Newsgroups: comp.lang.apl
Path: watmath!watserv2.uwaterloo.ca!torn!cs.utexas.edu!uunet!caen!uakari.primate.wisc.edu!usenet.coe.montana.edu!oususalg
From: Lou Glassy (fortran@giac1.oscs.montana.edu)
Subject: Use of J in Geostatistics
Message-ID: <1992Oct14.173052.5924@coe.montana.edu>
Originator: oususalg@nero.cs.montana.edu
Sender: usenet@coe.montana.edu (USENET News System)
Organization: Montana State University, Bozeman Montana USA
Date: Wed, 14 Oct 1992 17:30:52 GMT
Lines: 30

A question or two for the J community.  Do you know of anyone using J
in a geostatistics?  On a related vein, how would you J-users /
statisticians out there compare J as a statistical tool, relative to
ready-made stat packages (SAS, BMDP, SPSS, MiniTab...).  Clearly, J
doesn't have the extensive libraries of stat functions these other
packages have, but if one could generate the functions one needed...
what then?  How would you rate J vs these other tools?

Hamming's saying 'The purpose of computing is insight, not numbers'
comes to mind.  My experience thus far tells me J is a handy
'insight-tool'.

Specifically, there are some aspects of data in a spatial context that
require a slightly different treatment than standard classical
statistics would admit.  J's way of handling data in a uniform way (via
the rank operator, and other things besides) seems superbly suited for
the kinds of analyses I'd like to do.

When I play with J's capabilities for munging large aggregates of data
in a functional way, I am struck by the potential it (J) may have for
reduction of geographical data.

The adventure of learning J continues.  A Large-Thank-You goes to ISI
for making J available, and to Dr. Keith Smillie, whose document
'Statistics and J' was just the thing I needed to get going with J.

Lou.

-- 
Lou Glassy (oususalg@cs.montana.edu)                  C Delenda Est
