A library for solving differential equations using neural networks based on PyTorch, used by multiple research groups around the world, including at Harvard IACS.
Hello, I was wondering if the code currently supports solving a BVP of the form $L(u) = v,$ where $L$ is some differential operator, $u(x,y), v(x,y) \in \mathbb{R}^2$ and $x$ and $y$ are scalar coordinates.
I tried to modify the example BVP to be trained with a neural network with a vector output, but the functions in conditions.py and solvers.py assume that the output of the PDE is scalar.
This package is very nicely written, thank you!
Hello, I was wondering if the code currently supports solving a BVP of the form $L(u) = v,$ where $L$ is some differential operator, $u(x,y), v(x,y) \in \mathbb{R}^2$ and $x$ and $y$ are scalar coordinates. I tried to modify the example BVP to be trained with a neural network with a vector output, but the functions in
conditions.py
andsolvers.py
assume that the output of the PDE is scalar. This package is very nicely written, thank you!