Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: -cluster- dissimilarity measures


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: -cluster- dissimilarity measures
Date   Sun, 17 Nov 2002 17:10:22 -0000

Stas Kolenikov
>
> is there an easy way to code the distance between two
> points on the Earth
> surface along the great-circle arc given their longitudes
> and latitudes? I
> want to cluster a few geographical locations... I know a
> messy way to do
> it -- to go through all 2400 lines of cluster.ado and find
> out where the
> (dis)similarity measures are computed, and make a new one
> that I need. (Is
> this possible with the programmer's extensions to cluster that Stata
> provides for, instead? I looked it up rather quickly, and
> decided it is
> not quite what I need.) My gut feeling though is that it is the
> responsibility of the built-in -_cluster-, so there is not
> much I can do
> to define my own distance function, is there?

Ken Simons has written a -globdist- which may help. It assumes
the planet is a perfect sphere, which isn't true,
but that is a pretty good approximation by e.g.
econometric standards.

Nick
n.j.cox@durham.ac.uk

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index