A library for solving differential equations using neural networks based on PyTorch, used by multiple research groups around the world, including at Harvard IACS.
As a result of my excellent learning experience in the Harvard ComputeFest 2024: Physics-Informed Neural Networks, I wanted to contribute to neurodiffeq. The perfect opportunity presented by initiative of @shuheng-liu. I open this pull request to create the chebyshev_uniform method for Generator1D. This should:
Sample randomly from an uniform distribution between 0 and π
Generate the Chebyshev nodes of the first kind from this sample
Map output from (-1, 1) to (tmin, tmax)
I have also updated the documentation and added a test, accordingly.
As a result of my excellent learning experience in the Harvard ComputeFest 2024: Physics-Informed Neural Networks, I wanted to contribute to neurodiffeq. The perfect opportunity presented by initiative of @shuheng-liu. I open this pull request to create the
chebyshev_uniform
method forGenerator1D
. This should:I have also updated the documentation and added a test, accordingly.
Thank you!