Re: Adaptive Resolution and Sampled Sets

  • To: Matthew Leonard <mel2@xxxxxxxxxx>
  • Subject: Re: Adaptive Resolution and Sampled Sets
  • From: Bill Hibbard <billh@xxxxxxxxxxxxx>
  • Date: Thu, 23 May 2002 14:37:09 -0500 (CDT)
Hi Matt,

> I have created an Irregular3DSet using DelaunayCustom with a specific
> kind of tetrahedralization.  This tetrahedralization is
> multiresolutional.  What I am trying to do is "adapt" certain created
> polygon vertices in makeIsosurface().  I am keeping track of my original
> (x,y,z) coordinates in a separate class and reference each isosurface
> polygon to each of my created tetrahedrons.  What I have found is that
> once I create my Irregular3DSet and pass it my samplePoints[][] array,
> it "normalizes" them somehow from -1.0 to 1.0.  From where does this
> mapping come?  I would like to prevent such mapping so that the (x,y,z)
> coords in Irregular3DSet.Samples[][] are the same as the coords in my
> samplePoints[][] array originally passed to Irregular3DSet.  In keeping
> them the same I am able to efficiently adapt the polygon vertices as
> needed.  How might one change or rather prevent such a conversion?

The test in 'java visad.Irregular3DSet' demonstrates this
is not true. It's main() method constructs an Irregular3DSet
and then prints the values of the Irregular3DSet's Samples
array, and they have the values that were passed to the
constructor (values in the 100's).

However, if you link your Irregular3DSet to a DisplayImpl,
a different Irregular3DSet will be constructed inside the
display logic with values in the range (-1.0, 1.0). This is
done by the auto-scaling logic. If this is what you mean,
you can prevent this by calling setRange(-1.0, 1.0) for your
ScalarMaps to XAxis, YAxis and ZAxis.

Good luck,