On Nov 18, 10:40 pm, mbwel...@gmail.com wrote:
> On Nov 17, 10:30 pm, l...@lbnc.de wrote:
>
>
>
>> On 18 Nov, 04:28, mbwel...@gmail.com wrote:
>
>>> Hello,
>
>>> I have an output file that gives me data in the form of:
>>> Top, bottom, left, right sides of a cell (so basically the corners)
>>> and two separate magnitude data columns.
>
>>> What I need to be able to do is create two image files that
>>> graphically, either via contours, colored cell scheme, or other...,
>>> display the values of magnitude per cell ( either Magnitude value 1 or
>>> value 2).
>
>>> Is this feasible to do in IDL, if not is there another program better
>>> suited?
>
>>> If anyone has any ideas, I would really appreciate it.
>
>>> Thanks,
>
>>> ~Matt
>
>> Although it will not be the fastest solution and produce large
>> PostScript files, I would suggest looping over all values, plotting
>> them using POLYFILL, like so:
>
>> ;- establish axes, without actually plotting them, because the
>> polyfill would cover them
>> plot, [0], /nodata, xstyle=5, ystyle=5, $
>> xrange=some_xrange, yrange=some_yrange
>> for i=0L, ndata-1L do $
>> polyfill, [xstart[i], xend[i], xend[i], xstart[i], xstart[i]], $
>> [ystart[i], ystart[i], yend[i], yend[i], ystart[i]], $
>> color=data_value[i], noclip=0
>> ;- plot axes
>> plot, [0], /nodata, /noerase, xstyle=1, ystyle=1, xrange=some_xrange,
>> yrange=some_yrange
>
>> Cheers
>> Lasse
>
> Thanks for the info, but in looking at your code, it appears that it
> would assign case 1 as one color and case 2 as another color, correct?
>
> If so, what I need to do is create a grid that has a color range. Let
> us say the values are between 1 and 5, so that value one is purple and
> value two is red and all values in between are a gradation between the
> extremes. Of course this is a simple case and I will probably have
> values a couple orders in magnitude in range and will need to be able
> to "auto bin" them. Will this code work for that, or do I need
> something else?
>
> Thanks,
> ~Matt
It might be better to do it as an image... let's say that the data
bounds are stored in arrays left, right, top, and bottom, and the data
range in the x-direction is given in data_xrange and the same for
data_yrange, and you want your final image to be nx by ny pixels. Then
you could use this to create the image:
image = fltarr(nx,ny)
deltax = (xrange[1]-xrange[0])/float(nx)
deltay = (yrange[1]-yrange[0])/float(ny)
for i=0l,ndata-1 do $
image[(left[i]-xrange[0])/deltax:(right[i]-xrange[0])/deltax , $
(bottom[i]-yrange[0])/deltay:(top[i]-yrange[0])/deltay] = magnitude
[i]
Then you can use the normal image viewing routines (e.g. tvimage) to
plot them:
erase
loadct, 10 ; or whatever you want - the Brewer tables would probably
be useful
location = [0.1,0.1,0.9,0.9]
tvimage, bytscl(image,top=250)+4, position=location
plot, /noerase, /nodata, [0],[0], position=location, xrange=xrange,
yrange=yrange
-Jeremy.
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
