[Paraview-developers] Sparse Data Reader

[Moreland, Kenneth]

It sounds like your data represents (or can represent) a block in space
with a sparse sampling of points within it. If that is the case, the most
natural representation for ParaView would be a poly data or unstructured
grid containing vertex cells at the location of your sparse samples.

If you're lazy like me, a quick way to get this up is to write your data
as a csv file (or some other delimited text file) with each row
representing one of your samples. The columns would be the x, y, and z
positions of each sample as well as additional columns for any other field
data.

ParaView will read that as a table. Use the "Table to Points" filter to
convert that to a point structure I first mentioned.

Of course, a csv file is not a good representation if you have a lot of
data, so you may need to do some custom solution anyway. But it is a good
way to get started with a small dataset to see if that works for you.

-Ken

On 11/6/13 10:13 AM, "Juergen Endres" <jerng at gmx.de> wrote:

>Hi all!
>
>What's the best way to create a paraview reader plugin for sparse data?
>We have some files containing image data (3D) which are stored in a
>sparse way. Reading the file does work, but using which data structure
>should this data passed to paraview? My plan was to use vtkSparseArray,
>but I didn't get paraview to accept this data. (empty volume). And
>converting the sparse data to a full-blown matrix would not be the best
>choice..
>
>Best, 
>JEndres
>_______________________________________________
>Paraview-developers mailing list
>Paraview-developers at paraview.org
>http://public.kitware.com/mailman/listinfo/paraview-developers