Matthias Greiner
>
> I wonder whether someone has written a Stata programme for graphing
> Chernoff's faces. grstar is nice but faces are somewhat
> easier to interpret
> (or not?).
-findit- finds nothing. I have
not heard of anything in Stataland.
Also, I think this would be quite
difficult to program in Stata.
My impression is that these caught a lot
of attention for a few years as a very
original idea, but have not really been adopted
anywhere as a standard method. That's partly chicken
and egg: only if Chernoff faces were implemented,
would people have a chance to play, and I don't
think they have been widely implemented.
A common remark is that they worked
quite well with some non-statistical
audiences, perhaps for poor reasons
(novelty, cutiness, ...), but statistically minded
people tend to be critical of
arbitrariness of coding and difficulties
of interpretation, and to suspect that
e.g. cluster structure, outliers,
etc. are usually best detected in
other ways, say on some 2-D projection(s)
or a tree structure.
Nick
[email protected]
*
* 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/
