Re: Cluster analysis for beginners

"illywhacker" <illywacker@xxxxxxxxx> wrote in
news:1175183941.771746.316260@xxxxxxxxxxxxxxxxxxxxxxxxxxxx:



http://bayes.wustl.edu/etj/articles/confidence.pdf

Check it out, and the Pratt and Savage references therein.


I did read it. Parts of it anyway. I like Jaynes. I started my
undergraduate work in physics. I could not find any example in that chapter
that seemed on point for this problem.

I am having a very difficult time imagining that Jaynes would have offered
the advice that you gave. He begins by laying out a standard for judging
statistical arguments: If it violates common sense, it's probably wrong.
The OP is a scientist with a set of observations where he has observed a
"clustering" of proteins bearing motifs in some regions of an ordering of
these proteins by molecular weight. I think it violates common sense to
think we cannot systematically analyze the motif density in a dataset with
molecular weight as the covariate of interest.

Pratt and Savage will need to wait.

--
David Winsemius

.