Update the trajectory and DCp when self.samples is updated

[chaithyagr]

This issue will track a bunch of features needed for learning k-space sampling trajectories.

For now, what we need is:

• [x] self.samples property setter to have update the corresponding plans to update the trajectory in respective raw_op for each backend.
  • [x] This is doable in finufft and cufinufft easily as there is a set_pts function. @alineyyy will handle a PR for it. @alineyyy for other backends, please add an error for now. This can be done by having error raised in the main setter in the base class.
  • [X] For gpuNUFFT, we need some update on gpuNUFFT side. @chaithyagr will handle this. #133
  • [x] @paquiteau can you detail more about other backends and if want to support for them? (Most dont have autodiff wrt to trajectory anyway).
• [x] Update the density compensators accordingly. This is a bit complicated as it means we need to save how the density was estimated. @paquiteau lets discuss about this and @Lenoush can you please work on it?
• [X] A bunch of tests overall to test this. This is fairly straightforward. But its good sanity tests #133

[chaithyagr]

Please add on if theres anything I missed. Generally I would also prefer sensitivity maps to change. But thats a different ballgame as the sensitivity maps need to anyway change with k-space data also.

[chaithyagr]

All of this is handeled now!