[Paraview] Merging Coincident Faces

[Berk Geveci]

The short answer is no. No such filter exists in ParaView or VTK.

However, this can be done efficiently (no for loops within Python) if
you have ParaView compiled with a Python that has NumPy. All official
3.10 binaries should have this.

Here is the recipe:

- Load data
- Apply programmable filter with the following to assign unique ids to cells:

output.ShallowCopy(inputs[0].VTKObject)
array = numpy.arange(0, output.GetNumberOfCells(), dtype=numpy.int32)
output.CellData.append(array, 'ids')

- Apply Cell Centers

- Apply Clean to merge duplicate cell centers

- Apply programmable filter (input 0 = reader, input 1 = Clean),
output type: unstructured grid, script:

from paraview import vtk
filter = vtk.vtkExtractSelection()
sel = vtk.vtkSelection()
selNode = vtk.vtkSelectionNode()
selNode.SetFieldType(vtk.vtkSelectionNode.CELL)
selNode.SetContentType(vtk.vtkSelectionNode.INDICES)
selNode.SetSelectionList(inputs[1].PointData['ids'].VTKObject)
selNode.GetSelectionList()
sel.AddNode(selNode)
filter.SetInput(1, sel)
filter.SetInput(0, inputs[0].VTKObject)
filter.Update()
output.ShallowCopy(filter.GetOutput())

This should produce an unstructured grid of 2D cells that are unique.
Works only in serial. In parallel, you need to reduce the polydata to
1 node using All to N.

-berk

On Wed, Jun 29, 2011 at 10:50 AM, Randall Hand <randall.hand at gmail.com> wrote:
> I'm working with an Xdmf dataset in ParaView 3.10.1 that has multiple
> unstructured-grid blocks with overlap regions, however doens't seem to
> actually have "ghost" cells.  This means when I extract an isosurface, I
> wind up with perfectly overlapping faces.
>
> I can use the Clean filter to eliminate coincident points, but but
> coincident faces.  As an example, if I use the "Select Cells Through"
> option to select 2 triangles on an overlapping border, I get 4 points
> (normal for 2 triangles) but 5 cells, with perfect overlap.  The results
> from an ASCII VTK File saved to disk:
>
> vtk output
> ASCII
> DATASET POLYDATA
> POINTS 4 float
> 1.37916 0.965356 0.21928
> 1.37129 0.964077 0.218481
> 1.37613 0.958331 0.219823
> 1.36947 0.958944 0.21948
> POLYGONS 5 20
> 3 0 1 2
> 3 2 1 3
> 3 0 1 2
> 3 2 1 3
> 3 2 1 3
>
> Is there any filter in ParaView capable of fixing this? Or does it have
> to be an arduously slow Python script to run in a Programmable Filter?
>
> --
> Randall Hand
> http://www.vizworld.com
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView
>
> Follow this link to subscribe/unsubscribe:
> http://www.paraview.org/mailman/listinfo/paraview
>