Reading planar sampling output about amr-wind HOT 7 CLOSED

We have been mostly using NetCDF format for sampling output and manipulating the resulting NetCDF files with Python/matplotlib. The images you refer to in the exawind/wind-energy repo is loading the full 3D flowfield in ParaView (@gantech can confirm).

If you want to continue using native format, I have heard that yt is another option to visualize AMReX datasets? We will look at updating the native interface so that you don't have to manually copy files to load them in ParaView. @marchdf any thoughts here on why particle data won't load/render in VisIt?

from amr-wind.

I would be fine with using NetCDF and python/matplotlib. I will look at making the change so that AMR-Wind willl write out sampling data in NetCDF. That seems like the easiest path forward at this point.

I tried yt and Amvis as well, but neither worked. Amvis says that the dataset is 0D and it doesn't know how to display. yt says that it can't load the data.

I'm happy to provide any error output from the various viz tools if that would be helpful at all.

from amr-wind.

@rd-contr NetCDF support must be explicitly activated in AMR-Wind.

cmake -DAMR_WIND_ENABLE_NETCDF=ON -DNETCDF_DIR=<path_to_netcdf_install>

You can see this IPython notebook for an example of how to extract data from the netcdf files.

from amr-wind.

Copy. Thank you for the pointer on adding the NetCDF support. I'll get my binaries rebuilt. Will also check out the iPython notebooks.

from amr-wind.

@rd-contr If you are comfortable with python scripting, then you might also want to look into the AmrexParticleFile in the tools directory. I have not used it a lot after we moved to NetCDF, but you might be able to use this to process the files you already have

from amrex_particle import AmrexParticleFile

# load the 10-th time index 
pfile = AmrexParticleFile.load(10)
# Process the file and return a Pandas dataframe
df10 = pfile()

The Pandas DataFrame has index which is the unique identifier for a given particle. It also has two additional columns set_id and probe_id which you can filter on to get a particular set of sampling probes and the probe_ids start from 0 for each set of sampling points. The columns xco, yco, zco are the coordinates of the particle/probe, and the remaining columns are the field data you had requested.

from amr-wind.

Thanks for the tip on the python code for interacting with AMReX particle files. I will check it out.

from amr-wind.

@rd-contr closing issue due to inactivity. Please reopen if you have follow-up questions.

from amr-wind.