s2fft
s2fft copied to clipboard
Differentiable and accelerated spherical transforms with JAX
Hi, I am using s2fft in a script where I use healpix maps. Everything works perfectly when using a CPU. However, I now need to use the GPU and it...
Bumps [actions/setup-python](https://github.com/actions/setup-python) from 5 to 6. Release notes Sourced from actions/setup-python's releases. v6.0.0 What's Changed Breaking Changes Upgrade to node 24 by @salmanmkc in actions/setup-python#1164 Make sure your runner is...
Currently the data type of the output arrays of the JAX forward and inverse spherical tranforms in `s2fft.precompute_transforms` are fixed to be of type `complex128` irrespective of the type of...
Running benchmarks of the forward transform (with `mw` sampling scheme) the compile time appears to be growing linearly with the bandlimit `L` ``` forward (method: jax, L: 16, L_lower: 0,...
The `s2fft.precompute_transforms.construct.spin_spherical_kernel` function constructs an array of dimension $O(L^3)$. The currently implementation raises an out of memory (OOM) error when trying to construct the kernel for argument values ``` L...
Currently `so3`, `pyssht` and `healpy` are all test dependencies of the packages as we check consistency of our spherical harmonic and Wigner transform implementations against the implementations in the packages....
Our tests workflow is currently failing on Ubuntu / Python 3.8 - see for example [this run log](https://github.com/astro-informatics/s2fft/actions/runs/15513091938/job/43676429874). This is due to errors when building `so3` a test dependency using...
Currently we separately specify both type annotations for functions in our API and separately include type information in our NumPy style docstrings. Sphinx is able to parse the type annotations...
At the moment while we have some documentation about the [spherical sampling schemes supported in the main README](https://github.com/astro-informatics/s2fft?tab=readme-ov-file#sampling-) this does not give details of the differences between the sampling schemes...