Thank you Berk for your answer. You are right, make_vector is even mentioned in some parts of the Paraview Guide that I have read. But my first need was to create a 9-component field to be able to use the Tensor Glyph filter, so I still have to use numpy.
To reproduce the bug: - create a source sphere - create a filter Python Calculator with this formula: numpy.column_stack((Normals[:,0], Normals[:,1], Normals[:,2])) and Apply. - display the field result - in the toolbar, click on Magnitude to see the components: X, Y, Z. It is OK. - edit the formula of the Python Calculator with a formula that changes the number of components: numpy.column_stack((Normals[:,0], Normals[:,1], Normals[:,2], Normals[:,0])) and Apply without changing the name of the result field. - in the toolbar, click on Magnitude to see the components: X, Y, Z. It is not OK. - in the panel Information, the field "result" has 4 components. It is OK. - edit the name of the result field to call it result2 and Apply. - display the field result2, and click on Magnitude to see the components: 0, 1, 2, 3. It is OK. So it seems to be a bug in the drop down button to select the component that is not updated when the formula changes the number of components on an existing field. Christophe 2015-04-02 17:01 GMT+02:00 Berk Geveci <[email protected]>: > Dear Christophe, > > It seems like the documentation got out of date. Thank you for reporting > it. There is also a make_vector() function now. You can pass it 2 or 3 > components: > > def make_vector(arrayx, arrayy, arrayz=None): > """Given 2 or 3 scalar arrays, returns a vector array. If only > 2 scalars are provided, the third component will be set to 0.""" > > > I defer to Utkarsh to explain how to contribute to the tex documentation. > It is definitely very welcome. > > Finally, I am not sure that I understand the comment about the potential > bug. Can you please give the steps to reproduce? > > Thanks, > -berk > > On Thu, Apr 2, 2015 at 5:23 AM, Christophe Bourcier < > [email protected]> wrote: > >> Dear Paraview folks, >> >> It seems there is a mistake in the Paraview User's Guide at the end of >> the section Python Calculator. >> >> To create a multi-component field from multiple scalar fields, hstack >> used to work before Paraview 4.2.0. >> >> But since Paraview 4.2.0, we must use numpy.column_stack. This change may >> have been caused by the great work done in 4.2.0 to be able to use numpy >> arrays as vtk arrays with automatic conversion from numpy to vtk and >> inversely. >> >> Using hstack will create a dataset with many components (number of nodes >> times the number of field in hstack). >> >> For instance, create a sphere source with default properties. >> >> In Paraview 4.1.0, use this formula to create the same field as Normals: >> hstack((Normals[:,0], Normals[:,1], Normals[:,2])) >> >> In Paraview 4.2.0 and 4.3.1, if you use numpy.hstack((Normals[:,0], >> Normals[:,1], Normals[:,2])), you will get a field with 150 components, >> since the default sphere has 50 points. And it may lead to a memory >> oversize for a source with thousands points. >> >> In Paraview 4.2.0 and 4.3.1, the following formula gives the wanted >> result: >> numpy.column_stack((Normals[:,0], Normals[:,1], Normals[:,2])) >> >> Note that you must change the name of the result, otherwise the number of >> components is not updated between two different formulas. This might be a >> bug. >> >> I have not found where the latex documentation is. Is it available >> somewhere as a git repository? What is the policy on contributing to it? >> >> Christophe >> >> >> _______________________________________________ >> 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 >> >> Search the list archives at: http://markmail.org/search/?q=ParaView >> >> Follow this link to subscribe/unsubscribe: >> http://public.kitware.com/mailman/listinfo/paraview >> >> >
_______________________________________________ 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 Search the list archives at: http://markmail.org/search/?q=ParaView Follow this link to subscribe/unsubscribe: http://public.kitware.com/mailman/listinfo/paraview
