Re: 3D Set bug?

• To: Doug Lindholm <lind@xxxxxxxx>
• Subject: Re: 3D Set bug?
• From: Bill Hibbard <billh@xxxxxxxxxxxxx>
• Date: Sat, 16 Mar 2002 15:19:06 -0600 (CST)

Hi Doug,

> I have a horizontal slice of 3D data - Gridded3DSet with a 2D manifold.
> In a 3D display I get contours as expected. In a Display with a
> TwoDDisplayRendererJ3D, the contours are there but not correct. If I
> remove contouring, I get an image in the 3D display but only dots in the
> 2D display.
>
> Is this a limitation of the 2D display or a bug?

ScalarMaps to ZAxis are illegal with TwoDDisplayRendererJ3D,
so when you display Gridded3DSet you can map two of its
domain RealTypes to XAxis and YAxis, but one of its domain
RealTypes cannot be mapped spatially. Thus the Display logic
cannot use the topology of the Gridded3DSet in display space.
That's why the image turns into dots. For contours, it needs
a topology and tries to compute a new one (since it cannot
use the one from the Gridded3DSet). I am not sure exactly
what is happening in your case, but am not too surprised
there are problems. I suggest converting your Gridded3DSet
with manifold dimension = 2 to a Gridded2DSet, for display
with TwoDDisplayRendererJ3D.

Good luck,
Bill

• References:
  • 3D Set bug?
    • From: Doug Lindholm