AirfRANS_clipped

This dataset is a new version of the dataset originally made available through the library and described in the paper.

It has 3 main differences:

• it is in PLAID format
• all meshes were clipped in a box  [-2, 4] x [-1.5, 1.5] and converted to triangular meshes thanks to Paraview
• A ML4PhySim_Challenge_train split is provided, corresponding to the training set of the ML4PhySim challenge

It has 2 variants:

• AirfRANS_remeshed
• AirfRANS_original

Tips to access the data:

After decompressing the downloaded file:

from plaid.containers.dataset import Dataset
from plaid.problem_definition import ProblemDefinition

dataset = Dataset()
problem = ProblemDefinition()

problem._load_from_dir_(os.path.join(/path/to/data,'problem_definition'))
dataset._load_from_dir_(os.path.join(/path/to/data,'dataset'), verbose = True)

print("problem =", problem)
print("dataset =", dataset)

sample = dataset[0]
print("sample =", sample)

for fn in sample.get_field_names():
    print(f"{fn} =", sample.get_field(fn))
for sn in sample.get_scalar_names():
    print(f"{sn} =", sample.get_scalar(sn))

print("nodes =", sample.get_nodes())
print("elements =", sample.get_elements())
print("nodal_tags =", sample.get_nodal_tags())