| Re: Procedure FILLCONTOUR [message #1604] |
Thu, 20 January 1994 09:49 |
andy
Messages: 31
Registered: November 1993
|
Member |
|
|
Scott,
Many are those who have begged RSI to improve their contour filling method.
We at NASA Goddard have had numerous telephone conversations with the kind
folks in Boulder, and are 98% pleased with the results from IDL version 3.5.1.
A number of contour filling improvements have been made in this version.
Are you having difficulty with this or an earlier version of the software?
Andy
--
))))/ Andy Loughe
, , Goddard Space Flight Center
J Code 971 Bldg. 22 Room 380-B phone : (301) 286-3995
\_/ Greenbelt, MD 20771 e-mail: andy@toto.gsfc.nasa.gov
__/__/ __/ __/ __/__/__/ __/ __/
__/ __/ __/__/ __/ __/ __/ __/ __/
__/ __/ __/ __/ __/ __/ __/ __/ __/
__/__/__/__/ __/ __/ __/ __/ __/ __/
__/ __/ __/ __/__/ __/ __/ __/
__/ __/ __/ __/ __/__/ __/ __/
"I am sorry, but the number you dialed is imaginary."
"Please rotate your phone 90 degrees and try again."
|
|
|
|
| Re: Procedure FILLCONTOUR [message #1606 is a reply to message #1604] |
Thu, 20 January 1994 07:12  |
bowman
Messages: 121
Registered: September 1991
|
Senior Member |
|
|
In article <CJw4D6.17EF@yuma.ACNS.ColoState.EDU>,
scott@abyss.ATMOS.ColoState.Edu (Scott Denning) wrote:
>
> Has anybody gotten good results using FILLCONTOUR or CONTOUR, /FILL
> in IDL?
>
> I am trying to make maps of the world with color-filled contours of data
> overlayed on them. There are missing values that need to be masked out.
> Sounds simple enough, eh?
>
> The problem is that both routines leave large areas of "background" color
> unfilled wherever contours are "open" to the edge of the plot. This occurs
> even when /CLOSE is specififed as a keyword. These areas have many
> different ranges of values, but are all colored as if they have the same
> value (very misleading to readers).
>
> Also, the handling of missing values by "MAX_VALUE" appears not to work
> properly, and the contours are _very_ rough (as in not smoothed),
> consisting of just a handful of line segments rather than hundreds as I
> expected. MIN_CURVE_SURF no longer works to fix this.
There was discussion of this some time back. The /FILL option just doesn't
work very well in a number of cases. There are/were some bugs also, such
as not filling below the lowest contour. The bigest problem for me is that
/FILL won't work at all if there are missing data. This makes it pretty
well useless for my purposes. I can send you a hacked version of map_image
that handles missing data by not trying to interpolate at the edges of
gaps. You can overlay contours on this using any map projection (although
I haven't tested them all). It works best with smooth color maps, though.
It is impossible to get the mapped image and contours to agree exactly, and
the contours won't fall exactly on the boundaries between two colors when
using a steppd contour map (to try to simulate filled contours).
I know a lot of people who would like a filled contour routine that is more
general. I also think the MAX_VALUE keyword is an awkward way to do
missing values, and would prefer to see a MISSING_VALUE keyword (in a lot
of routines).
------------------------------------------------------------ ---------------
Dr. Kenneth P. Bowman 409-862-4060
Climate System Research Program 409-862-4132 fax
Department of Meteorology bowman@csrp.tamu.edu
Texas A&M University PP-Glider
College Station, TX 77843-3150
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
