Re: Some details with Display.shapes.

• To: visad-list@xxxxxxxxxxxxx
• Subject: Re: Some details with Display.shapes.
• From: "William L. Hibbard" <WHIBBARD@xxxxxxxxxxxxx>
• Date: Mon, 12 Oct 1998 07:11 CST

Eugene,
 
>         I have a couple of questions, hopefully you guys can help out.
>
> I used Display.Shape to create a plate-object
> The size/scale of the platess correspond to my sample data set.
> And the "display" consists of multiple plates.
 
Good.
 
> (1)     Is there a way to select an individual plates?
>         I hope to trigger visualization-relative to selected plates.
>         As far as I know - the mouse interaction allows users to
>         rotate/zoom/translate/move cursor around the whole display
 
Probably the best way to do this is to create a custom cursor as
a RealTuple data object (whose RealTupleType consists of those
RealTypes that determine plates' display locations).  This RealTuple
would be displayed using DirectManipulationRendererJ3D (or
DirectManipulationRendererJ2D) and would trigger a CellImpl that
computed the closest plate.
 
> (2)     Is there a way to specify multiple colors in different
>         regions on ONE plate? (if possible without having to
>         create the plate as a composite object of different shapes)
>         If there isn't - can you direct me if I want to implement that
>         feature somewhat efficiently
 
You can control the color of each vertex in your VisADGeometryArray
by setting values in its colors array - this float[] array has three
elements per vertex for red, green and blue components, all in the
range 0.0f to 1.0f.  So colors[0] is the red component of vertex 0,
colors[1] is it green component, colors[2] is its blue component,
then colors[3] is the red component of vertex 1, etc.
 
Of course, if the plate Shape occurs more than once in the display
all occurrences will have the same colors.  In that case you might
have two (or more) plate Shapes and change values in your data
Fields (probably FlatFields?) to control which data point is mapped
to the plate Shape with the special color.  If this happens as a
result of plate selection in question (1), the switch of Field values
could be done in the CellImpl that finds the closest plate.
 
> (3)     Thank you very much for the software. It's wonderful !!
>         But I must say that it's hard for one to learn to use the
>         many powerful features of visad. I have been reading the
>         documentation and experimenting with DisplayTest.java.
>         I find that it's difficult to know what is happening in
>         DisplayTest.java
>         I have request - is it possible that you include some
>         samples that is less abstract - and somewhat more contained.
>         Contained => no fortran, no images or other data format, no
>         vis5d.
>         Basic ideas like - creating Real Data Objects that has numerical
>         values, mapping them to some display, Creating DataReference,
>         illustrating how to use the reference to change data values,
>         and have the display updated accordingly.
>         How to implement direct manipulation.
 
I'm happy you like VisAD and also well aware of the shortcomings of
the documentaiton.  I revised the Developers Guide several times based
on Tom Whittaker's suggestions, including adding Simple.java and several
other application examples.
 
Part of the problem is that VisAD is different from any other system
so its basic ideas are unfamiliar ("too many concepts" as Tom Yoksas
might say).
 
Part of the problem is my limitations as a writer, plus I am very
familiar with the system so sometimes assume too much in my
explanations.  One cure for this would be for someone else to write
tutorial(s).  I will add links from the VisAD web page to any tutorials
that anyone cares to write.
 
> (4)     BTW the javadoc link in the homepage is a broken link again
>         Can someone fix it?
 
Oops, we'll see what we can do about this.
 
Cheers,
Bill
p.s., still working late I see ;)
 
----------------------------------------------------------
Bill Hibbard, SSEC, 1225 W. Dayton St., Madison, WI  53706
whibbard@xxxxxxxxxxxxx  608-263-4427  fax: 608-263-6738
http://www.ssec.wisc.edu/~billh/vis.html
 
"kill cross-platform Java by growing the polluted Java market"
   - from an internal Microsoft planning document