2023/05/22 by Fabien Casenave, Casenave, Fabien, Brian Staber +3 · 1 citation
Computer Science · Engineering · #3D Shape Modeling and Analysis #Advanced Multi-Objective Optimization Algorithms #FOS: Computer and information sciences #Machine Learning (cs.LG) #Manufacturing Process and Optimization
paper · pdf · doi:10.48550/arxiv.2305.12871
openalex publication_date 2023/05/22 · openalex created_date 2023/05/24 · openalex updated_date 2026/07/28
This dataset contains 2D quasistatic non-linear structural mechanics solutions, under geometrical variations. A Description is provided in the MMGP paper Sections 4.1 and A.2. The file format is PLAID, see the plaid documentation. The variablity in the samples are 6 input scalars and the geometry (mesh). Outputs of interest are 4 scalars and 6 fields. Seven nested training sets of sizes 8 to 500 are provided, with complete input-output data. A testing set of size 200, as well as two out-of-distribution sample, are provided, for which outputs are not provided. Tips to access the data: After decompressing the downloaded file: dataset = Dataset()problem = ProblemDefinition() problem.loadfromdir_(os.path.join(/path/to/data,'problemdefinition'))dataset.loadfromdir_(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.getfieldnames(): print(f"fn =", sample.getfield(fn))for sn in sample.getscalarnames(): print(f"sn =", sample.getscalar(sn)) print("nodes =", sample.getnodes())print("elements =", sample.getelements())print("nodaltags =", sample.getnodaltags())