Hello, Doc.Li ! I have read your paper Neural Operator: Graph Kernel Network for Partial Differential Equations and it is very cool. Now, I am quite understand the ways that Graph Kernel Network learns the kernel function of PDEs. However, on Page 5 Example 1 Poisson Equation, we can learn a neural network to approximate the kernel function so when we get data of x and a(x),we can get sol. U(x) out quickly. But the neural network we learned is not a explicit function and how could we plot it out in your paper Fig.2?
I would be very grateful if you could show the code about Example 1:1d poisson equation generously.