Welcome to Statlist, Jens. We corresponded some years ago about your
vennerable Venn program regarding the same question - how to make the
circles proportional and the overlaps reflecting the true intersected
areas. (We also corresponded later on early iterations of the
excellent EpiData project that you have lead ( http://www.epidata.dk/
))
At that time I pointed out that my partner and I had developed a
method that did the trick for a paper we were working on, but in the
end it required using an approximation with numerical analysis. I
never did find a way to simultaneously calculate the relative location
of the circle centres such that the area of the intersections is
exact. I should note, also, that when one of the sets approaches
100%, that a rectangular universe will not work, and that one should
actually start with a circular universe to be able to handle all
two-way and three-way sets.
DCE
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
