Hi Patrick,
Bill's suggestion for your Gridded3DSet contruction should take care of
the continuous surface instead of points problem (though I'd get rid of
the ConstantMap to Display.PointSize, I don't recall if it triggers a
change to point mode). The topology of the set will affect how it is
displayed.
What does the set look like that you are using to resample? I suspect a
similar error in that set might be the problem.
Cheers,
Doug
"P. Tewson" wrote:
>
> Hello Bill and the list,
>
> > If you have a FlatField
> > whose domain Set is a Gridded3DSet with manifold dimension
> > = 2 (the result of the resample() call above), then with
> > proper ScalarMaps your display should be a 'solid' surface,
> > possibly transparent.
>
> The result of 'getManifoldDimension()' on my Gridded3DSet is definately
> '2', though the 'getDimension()' call returns '3'. However, the horizontal
> slice still displays as a grid of points and not a 'solid' surface, and
> the transparency still does not appear. At the end of this message I am
> going to include a summary of how the MathTypes and ScalarMaps are
> declared at the end of this message.
>
> First I wanted to describe he other odd behaivior I'm experiencing. If the
> FlatField is not re-initialized before each call to resample(), the
> following exception occurs:
>
> visad.SetException: Gridded3DSet.valueToGrid: ManifoldDimension must be 3
> at visad.Gridded3DSet.valueToGrid(Gridded3DSet.java:686)
> at visad.GriddedSet.valueToInterp(GriddedSet.java:319)
> at visad.FlatField.resample(FlatField.java:3350)
> at visad.FunctionImpl.resample(FunctionImpl.java:190)
> at sennev.SEView1b.setCDLevel(sennev/SEView1b.java:213)
>
> However, if I re-new the FlatField first, the exception does not occur.
> The odd thing is that in BOTH cases, the Manifold dimension of the
> accompanying domain set is 2, and the Domain dimension is 3. So although
> the exception would indicate that the problem is with some aspect of the
> manifold dimension, the manifold dimension is the same for the succesful
> and unsuccesful cases.
>
> Here are the declarations --
>
> The RealType declarations:
>
> // I know that the units should be the built-in lat/lon,
> // I'm getting to that
> latitude = RealType.getRealType("latitude", SI.meter, null);
> longitude = RealType.getRealType("longitude", SI.meter, null);
> depth = RealType.getRealType("depth", SI.meter, null);
> se = RealType.getRealType("Signal_Excess");
>
> The display Implementation:
>
> display = new DisplayImplJ3D("display1");
>
> The ScalarMaps:
>
> latMap = new ScalarMap( latitude, Display.YAxis );
> lonMap = new ScalarMap( longitude, Display.XAxis );
> depthMap = new ScalarMap( depth, Display.ZAxis );
> seAlphaMap = new ScalarMap( se, Display.Alpha );
> seRGBMap = new ScalarMap(se, Display.RGB);
>
> Declaring the arrays:
>
> depthSlice = new float[NROWS][NCOLS];
> seDomain = new float[3][NROWS*NCOLS];
> flat_samples = new float[1][NROWS * NCOLS];
>
> The 3D RealTupleType for domain:
>
> domain_tuple3D = new RealTupleType(latitude, longitude, depth);
>
> Declaring the Gridded3DSet, FlatField, and DataReference:
>
> domain_set3D = new Gridded3DSet(domain_tuple3D, seDomain,
> 1,NROWS* NCOLS);
>
> func_domain_rgbVal = new FunctionType( domain_tuple3D, se);
>
> vals_ff = new FlatField( func_domain_rgbVal, domain_set3D);
>
> se_data_ref = new DataReferenceImpl("data_ref");
> vals_ff.setSamples( flat_samples , false );
> se_data_ref.setData((FlatField)vals_ff.resample(domain_set3D
>
> Adding the Reference to the Display:
>
> display.addReference(se_data_ref, new ConstantMap[] { new
> ConstantMap( 1.0f, Display.PointSize ) });
>
--
*----------------------------------------------------------------------*
| Doug Lindholm, Software Engineer | E-mail: lind@xxxxxxxx |
| Research Applications Program | Phone: 303-497-8374 |
| National Center for Atmospheric Research | |
| P.O. Box 3000 | There's no place |
| Boulder, Colorado 80307-3000 | like $HOME |
*----------------------------------------------------------------------*
