[vtkusers] (no subject)

[Schubert, Raphael]

I use the following script to generate three vtk files containing different dataset types and an additional FIELD.

import numpy as np
import vtk
from vtk.util.numpy_support import numpy_to_vtk

for dataset in (vtk.vtkStructuredPoints,
                vtk.vtkRectilinearGrid,
                vtk.vtkPolyData):

    vtkData = dataset()

    if dataset is vtk.vtkStructuredPoints:
        vtkWriter = vtk.vtkStructuredPointsWriter()

        vtkData.SetOrigin(0.0, 0.0, 0.0)
        vtkData.SetSpacing(1.0, 1.0, 1.0)
        vtkData.SetExtent(0, 1, 0, 1, 0, 1)

    elif dataset is vtk.vtkPolyData:
        vtkWriter = vtk.vtkPolyDataWriter()

        vtkPoints = vtk.vtkPoints()
        vtkPoints.SetData(numpy_to_vtk(np.asarray(
            [[0.0, 0.0, 0.0],
             [1.0, 0.0, 0.0],
             [0.0, 1.0, 0.0],
             [1.0, 1.0, 0.0],
             [0.0, 0.0, 1.0],
             [1.0, 0.0, 1.0],
             [0.0, 1.0, 1.0],
             [1.0, 1.0, 1.0]])))
        vtkData.SetPoints(vtkPoints)

    elif dataset is vtk.vtkRectilinearGrid:
        vtkWriter = vtk.vtkRectilinearGridWriter()

        vtkData.SetDimensions(2,2,2)
        vtkData.SetXCoordinates(numpy_to_vtk(np.asarray([0.0, 1.0])))
        vtkData.SetYCoordinates(numpy_to_vtk(np.asarray([0.0, 1.0])))
        vtkData.SetZCoordinates(numpy_to_vtk(np.asarray([0.0, 1.0])))

    # file scope definition of Pi
    vtkArray = numpy_to_vtk(np.asarray(3.141).reshape(1,1),deep=1)
    vtkArray.SetName("Pi")
    vtkData.GetFieldData().AddArray(vtkArray)

    vtkWriter.SetFileName(str(dataset) + ".vtk")
    vtkWriter.SetHeader("")
    vtkWriter.SetFileTypeToASCII()

    vtkWriter.SetInput(vtkData)
    vtkWriter.Write()

/endscript

The files generated from this are
# vtk DataFile Version 3.0

ASCII
DATASET POLYDATA
FIELD FieldData 1
Pi 1 1 double
3.141
POINTS 8 double
0 0 0 1 0 0 0 1 0
1 1 0 0 0 1 1 0 1
0 1 1 1 1 1

# vtk DataFile Version 3.0

ASCII
DATASET RECTILINEAR_GRID
FIELD FieldData 1
Pi 1 1 double
3.141
DIMENSIONS 2 2 2
X_COORDINATES 2 double
0 1
Y_COORDINATES 2 double
0 1
Z_COORDINATES 2 double
0 1

# vtk DataFile Version 3.0

ASCII
DATASET STRUCTURED_POINTS
FIELD FieldData 1
Pi 1 1 double
3.141
DIMENSIONS 2 2 2
SPACING 1 1 1
ORIGIN 0 0 0

Note the FIELD preceding the dataset details. All files can be read back into python without problems, but when trying to view each file in using Paraview 4.3.1, things get interesting:

- POLYDATA works.
- RECTILINEAR_GRID produces an error (which, btw, contains a typo), refuses to render anything but displays everything correctly in the Information tab.

ERROR: In /home/kitware/Dashboards/buildbot/paraview-debian4dash-linux-shared-release_qt4_superbuild/source-paraview/VTK/IO/Parallel/vtkPDataSetReader.cxx, line 701
vtkPDataSetReader (0x8874010): Expecting 'DIMENSIONS' insted of: FIELD

- STRUCTURED_POINTS does not produce any errors, still refuses to render but displays everything correctly in the Information tab.

If I now (manually) put the FIELDS definition after the dataset details, everything works fine.

Is this  problem with Paraview, or am I doing something with the vtk format that should not be possible at all? I got the idea from http://www.visitusers.org/index.php?title=Time_and_Cycle_in_VTK_files.

Thanks
Raphael

Software versions:

Paraview 4.3.1 Linux 64bit

Python 2.7.8 |Anaconda 2.1.0 (64-bit)| (default, Aug 21 2014, 18:22:21)
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)] on linux2

vtk.vtkVersion().GetVTKVersion()
'5.10.1'

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtkusers/attachments/20150806/73306656/attachment.html>