s.sv - sample semivariogram

[James Darrell McCauley]

There's a new program for calculating sample semivariograms
in GRASS: s.sv (version 0.1B, 15 Oct 1994)
Available at:
  ftp://pasture.ecn.purdue.edu/pub/mccauley/grass/s.sv.tar.gz

Notes: 
1. THIS IS DIFFERENT THAN S.SEMIVAR. 
2. To use the -p flag, you must have g.gnuplot installed 
   (see my WWW home page, given in my sig, for info on g.gnuplot).
3. This does not do model fitting. See below.

After some discussion with beta testers, I've adopted a new approach
to this problem: instead of have a unified exploratory analysis tool, 
I've decided to break things up into many smaller programs.

Benefits: 
o no tacky (curses-based) menu system, uses command line parser
  (easier to use and understand)
o uses (much) less memory: order of N(sites)+N(h)
o may be more easily incorporated into shell scripts and
  cutesy mousy things (xgen or tcl scripts).
Drawbacks:
o less computationally efficient
o no unity. users must hunt down individual programs.

Hopefully everone will like this approach much better
than the first. However, I'm always open for suggestions.

SEE ALSO
s.univar, s.probplt, s.normal.

The old s.semivar has not been completely dismantled.
The next tool coming from this old code will
probably be "m.svfit," a tool that fits one of
several semivariogram models to the output of s.sv.

A big thanks goes to everyone who tested early versions of s.semivar
as well as other hacks that I have done (and to Bill Baker 
for always providing encouragement).

--Darrell

James Darrell McCauley, Purdue Univ, West Lafayette, IN 47907-1146, USA
mccauley at ecn.purdue.edu, mccauley%ecn at purccvm.bitnet, pur-ee!mccauley
http://soils.ecn.purdue.edu/~mccauley