I am trying to develop a strategy for examining cloud top radiances and
temperatures in GRASS4.0. I would like to be able to generate a hard copy
of the histogram of the count values (similar to that from d.histogram),
so that I could select different threshold levels. So far I have only been
able to eyeball the values from d.profile and d.histogram, but would
appreciate if you could suggest a more quantitative module by which to be
able to categorize the data.
Looking forward to your reply.
Lesley-Ann Dupigny.
Lesley-Ann Dupigny (dupigny@felix.geog.mcgill.ca) writes on 16 Nov 93:
temperatures in GRASS4.0. I would like to be able to generate a hard copy
of the histogram of the count values (similar to that from d.histogram),
The following example produces postscript output. This could be easily
modified for TeX, FrameMaker, groff, or whatever.
% r.stats -qc i=temperatures o=temp.histo
% g.gnuplot
gnuplot> set term postscript
gnuplot> set out 'temp.ps'
gnuplot> plot 'temp.histo' with impulses
gnuplot> quit
(then print temp.ps)
g.gnuplot is not part of 4.0 but will be in 4.2. It is currently
available via anonymous ftp from moon.cecer.army.mil in
grass/incoming/g.gnuplot3.5.tar.Z. Included in this is a manual.
I'm assuming that the -qc flags exist in the 4.0 r.stats...
--Darrell
Don't overlook the i.class function for quick looks at the data. I
provides a small scale image window, a zoomed image window, and a
window to hold the histograms. It outputs the means to a file with
annotation! It is meant to perform maximum likliehood classification,
so you must enter in two images. I use g.copy to create a copy of the
first image with a different second name, then i.group them together
for i.class.
Another way to get histograms is to use r.stats -a input=fname |
xgraph -bar. This tabulates each unique value, and the number of
counts for each value.
Good Luck!
dave