You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi,
I have been working on some complex problems with adaptivity and neutronics (using fetch2 which is built on fluidity) and while trying to visualise some arrays I came across a limitation of vector_fields which does not allow for the output of vector field if the dimension is > 3.
Which had me wondering if it would be possible to write a n-dimensional vector_field to a vtk file.
I have had a look at the VTK_Interfaces.F90 file and although the remap_field method will be a bit of an issue, the real problem is vtkwritefvn/ vtkwritedvn. From what I gather these are defined in vtkfortran.cpp and they take as arguments explicit pointers to x,y,z entries of the vector_field.
I think what I am suggesting is possible with some minor changes/ overloading, e.g. SetTuple3 will have to be changed to SetTuple and so will *vx, *vy, *vz to **v.
Do you think this is something that could go in master, because currently I am using an array of scalar fields to output my n x no_nodes vector and it is an absolute nightmare to post-process.
The text was updated successfully, but these errors were encountered:
Oh, that's some horrible old code in there. Yes, absolutely, sounds like a sensible addition. As long as the interface for vtk_interfaces doesn't change, I don't think anyone will care about what you do to libvtkfortran.
Hi,
I have been working on some complex problems with adaptivity and neutronics (using fetch2 which is built on fluidity) and while trying to visualise some arrays I came across a limitation of vector_fields which does not allow for the output of vector field if the dimension is > 3.
Which had me wondering if it would be possible to write a n-dimensional vector_field to a vtk file.
I have had a look at the
VTK_Interfaces.F90
file and although theremap_field
method will be a bit of an issue, the real problem isvtkwritefvn
/vtkwritedvn
. From what I gather these are defined invtkfortran.cpp
and they take as arguments explicit pointers to x,y,z entries of the vector_field.I think what I am suggesting is possible with some minor changes/ overloading, e.g.
SetTuple3
will have to be changed toSetTuple
and so will*vx, *vy, *vz
to**v
.Do you think this is something that could go in master, because currently I am using an array of scalar fields to output my
n x no_nodes
vector and it is an absolute nightmare to post-process.The text was updated successfully, but these errors were encountered: