Hi Jens,

I don't know about Visit, but the reason SimulationTools prefers to have CarpetIOHDF5::one_file_per_group=no is that it can then infer the filename directly from the variable name. I don't think this is a strict requirement. If your variable names are not related to the filename (for example, setting the parameter to yes) then you just have to manually specify the filename. For example, if you're following the GridFunctions tutorial you would change the line:

phi = ReadGridFunction[$SimulationToolsTestSimulation, "phi", "xy",

Iteration -> 4096, RefinementLevel -> 3]

to:

phi = ReadGridFunction[$SimulationToolsTestSimulation, "phi", "xy",

Iteration -> 4096, RefinementLevel -> 3, FileName -> "phi.xy.h5"]

Barry

On Sun, 7 Oct 2018 at 15:37, Jens Mahlmann <jensmahl@alumni.uv.es> wrote:

Dear all,

I hope I may approach you with a rather general question on output
generation. For visualization, I find it convenient to use both
SimulationTools for Mathematica as well as VisIt. Both are fine with
reading hdf5 files and the interface to Carpet data is in general very nice.

However, I find that depending on if I activate
CarpetIOHDF5::one_file_per_group or not, I have difficulties of opening
the files in one or the other (with yes I am great for visit but not for
SimulationTools and vice versa).

Especially, I would like to have the 3D output in one file per group,
and the 2D output not. Is there an option to do so? Or does anybody have
a best practice for a student of how to overcome this in post process? I
want to avoid running all simulations twice...

Thank you and my best!

Jens

_______________________________________________
Users mailing list
Users@einsteintoolkit.org
http://lists.einsteintoolkit.org/mailman/listinfo/users