
the Swarm User Guide and wrote down how to get a beta variate.  There is
a reference in there to a great book by law and kelton (Simulation
Modeling and Analysis, New York: McGraw Hill.) that has pictures of all
different kinds of shapes the beta can have, depending on the values of
the two coefficients.

Here's how to get a beta:
http://lark.cc.ukans.edu/~pauljohn/Swarm/Beta/SwarmUserGuide/x2220.html

-- 
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


                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================