Published 2021 | Version v2
Dataset Open

Darcy Flow Dataset

  • 1. ROR icon California Institute of Technology
  • 2. ROR icon Nvidia (United States)

Contributors

Data collector:

  • 1. ROR icon California Institute of Technology
  • 2. ROR icon Nvidia (United States)

Description

Dataset of input-output pairs of functions governed by Darcy's law, which describes flow through a semiporous medium. The data is saved as keyed pairs of PyTorch Tensors. Inputs (keyed 'x') are a(x), the input function that describes the diffusivity constant at each point on the discretized domain. Outputs (keyed 'y') are u(x), the solution function that describes the flow at each point on the domain. 

Available discretized resolutions: 16x16, 32x32, 64x64, 128x128, 421x421.

Files

Files (12.7 GB)

Name Size Download all
md5:43e85ec9bff754d0f6d3925e26d5a8ac
373.9 MB Download
md5:091575565d8c4ab17f8c3bf2bbb566cb
1.0 MB Download
md5:24c0033f0c0b75f3df546073111ae154
93.7 MB Download
md5:d6600431a21ebae44b05eac601451b17
12.1 GB Download
md5:49027fc1972647915f6c0893a79bbc59
187.1 MB Download

Additional details

Related works

Is part of
Publication: arXiv:2010.08895v3 (arXiv)

Dates

Available
2024

Software

Repository URL
https://github.com/neuraloperator/neuraloperator
Programming language
Python
Development Status
Active