Jon Allen wrote:
Ok, here's a rough outline of what I'm doing:
take spherical coordinates and assign them a value
mangle the spherical coordinates to their cartesian coordinates
output into constant r rows, and chunks of constant phi (0->2*pi), and
then vary theta (0->pi)
I have no clue where to start (my attention span is making it quite
difficult to make it through any reasonable amount of the user guide,
and gain anything from it), but I did take a look at the irregular.dx
file and got some pointers from that... so here's what my output looks
like:
object 1 class array type float rank 1 shape 3 items 12960 data follows
0.000000 0.000000 0.000000
............. (these dots aren't here... they represent the other 12958
lines)........
16.245957 -2.864601 -93.556736
attribute "ref" string "positions"
object 2 class array type float rank 0 items 12960 data follows
1.0
.................................
1.0
attribute "dep" string "positions"
attribute "ref" string "data"
ok, so my question, I /want/ to connect these points (either through
x,y,z (only since delta isn't constant it will be a bent grid), or
through a spherical connection between points (ala, globe)).... if you
have /any/ advice that would be great.
You could import the r,phi,theta data-cube into DX, with regular
connections, and then warp the positions of that field to a sphere:
Mark(positions) --> Compute(x=x(r,th,ph),y=y(r,th,ph),y=y(r,th,ph))
--> Unmark(positions)
In other words, do the mangling inside DX.
However, the "seam" at phi=0 might give some problems or ugly results
because DX won't interpolate across it
-Jeroen
