[comp.graphics.visualization] XDataSlice and sparse data

williams@casbah.acns.nwu.edu (Gus Williams) (05/18/91)

I have two questions.  The first concerns XDataSlice from NCSA.

I have read the documentation carefully and though it never actually
gives these requirements they appear to be true.

Does XDataSlice require a "square" data set?  That is do the
X, Y, and Z dimensions of the array of data need to be the
same? and does it now or in the future support variable spacing
of the data grid?  I know the vset extensions to HDF support these
items but do any of the visualization tools available from NCSA support
them?

Second question.  I have been reading the Dynamics Graphics sales
literature.  They use a method for iso-surfaces called (in their
literature) minimum tension surfaceing.  What is this?  Can anyone
point me to any good papers or references that give the algorithm?


Third question.  (OK I lied when I said I only had two)  Has anyone
implemented a 3D spatial averaging formula?  Something like distance
squared weighting or krieging in 2D.  I work with sparse data sets,
well samples, drill corings, etc, and to get good pictures you generally
need a much denser (sp) grid than I use.  However, most of the variables
I am interested in vary continiously in 3D space.  A pointer to a good
paper or algorithm would again be appreciated.

Thanks in advance.
-- 
GUS WILLIAMS			williams@casbah.acns.nwu.edu
(708) 491-3838			williams@voltaire.cme.nwu.edu
(708) 972-4609			williams@adelaide.ees.anl.gov
     Escape from the real world, get a PhD

lawrance@ncsa.uiuc.edu (David Lawrance) (05/20/91)

In article <1991May17.201453.15628@casbah.acns.nwu.edu> 
williams@casbah.acns.nwu.edu (Gus Williams) writes:
> Does XDataSlice require a "square" data set?  That is do the
> X, Y, and Z dimensions of the array of data need to be the
> same? and does it now or in the future support variable spacing
> of the data grid?  I know the vset extensions to HDF support these
> items but do any of the visualization tools available from NCSA support
> them?

1) Data dimensions can be different.
2) Data does not have to be regularly gridded.
3) PolyView supports the newer vset extensions.

Documentation and software available via anonymous ftp at ftp.ncsa.uiuc.edu