search

mmuckley / torchkbnufft

A high-level, easy-to-deploy non-uniform Fast Fourier Transform in PyTorch.
https://torchkbnufft.readthedocs.io/
MIT License
204 stars 44 forks source link

3D radial density compensation and stacks of radial density compensation #5

Closed zaccharieramzi closed 3 years ago

zaccharieramzi commented 3 years ago

This tackles #2 (don't know if it solves it though).

I tested the equivalent in my package (tfkbnufft) in TensorFlow and the density compensation worked perfectly.

mmuckley commented 3 years ago

Note: if you find the default behavior undesirable, then we could also remove the default argument altogether to force users to be aware of the new functionality.

zaccharieramzi commented 3 years ago

No no, I think surely you know how users were using it so far, so I will definitely change it back to stacks=True by default in 3D.