# Re: Filling colors in Delaunay triangles

From: Roshil Paudyal <p.roshil1994_at_nyahnyahspammersnyahnyah>
Date: Wed Jul 10 2013 - 02:21:04 MDT

Hello again,

@Adam: Thank you so much for the solution. I didn't really expect you to
write out the full script for me; this mailing list is super helpful!
As for the big triangles, the data file is actually much larger. I just
gave you the first 100 lines. So, using the whole data gave smaller
triangles and it came out all fine.
Thanks again.

@David: Thank you too for your concern. But everything looks fine for the
time being.

Just a quick question: Is there a function for calculating the area of a
voronoi polygon or do I have to divide it into triangles and add up their
areas?

Thank you,
Roshil

On Tue, Jul 9, 2013 at 3:21 PM, David Brown <dbrown@ucar.edu> wrote:

> Hi Roshi,
> I will be looking at your script and output as soon as possible.
> -dave
>
>
> On Jul 8, 2013, at 12:34 PM, Adam Phillips <asphilli@ucar.edu> wrote:
>
> Hi Roshil,
> This is a rather complicated issue. First off, you are setting gsFillColor
> = False, when you should be setting it to a color index or color name.
> See:
> http://www.ncl.ucar.edu/Document/Graphics/Resources/gs.shtml#gsFillColor
>
> Second, according to a search I did on the NCL web page, it's better to
> have your input longitude coordinates to be in the 0:360 range, EXCEPT when
> your coordinates cross the prime meridian where they should range from
> -180:180. You will see where statement coding in the attached script to
> account for this.
>
> Third, there are two triangles that you are trying to draw that are rather
> large. Here are their coordinates:
> (0) (40.719,245.982) (45.591,237.4) (-37.733,145.4) (40.719,245.982)
> (0) (40.719,245.982) (-37.733,145.4) (-13.533,288.067) (40.719,245.982)
>
> These 2 triangles are large enough that they are causing NCL to color fill
> outside the triangles instead of inside. I am not sure how to get NCL to
> fill inside those two triangles as opposed to outside. Someone else will
>
> Finally, I used gsn_add_polygon due to personal preference. In this case
> it automatically switched the drawing of the continental outlines to be
> drawn last, which I assume you prefer.
>
> I have attached a script and the example graphic.
> Good luck,
>
> On 7/8/13 9:38 AM, Roshil Paudyal wrote:
>
> Hi,
>
> I tried that but it still doesn't solve the problem. Using polyline gives
> good results; so I don't think there is anything wrong in the rest of the
> code. The problem must lie somewhere in executing the gsn_polygon function.
>
> It would be great if you could run this script once, which may make the
> problem clearer. It does give some output; but that is very different from
> what is expected (or what polyline gives). I have already attached the only
> datafile I used (illdt1).
>
> Thank you,
> Roshil
>
> On Mon, Jul 8, 2013 at 10:38 AM, Adam Phillips <asphilli@ucar.edu> wrote:
>
>> Hi Roshil,
>> I believe you have to enclose a complete polygon for gsn_polygon to work.
>> In other words, the first and last latitude/longitude points are usually
>> the same. Also, you need to make sure that you are advancing the frame at
>> the end of your script.
>>
>> Try changing this:
>>
>> qlat = new(3,float)
>> qlon = new(3,float)
>> gsres@gsEdgesOn = True
>> gsres@gsEdgeColor = 4
>> gsres@gsFillColor = False
>> do np=0,num_triangles-1
>> qlat(0) = lat_in(triangles(np,0))
>> qlon(0) = lon_in(triangles(np,0))
>> qlat(1) = lat_in(triangles(np,1))
>> qlon(1) = lon_in(triangles(np,1))
>> qlat(2) = lat_in(triangles(np,2))
>> qlon(2) = lon_in(triangles(np,2))
>> gsn_polygon(wks,map,qlon,qlat,gsres)
>> end do
>>
>> to this:
>>
>> qlat = new(4,float)
>> qlon = new(4,float)
>> gsres@gsEdgesOn = True
>> gsres@gsEdgeColor = 4
>> gsres@gsFillColor = False
>> do np=0,num_triangles-1
>> qlat(0) = lat_in(triangles(np,0))
>> qlon(0) = lon_in(triangles(np,0))
>> qlat(1) = lat_in(triangles(np,1))
>> qlon(1) = lon_in(triangles(np,1))
>> qlat(2) = lat_in(triangles(np,2))
>> qlon(2) = lon_in(triangles(np,2))
>> qlat(3) = qlat(0)
>> qlon(3) = qlon(0)
>> gsn_polygon(wks,map,qlon,qlat,gsres)
>> end do
>> frame(wks)
>>
>>
>> If that does not solve the problem, please respond to ncl-talk to let
>> everyone know.
>>
>>
>> On 07/07/2013 11:31 AM, Roshil Paudyal wrote:
>>
>> Hi,
>>
>> I am trying to plot some data on a sphere and create a Delaunay
>> triangulation out of it and fill colors into each of the triangles using a
>> custom color map. Following code does a good job of creating the
>> triangulation: (ncl v. 6.1.2 in cygwin)
>>
>>
>>
>> begin
>>
>> NARC = 50
>>
>> id_in = jps(1:100,0)
>> lat_in = jps(1:100,1)
>> lon_in = jps(1:100,2)
>> dt_in = jps(1:100,3)
>> N = 100
>>
>>
>> gsres = True
>>
>>
>> ; Create the triangulation.
>> ;
>> triangles = csstri(lat_in,lon_in)
>> tri_sizes = dimsizes(triangles)
>> num_triangles = tri_sizes(0)
>>
>>
>>
>> cmap = (/ \
>> (/ 1., 1., 1. /), \
>> (/ 0., 0., 0. /), \
>> (/ 1., 0., 0. /), \
>> (/ 0., 0., 1. /), \
>> (/ 1., 0., 0. /), \
>> (/ 0., 1., 0. /), \
>> (/ 0., .8, 0. /), \
>> (/ .65, .65, .65 /) \
>> /)
>>
>> NCGM=1
>> X11=0
>> PS=0
>> PDF=0
>>
>> if (NCGM .eq. 1) then
>> wks = gsn_open_wks("ncgm","nm21n")
>> end if
>> if (X11 .eq. 1) then
>> wks = gsn_open_wks("x11","nm21n")
>> end if
>> if (PS .eq. 1) then
>> wks = gsn_open_wks("ps","nm21n")
>> end if
>> if (PDF .eq. 1) then
>> wks = gsn_open_wks("pdf","nm21n")
>> end if
>> gsn_define_colormap(wks,cmap)
>>
>>
>>
>> ;
>> ; Draw a globe
>> ;
>> map_resources = True
>> map_resources@gsnFrame = False
>> map_resources@mpOutlineBoundarySets = "National"
>> map_resources@mpNationalLineColor = 1
>> map_resources@mpGeophysicalLineColor = 7
>>
>> map_resources@mpLimbLineColor = 7
>> map_resources@mpGridLineColor = 0
>> map_resources@mpGridAndLimbDrawOrder = "PreDraw"
>>
>> map_resources@mpCenterLatF = 40.
>> map_resources@mpCenterLonF = -105.
>> map_resources@vpXF = 0.06
>> map_resources@vpYF = 0.90
>> map_resources@vpWidthF = 0.88
>> map_resources@vpHeightF = 0.88
>> map_resources@mpSatelliteDistF = 4.0
>>
>> map_resources@mpGreatCircleLinesOn = True
>>
>> map = gsn_map(wks,"Satellite",map_resources)
>>
>>
>>
>> ; Draw the triangles.
>> ;Ignore the blue for now
>>
>> qlat = new(4,float)
>> qlon = new(4,float)
>> gsres@gsLineColor = 5
>> do np=0,num_triangles-1
>> qlat(0) = lat_in(triangles(np,0))
>> qlon(0) = lon_in(triangles(np,0))
>> qlat(1) = lat_in(triangles(np,1))
>> qlon(1) = lon_in(triangles(np,1))
>> qlat(2) = lat_in(triangles(np,2))
>> qlon(2) = lon_in(triangles(np,2))
>> qlat(3) = lat_in(triangles(np,0))
>> qlon(3) = lon_in(triangles(np,0))
>> gsn_polyline(wks,map,qlon,qlat,gsres)
>> end do
>>
>> end
>>
>> However, is there a way to fill colors into the triangles created using
>> polylines? I didn't find any. So, I thought maybe I could give the end
>> points of triangles as an input to gsn_polygon. If that created the
>> triangles, that would have given me a full set of resources to modify,
>> including the fill colors. So, I replaced the blue part with the following
>> code:
>>
>> qlat = new(3,float)
>> qlon = new(3,float)
>> gsres@gsEdgesOn = True
>> gsres@gsEdgeColor = 4
>> gsres@gsFillColor = False
>> do np=0,num_triangles-1
>> qlat(0) = lat_in(triangles(np,0))
>> qlon(0) = lon_in(triangles(np,0))
>> qlat(1) = lat_in(triangles(np,1))
>> qlon(1) = lon_in(triangles(np,1))
>> qlat(2) = lat_in(triangles(np,2))
>> qlon(2) = lon_in(triangles(np,2))
>> gsn_polygon(wks,map,qlon,qlat,gsres)
>> end do
>>
>> But this doesn't even create all the triangles. So, is there a way to
>> fix this problem? I have attached the data I used for this (illdt1.txt).
>>
>>
>> Thank you,
>> Roshil
>>
>>
>>
>>
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>>
>> --
>> ______________________________________________________________
>> NCAR/Climate and Global Dynamics Division (303) 497-1726
>> P.O. Box 3000
>> Boulder, CO 80307-3000 http://www.cgd.ucar.edu/cas/asphilli
>>
>>
>
> <nm21.gif><script1.ncl>_______________________________________________
>
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Wed Jul 10 02:21:25 2013

This archive was generated by hypermail 2.1.8 : Fri Jul 12 2013 - 16:37:39 MDT