Re: Fill plotting question

[hankin@xxxxxxxxxxxxxxxxxxx (Steve Hankin)]

	*** End point considerations when contouring on a modulo axis ***

Hi Mark,

The SHADE command does not perform any interpolation so it colors the L=1 grid
cell from the start of that cell (which is before t="6-jan:02:00") to the end
of the L=36 cell (which is actually after "26-dec:22:00")

CONTOUR (and FILL), by contrast, do perform interpolations. Thus to contour
before L=1 (t="6-jan:02:00") requires that the L=36 point be adjacent and to
contour after L=36 ("26-dec:22:00") requires that the L=1 point be adjacent.

By default Ferret fetches only the actual data requested (L=1:36). It needs to
be coached to fetch before the L=1 and after the L=36 points. In this solution
you'll see that can be achieved by requesting "myvar[l=0:37]"

  define axis/x=125:286.5:.25/units=longitude xc
  define axis/modulo/t="6-jan:02:00":"26-dec:22:00"/np=36/t0="01-jan-0000" tc
  define grid/x=xc/t=tc gc

  set grid gc
  let myvar = i+l	! a quickly made-up variable

  fill/ylimits=1-jan:31-dec myvar[l=0:37]


	Happy Ferreting

	- steve

==============================================================

--- Forwarded mail from "Mark Verschell" <markv@shear.coaps.fsu.edu>

To: ferret_users@ferret.wrc.noaa.gov

I have a climatology file defined on the following grid:

define axis/x=125:286.5:.25/units=longitude xc
define axis/modulo/t="6-jan:02:00":"26-dec:22:00"/np=36/t0="01-jan-0000" tc
define grid/x=xc/t=tc gc

If I shade a variable defined on this grid I get a plot that has a time axis
from January to December that is completely filled in (i.e. no blank space
inside the axis lines). However, if I contour or fill this same variable I get
a plot where the plotted information runs from "6-jan:02:00":"26-dec:22:00"
with the same time axis as the shade plot (January to December), with lots of
blank space inside the axis lines. I tried a bunch of different set region
commands and the closest I could come was when I set the region to l=35:73, in
which case the plotted information and the time axis in the plot also included
the leading December and trailing January (i.e. 14 months). The solution I came
up with was to explicitly set the region and the time axis limits thusly:

fill/ylimits=8785:17544 ct[l=35:73]

where ct is the variable I want to plot.

I am not that surprised at this behavior, but I would like to know if anyone
has a more elegant method of accomplishing this.

Thanks, Mark

--
-------------------------------------------------------------------------------

	      Mark Verschell (verschell@coaps.fsu.edu)
	      Center for Ocean and Atmospheric Prediction Studies
	      Florida State University
              2035 E. Paul Dirac Drive
              R.M. Johnson Bldg. - Suite 200
	      Tallahassee, FL  32310
	      (904) 644-6532              (904) 644-4841 (fax)


--- End of forwarded mail from "Mark Verschell" <markv@shear.coaps.fsu.edu>


-- 

		|  NOAA/PMEL               |  ph. (206) 526-6080  
Steve Hankin	|  7600 Sand Point Way NE  |  FAX (206) 526-6744
		|  Seattle, WA 98115-0070  |  hankin@pmel.noaa.gov