| map_grid limitations [message #30207] |
Wed, 10 April 2002 14:18 |
Ken Mankoff
Messages: 158
Registered: February 2000
|
Senior Member |
|
|
Hi,
I am interested in using IDLs mapping procedures on Mars, which has a
fair number of datasets on the 0 to 360 *west* longitude system as
opposed to the -180 to 180 *east* system.
I posted this question a week or two ago, but got no responses. I
don't think I stated the problem terribly clearly in the original
post, so I am trying again. Also, IDL wrote me and said that:
> Some national labs have changed these routines (which are *.pro
> code) to work with the Moon, so check the newsgroup.
After an extensive web-search, I found a "mapl_set" procedure
mentioned here:
http://www.ias.fr/cdp/Futur/perso/biblio2.html
But there is no link and no code, only a description!
------------------------------------------------------------ ----------
Here is a demonstration of the problem:
A simple map in IDL is created like this:
IDL> map_set, 0, 0, /cylindrical
IDL> map_grid, /label
You will notice that the longitudes increase to the east, and are on a
-180 to 180 degree grid. I can get the appearance of longitudes
increasing to the west and on a 0 to 360 grid by these commands:
IDL> map_set, 0, 0, /cylindrical
IDL> map_grid, lons=[0,45,90,135,180,-135,-90,-45], $
lonnames=[0,315,270,225,180,135,90,45], /label
Everything appears fine. But the problem is that it is only the labels
(that I hard-coded above) that *appear* to be westward, and 0 to 360.
The entire map is still doing its math thing on a -180 to 180 east
grid.
This becomes apparent when you draw a box (in my case, the viewing
area of the satellite instrument. If, for example the instrument is
viewing the box from lat,lon coordinates [-30,45] (bottom left corner)
to [0,0] to right corner:
loadct, 39 & !p.color = 254 ; red box
plots, 45, -30, /data
plots, 45, 0, /data, /cont
plots, 0, 0, /data, /cont
plots, 0, -30, /data, /cont
plots, 45, -30, /data, /cont
If you do this after issuing the first set of mapping commands above,
it appears as expected. But if you do it after the second map_set
command, it appears in the wrong place (the same place as before).
------------------------------------------------------------ ----------
Finally, if the above make sense, I will give you an exact description
of my problem. The above problem with drawing the box can be solved
easily, because a very simple function will translate between the four
possible grids (0-360 vs 180-180 vs 'east' vs 'west'). It would be
nice to specify this via a keyword *once* when I call map_set, and not
worry about it, but I can do the translations. But here is the real
problem:
If you hardwire the longitude labels, you have to hardwire their
location (i.e. every degree or every 45 degrees, as above). My
software is used to view the surface of mars, and allows the user to
view it on a global scale, or down to a 1x1 degree grid.
map_set, if allowed to 'do its own thing', will always generate nice
grids for you. If you are looking at the globe, it will use 45 degree
grids as shown above. If you are looking at a 1x1 degree grid, it will
use .1 degrees.
But if I hardcode in the longitude names so they are westward, I also
have to code their locations. If I do every degree, you cannot see the
globe, all you see is the grid covering it. If I do every 45 degrees,
then you see no grids when you view a small area...
Hence, my desire do use map_grid in a 'dumb' way (let it do the math
to determine the grid spacing), and have it output westward 0-360
grids.
Finally, polar /ortho projections make this even more complicated than
equator-based /cylindrical projections.
Sorry for the length of this post...
-k.
--
Kenneth Mankoff
LASP://303.492.3264
http://lasp.colorado.edu/~mankoff/
http://lasp.colorado.edu/snoe/
http://lasp.colorado.edu/mars/
http://lasp.colorado.edu/marsrobot/
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
