| Re: new changes to contour under IDL Version 3.1.1 (sunos sparc) [message #1459 is a reply to message #1302] |
Mon, 04 October 1993 12:06   |
turet
Messages: 7
Registered: January 1993
|
Junior Member |
|
|
It's nice (not good) to hear about similar problems I've been having with
filled contours, etc.
I've got many gripes but I'll just run off a few at the top before I have
to leave for a class.
The color fills aren't working for me... on one set of plots I couldn't get
the lowest level to fill, one that happened to intersect the plot boundary,
i.e. if level=[-2,-1,0,1,2] and there was a region of -0.5, say that inter-
sected the boundary (and that was the minimum on the plot), then that region
wouldn't be filled at all. I tried fooling around with the colors keyword
with no success (the documentation for color contouring seems obscure to me).
I also had problems with a contour filling in the wrong color, i.e.
1 1 1 1 1 1 1 1 1 1 1
1 0 0 0 0 0 0 0 0 0 1
1 0 0 1 1 1 1 1 0 0 1
1 0 0 1 1 1 1 1 0 0 1
1 0 0 0 0 0 0 0 0 0 1
1 1 1 1 1 1 1 1 1 1 1
then the center block would be colored wrong.
I'm not happy with the downhill keyword, at least it should be adjustable,
e.g. downhill=2 or -3 for different length ticks (and uphill)
Speed up MIN_CURVE_SURF or else I'll go to Numerical Recipes (a pain).
I agree with a previous poster about NCAR graphics. The old (v1.0) IDL
had some contour stuff that appeared to be right out of NCAR graphics.
I could really use the markings of highs and lows with the values, and
perhaps symbols H and L.
Contour annotation positioning control is also very important to me. I got
a review of a paper back saying that I should line up the contour labels
so they're not jumbled around the page. I contacted RSI but was told that
contour label positions are determined by a "complicated heuristic", in other
words S.O.L. In one case I couldn't suppress the contour labels outside
the range I used to clip the contours themselves and I ended up with labels
crammed into the edges of the plot. I used a lot of white-out, ended up
losing the labels altogether and having them drawn in.
That reminds me, clip doesn't seem to work at all with color fills, although
I've just been working with v3.1 for a week.
What else? I had the same problem as a previous poster about the misalignment
using MAP_IMAGE with PostScript. RSI suggested that I just measure the
misalignment and apply some type of scaling to get it back in shape. I would
like to have a bit more control of the device parameters.
It also be nice to have the !MAP system variable documented a bit better,
perhaps to use parts of the structure to do some graphics transformations.
Also, it looks like !MAP gets set when you use a MAP_ call but it doesn't
get unset or any other changes made when you do an ordinary graphics call
after that. I would like to have the !MAP variable reflect the current
state of the graphics so I build other routines around this.
One more I promise: I had a routine that did some plotting on top of
graphics that had some xyouts and plots using /normal coordinates. When
the map was set I couldn't plot in normal coordinates outside the [0,1]
range, i.e. xyouts,.1,-.1,'*',/normal wouldn't work. Not only didn't it
work but it froze up my IDL session, I couldn't even Control-C, I had to
Control-\ or kill my window to get out.
What say? do we have a movement here. IDL is starting to set us back some
bucks. I read in the paper that there are all sorts of joint ventures going
on (IBM?) so IDL is starting to go big time. All us readers here are in it
pretty much on the ground floor (quite a way from the LASP days, eh Dave?).
Let's get this thing working right. Also, check out IVE, developed at the
U. of Washington Atm. Sci. department. I don't know the status/availability
of it, but it's a window/interactive thing for NCAR graphics, and it's
beautiful
--Phil Turet
turet@pmel.noaa.gov
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
