This page releases codes and data of an academic paper, 'On the locality of local neural operator in learning fluid dynamics'. With the provided codes, one can quickly reproduce the training & validation of local neural operator on the task of compressiblem fluid dynamics. Also, two demos for applying the well-trained LNOs to respectively solve the 2-D flows around a circular cylinder and vehicles are provided.
Pytorch 1.6.0
We provide 30 pieces of data samples for learning the compressible Navier-Stokes equations (Re=100, Ma=0.2).
To reproduce the reported results, please download the provided data from
URL: https://pan.baidu.com/s/13ZRXGuaBujk6zlA98tWIlQ
code: sy84
Unzip the data files at /Train_Validation/Data/*
This paper will be published as a preprint on arXiv.org soon.